Abstract
This article examines a content analysis of the televised political debates prior to the 2012 elections in Slovakia. The analysis is based on the functional theory of campaign discourse by William L. Benoit. The goal of the work is to test the theory and its applicability in contexts other than the political system of the United States. In order to achieve this goal, we also make use of the findings of previous studies from the non-American environment. We have found that if Benoit’s methods were strictly observed, more than 30% of the content of debates would not be included in the analysis. Therefore, its original categories of content analysis should be expanded. The Slovak debates differ markedly from the American model, but at the same time differ from the kind of debates found in a consensual political system. The area of the greatest difference is the use of various models of discursive functions in the debates.
Introduction
Television has become one of the key channels of political communication, offering political actors means of speaking directly to the voters. In countries that regulate political advertisements on television, political debates function as the only way for political actors to directly influence their TV campaign coverage. Political debates are an important form of political communication because they have become major events that enjoy widespread interest among the public and receive broader press coverage than other campaign events (McKinney and Carlin, 2004: 204).
In this article, we analyse TV debates in Slovakia during the 2012 early elections campaign. The goal is an analysis of the Slovak debates, which have not yet been systematically analysed, based on the functional theory of political campaign discourse designed by William Benoit (e.g. Benoit, Blaney and Pier 1998; Benoit, Pier and Brazeal 2002; Benoit, McHale, Hansen 2003). We will apply this theory to conditions different from those for which it has been constructed. There are many differences between Slovak and American political systems. The most basic and most visible differences lie in the fundamental institutional settings: Slovakia is a parliamentary democracy, with a multi-party system, whereas the United States has a presidential system, with a two-party system. Therefore, our secondary goal is to contribute to the theory’s general development and to find out whether the theory and the method developed in the US context is applicable in different political environment.
The functional theory of campaign discourse
The most influential attempt at systematic analysis of political debates based on a specific theoretical construct is the functional theory by William Benoit (Benoit, 2005, 2007; Benoit et al., 2002). Benoit’s theory is based on a set of basic principles (Benoit, 2007; Benoit and Klyukovski, 2006; Benoit and Sheafer, 2006):
Elections are a comparative act in which voters select the candidate they most prefer from among all the possible candidates.
Candidates must therefore differentiate themselves from their opponents.
Candidates differentiate themselves through messages sent during the election campaign.
The campaign discourse consists of three discursive functions: acclaim, attack, and defence. Acclaim represents positive statements by a candidate about himself, attack criticises the opponent and defence represents reaction to attacks.
These discursive functions work through two possible areas: policy (issues, government activity) and the candidate’s character (image). The theory breaks these themes down further: policy is divided into past deeds (accomplishments or failures), future plans (specific campaign promises) and general goals (general objectives). Character involves personal qualities (personality traits), leadership abilities (experience in office) and ideals (values, principles).
Candidates must win a majority, or a plurality, of votes.
We will focus on Assumptions 4 and 5. According to functional theory cost–benefit analysis, acclaims increase a candidate’s benefits, whereas attacks increase an opponent’s costs, and defences reduce alleged costs within a political debate (Benoit and Klyukovski 2006). In the case of attacks and defences, however, there is a possibility of unexpected side-effects on the side of public perception. Attacks, for example, might lead to a backlash effect (if voters perceive attack as unethical mudslinging it may instead damage the attacking candidate’s support). Defence has also certain downsides. Reacting to the attacks of an opponent can take a candidate off message. Responding to the attacks of others can appear less pro-active. Therefore, acclaim should be the most frequently used discursive function.
The theory assumes that policy will be a more frequently mentioned topic than the candidate’s character. Within the policy topics, general goals are more often mentioned in the context of acclaim than attack; within the character topics, ideals are more often used to acclaim than to attack. General goals are mentioned more often than specific future plans.
