Masarykova univerzita

ah.Conclusions

This thesis investigates linguistic means that contribute to a higher degree of speaker’s involvement in political interviews. Speaker’s involvement manifests itself in the modification of the illocutionary force which may be accentuated, attenuated or modalised. As already mentioned in the introduction of this thesis, some studies claim (cf. Tannen 1985; Chafe 1982, 1984; Elias 1987; Besnier 1994; Katriel and Dascal 1989, Gumperz 1992, among others) that formal communication displays a low degree of involvement, whereas informal interaction is characterised by a high degree of involvement. It is also thought that female speakers are more indeterminate and vague and they do not speak matter-of-factly and to the point. By contrast, it is claimed that male speakers are more direct and precise in their expression.

Based on the literature dealing with involvement (Tannen 1984, 1985; Chafe 1982, 1984; Gumperz 1992) and with language and gender (Holmes 1995; Coates 1993; Lakoff 2003), the following hypotheses were formulated:

hypothesis I

The genre of political interview is detached and impersonal and it shows features of a low-involved style as is typical of any other type of formal interaction.

hypothesis II

Female politicians are more indeterminate and vague in their expression and they do not speak to the point. The expression of male politicians is matter-of-fact and they express themselves more precisely than female politicians.

The aim of the present thesis was to confirm or reject the validity of these hypotheses in the genre of political interview.

This thesis understands political interview as a genre of political discourse since it has its specific communication goals and conventionalised arrangement with certain requirements on language use, degree of formality, argumentation, and topics discussed. The main purpose of a political interview is to inform the public about the political situation in the country, internal problems, international politics, etc. Another aim politicians want to achieve is to persuade their listeners, to retain and gain voters. To fulfil this aim, politicians use linguistic means showing involvement with their propositions.

The corpus of political interviews analysed in this thesis contained 40 interviews (20 interviews with male politicians, 20 interviews with female politicians, both of the same extent) which were downloaded from the webpages of American and British TV and radio stations. An analysis of prosodic means was not the subject of this study, therefore I worked only with the transcripts of the interviews. The transcripts may contain mistakes since I used them as they were published on the Internet. These mistakes may have been caused by editing or by the speaker but they were not corrected by the editor since they belong to a typical feature of spoken discourse. The topics debated in interviews range from current affairs and internal issues in the UK, international politics, and presidential campaign to elections in the USA.

It should be pointed out that only utterances of politicians were analysed. After the preliminary count of linguistic means of speaker’s involvement, the number of these means used by interviewers was very low (only about 200). The reason why interviewers do not use so many involvement-showing linguistic means is that their questions are prepared in advance, and they continue to employ very few expressions of involvement even if they react to politicians’ replies immediately. Another reason why interviewers do not use these means too often is that their role in the interview is different from that of politicians. They lead the interview, ask questions but they do not aim at persuading voters or asserting themselves in front of their audience. For these reasons, the language of interviewers was excluded from the analysis.

The concept of involvement is very diverse and linguists approach it from various angles. Chapter 2 of this thesis discusses the view of interactional sociolinguistics which emphasises the social organization of involvement. Involvement has also been studied in discourse analysis that focuses on variation between spoken and written discourse in connection with involvement (cf. Chafe 1982; Tannen 1985). The present analysis reveals the importance of context when interpreting pragmatic functions of linguistic means of involvement since these means may have different pragmatic functions in different contexts. This context-boundness has also been stressed by Tannen’s contextualization hypothesis and also by the scholars of the