Benoit concludes that the American presidential debates up until 2004 were, in total, more positive than negative, consisting of 57% acclaims, 35% attacks and 8% defence. Policy (75%) was a much more frequent topic than character (25%). Differences were identified between the discourse of the typical incumbent and challenger: incumbents tended to make more use of acclaims (64%) than challengers (51%); incumbents attacked less (25%), while challengers used attack 44% of the time. On the other hand, incumbents tended to spend more time defending against attack (12%), while challengers used defence only 6% of the time. Incumbents referring to past deeds tend to produce acclaims when talking about their own policies (72% vs 28% incumbents), while challengers tended to attack (82% vs 18% by incumbents; Benoit and Sheafer, 2006: 282).
Contextual factors and application of the functional theory
Functional theory was applied to various countries, for example, Ukraine, Israel, Poland and Finland. The Ukrainian politicians (Benoit and Klyukovski, 2006) tended to use attacks instead of acclaims and focused on the candidate’s character instead of policy. In comparison with the United States, in the Israeli debates, the candidate’s character was debated more often, but candidates’ ideas were mentioned rarely; on the other hand, future plans were mentioned more often than general goals (Benoit and Sheafer, 2006). Despite these variations, which the authors ascribe to cultural differences, the theory was confirmed as internationally applicable, mainly as a result of the homogenisation of political campaigning. The general applicability of the functional theory was also supported by a study of Polish presidential debates (Dudek and Partacz, 2009). Despite certain differences in the results, caused by momentary circumstances, the debates’ discourse is considered to be standardised across borders.
However, Benoit’s conclusions are not always completely accepted. Isotalus and Aarnio (2006) seek an approach to a content analysis of political debates in multi-party systems, such as Finland. They assume that the discourse is to a great degree country-specific. In the Finnish case, they talk about discussions instead of debates. Their main argument states that multi-partyism offers a different structure for the discourse of political debates, since these are usually multi-party debates. First, there is a generally lower level of attacks thus, logically, less need for defence. Therefore, they disagree with the conceptualisation of the debates in terms of attack and defence. Moreover, in a system with a tradition of coalition governments, debate participants often cannot use much negativism because of the risk of alienating possible coalition partners (Isotalus and Aarnio, 2006: 65). These useful characteristics can be extrapolated to multi-party systems with a tradition of coalition governments in general.
They add two discursive functions to the functional theory: agreement and disagreement. By agreement, candidates express their mutual willingness to maintain cooperation with one another. On the other hand, disagreement is based on opposing opinions and political programmes. Another key factor influencing the use of these discursive functions in political debate should be a party’s membership in either the governing coalition or the opposition.
Another modification of Benoit’s model is a category attempting to describe current political circumstances. Besides past deeds and statements referring to the future, the theory should take into account that a large part of a debate revolves around references to the current situation in various policy areas (Isotalus and Aarnio, 2006: 66–68).
Isotalus (2011) also analyses the 2006 Finnish presidential debate held between two presidential candidates. He insists that the theory still needs further development in order to achieve higher inter-cultural validity – not every statement fits the categories defined by the functional theory. For example, some statements can be used as a neutral description of a political situation. The analysis of the presidential debate reveals some other deviations from Benoit’s original assumptions because of the strong role of the debate moderators in the Finnish debates. Also Isotalus points out that a large part of the debate (21%) could not be coded using the discursive functions set forth by functional theory, which clearly distorts the results as well.
We decided not to strictly apply Benoit’s categories because, if we did so, we would assume that a large part of the discourse would not fall into any of the pre-defined functions. Various statements would have to be coded to correspond exactly to the functional theory, which would distort the analysis by cramming every statement into one of the three predetermined categories. We thus approach Benoit’s functional theory as a methodological instrument for analysing election campaign debates, but we assume that, on the basis of studies of the non-American environment, the robustness of its theoretical presuppositions is more or less limited to the circumstances of American politics.
With new categories of content analysis added, we pose the following research questions: What is the ratio of the individual functions, defined as acclaim, attack, disagreement, defence, agreement and neutral statement, in the election debates in Slovakia? What is the ratio between policy and character utterances in these debates? Does the format of the debate relate to the discursive pattern of the participants? Do the discursive patterns differ between opposition parties which, during the last electoral term, served in the coalition government?
Context of the elections
Slovakia uses a party-list proportional representation electoral system to elect 150 members of a parliament. Political parties must obtain more than 5% of the votes in order to win seats. Five parties were represented in the parliament after the 2010 elections: SMER-SD, Smer – Social Democracy; SDKÚ-DS, Slovak Democratic and Christian Union – Democratic Party; SaS, Freedom and Solidarity; KDH, Christian Democratic Movement; MOST, Bridge and SNS, Slovak National Party.
The elections were won by SMER, but it was the right-wing parties – SDKÚ, SaS, KDH, and MOST – that formed a coalition government. Another party – OľaNO, Ordinary People and Independent Personalities – was formed later by several candidates originally elected through SaS ballot. For the purposes of this article, we will consider it a coalition party.
In the fall of 2011, there was a government crisis in which one of the coalition parties, SaS, refused to support the so-called Euro-fund in parliamentary voting. On 12 October 2011, the government failed to garner the 76 votes to win a vote of confidence which Prime Minister Iveta Radicova linked to the vote on the Euro-fund. On 19 October 2011, the chairman of the National Council called for early elections which were held on 10 March 2012.
During the electoral campaign, the parties constantly defined themselves in contrast to one another. Coalition parties SDKÚ and SaS clearly showed their unwillingness to work with the opposition SMER and SNS. MOST rejected working with the SNS, and the SNS itself came out against all the coalition parties. On the other hand, SMER did not refuse to cooperate with any of the parties after the election. As we can see, the Slovak party system was divided into two camps: (1) coalition parties, which would prefer working together, with two of the parties clearly stating their disinclination to work with opposition parties and (2) opposition parties that would also prefer forming a coalition with each other, with one party clearly stating their unwillingness to work with any of the former coalition parties. After the election in March 2012, the government was formed by only one party – SMER, which filled 83 out of 150 seats in the parliament. The party system was clearly polarised with two poles of the party competition defining themselves against each other. In this context, it is possible to consider the centre-right coalition parties as incumbent parties which were challenged by the opposition parties.
Method
We analyse four debates which took place on the public broadcast television Rozhlas a televízia Slovenska (RTVS) during the official campaign period – 21 days before the elections. Only two of them to were ‘true’ election debates, including representatives from all of the relevant parties, and covering both the actions of the ending government and the parties’ future plans and goals. One of the debates included representatives of all relevant political parties, a total of seven debaters. The other debate did not include OľaNO representatives. We decided not to analyse debates that had been broadcast by commercial television companies because during a pretest we had realised their format was closer to an interview than a discussion: (1) sometimes there was only one guest and (2) even if there were two guests, confrontation and discussion between politicians was limited, dialogue took place only between the moderator and the individual politicians separately.
Two more debates were held during the campaign period. One of them included only two politicians: one from the governing SDKÚ-DS and one from the opposition SMER. The second debate featured three debaters (SDKÚ, SaS and SMER). However, both of these debates were focused more on the current issues than the election.
We performed a content analysis based on Benoit’s model. Because of a specific multi-party political system and coalition practice in Slovakia, we also reflected findings from the Finnish environment (Isotalus, 2011; Isotalus and Aarnio, 2006). The debates were unitised into single themes (assertions, claims, arguments, statements), which were defined as units expressing a single idea in the debate.
Then, we categorised them according to functions: acclaim, attack, defence, agreement, disagreement and neutral statement. Acclaims show the candidate in a positive light, or it can be a specific issue that the candidate wants his own party to pursue in the future. Attacks show the target subjects in a negative light. Defences are basically responses to the attacks by opponents or the debate moderator. Agreement is simply acceptance or acknowledgment of what the previous speaker has said. Disagreement is defined as not sharing the same opinion about what has been said. Neutral statement of a fact mostly presents descriptions of the situation, or introductions to specific issues (Table 1).
Examples of discourse functions.
VAT: value-added tax.
The topics were classified as policy, character and other. Policy themes concern the actions of the government, or policy issues. We divided these into past deeds, present situation, future plans and general goals. Character themes have to do with traits, characteristics or abilities; we divided these into personal qualities, leadership skills and values/ideals. ‘Other’ is simply cases that cannot be categorised according to any of the above. Our pretest showed that some of the discourse turns might not belong to any of the original categories. At the same time their appearance was not too frequent and their focus was so diverse that there would be no reasonable point in creating new specific categories. We also kept track of temporal orientation. Orientation can be to the past, present or future.
Results
Out of the overall number of themes (N = 447), 43.4% were related to the past, 31.8% to the present situation and 24.8% of the identified themes referred to the future. This justifies the concerns about insufficient coverage of the debate under Benoit’s basic theory, which does not take into account orientation on the present. If we had not introduced this category, we would have been unable to code almost a third of the debate.
The data showed that the ratio of various functions in the debate varied according to debate format. In the overall sample, acclaims (32.9%) predominate followed by attacks (21.9%), as the original functional theory predicts. However, next is disagreement (17.9%), and only after that we observed defences (12.3%). Moreover, neutral statements (11.6%) were also a significant part of the debate. If we compare the overall numbers of positive and negative messages, we find that the result does not correspond well to the logic of Benoit’s theory. Overall, the negative statements predominate, that is, attacks and disagreement (together 39.8%). Positive statements occur somewhat less frequently (36.3%). In reference to Benoit’s theory, we note that defences were not exclusively used in reaction to attacks by political opponents, but also in response to questions by the moderator.
In the case of the two debates classified as election campaign debates, we identified 39.8% mostly acclaims. Combined positive statements made up 42.7% of these debates, the negative was 38% and 9.5% was neutral. The data from the debates in which representatives of only three parties were present showed that neutral discourse appeared in 13.5% of all cases, but negative statements were slightly more frequent (35.9%) than positive ones (31.4%).
In the debates between two politicians, attacks and disagreements clearly predominated (28.6% and 21.4%). Altogether 16.7% of the debate was neutral. A total of 50% of the debate was negative and only 20.3% positive. We thus observe that the use of discursive functions in the Slovak case we are examining differs according the format of debate (Table 2).
Discourse functions in 2012 Slovak debates.
Within the debate, topics policy was discussed 79.4% of the time, but character only 2.5% of the time; in six cases, personal qualities were mentioned and in five cases leadership skills. As for policy themes, most frequently mentioned were past deeds (45.4%), general goals followed (30.1%) and then references to the present situation (18.6%). Future plans was the least-frequent policy theme (6%). As for references to the future, there was a significant prevalence of general goals compared to the less-frequent detailed proposals. For the purpose of making acclaims, general goals (44.7%) were the most frequent tool, which lines up with the predictions of functional theory. In the case of attacks, references to past deeds (69.3%) were used most frequently. A total of 18.1% of the debates was coded as ‘other’, since neither policy nor character was clearly the topic of debate (e.g. references to the upcoming elections).
Of the overall 447 statements coded, 287 were made by coalition politicians and 160 by the opposition. Coalition politicians mostly used acclaim (37.6%) and least often agreement (3.8%). Other types of discourse were relatively balanced, with only minimal differences in some of the debates.
Opposition debaters used mostly attacks (33.1%), acclaim (24.4%) and disagreement (23.8%). Because coalition politicians did not attack aggressively, defence was a relatively small share of these debates (8.8%). This shows that while coalition politicians tried to highlight their successes, the opposition focused on criticising their opponents. Another important fact is that the coalition politicians attacked only the opposition parties and not one another. This was also true for the opposition parties. When the parties within the two camps differed, they resorted to disagreement rather than attack.
The specific distribution within the individual parties only underscores this fact. While the opposition parties SMER and SNS tended to attack, the coalition parties preferred to use acclaim. An interesting case is an OľaNO representative – he supported the coalition but spoke more like the opposition. We can interpret this as an effort to present his party mainly as a viable alternative to all parties (Table 3).
Discourse functions used by political parties across the debates.
It is interesting that 70% of SMER’s attacks were on the government as a whole. All other political parties more or less directed their attacks towards specific opponents. For example, the strongest coalition party SDKÚ directed some 82% of its attacks directly at SMER (Table 4).
Topics used by political parties across the debates.
Conclusion
Our analysis shows that the most frequently used functions in the Slovak debates in 2012 were acclaims, followed by attacks. This corresponds to the functional theory of political debates. However, our analysis also added new categories such as agreement and disagreement. These played a significant role, comprising 21.3% of the discourse during the debates. We also included another modification – a category called ‘neutral statement’, which in our data came out to 11.6% of the debates. If we were to stick to Benoit’s set of categories, we would not be able to work with these new functions, and there would be a significant distortion of the analysis. Isotalus (2011) shows that in an analysis of the 2006 Finnish presidential debate, a total of 21% of statements could not be assigned to any of the three main classifications. In our case, 32.9% of the themes would not be classifiable.
Our data revealed differences in the discourse according to the debate format. The debates were held in several forms differing in the number of represented parties. The most distinctive feature is the predominance of negativity in the debates of two candidates; attacks and disagreements were more frequent than in the other debate formats.
Besides the format of the debates, another set of factors were the relationships between the participants within the party system. We see that the Slovak case is consistent with Benoit’s theory: incumbents make use predominantly of acclaims, while challengers tend to resort more to attacks. Relationships between individual parties also matter. Coalition partners did not attack one another, just as opposition parties did not mention each other.
An interesting feature of the debates was the small number of character themes. The great majority of the debate discourse was about policy. A total of 18.1% of the debates could not be coded either as a policy statement or a character theme. Benoit took into account only statements about policy and the character of politicians. The Slovak case, however, shows that this need not be the case, and that the debate is not limited to just those two categories.
Our research shows that functional theory can serve as a methodological instrument but categories capturing the real debate discourse of the particular country ought to be modified. We assume that the debates differ depending on the party system of the country, the format of the debates, the number of participants, the individual communicative style and the political culture of the country in general.
Slovakia represents a specific case in which the cultural and political context determines the shape of political debates. Because it is a system with multiple parties, we can identify elements corresponding to the more moderate discourse of multi-partyism, such as agreement and disagreement. However, Slovakia, unlike Finland, does not represent a consensual democracy with reluctance to engage in political attacks; instead, attacks are an important part of the debates. Another distinctive feature of the Slovak discourse is that it seldom focuses on the character of the political actors. And the topic of political ideals is completely absent. Slovak debates could be described as very particular, practical and focused on policy. Unfortunately, it is not possible to derive more concrete conclusions from our data. We consider the construction of complex relations between the political discourse, on one hand, and political culture and show format on the other hand as an interesting research question for further inquiry. However, data based on a single case study make it impossible to estimate more specifically how these variables affect the form of the political debates.
We are aware of the limits of our research, and further analysis of parliamentary elections debates would be necessary. Moreover, presidential debates would offer more suitable conditions for further comparison of the Slovak political debates with those of other countries. This indicates that there is plenty of space for further research. Nevertheless, we believe our findings present some preliminary conclusions. Also, we find it necessary to deal with some further methodological challenges: How to deal with turn-arounds, which are somewhere between attack and defence. This refers to a situation where one debater responds to the attack of another debater with a counter-attack. Another challenge is how to code statements that the theory does not include, such as insults or name-calling. We believe that further research can successfully solve these problems and contribute to further development of the functional theory of political debates discourse as such.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.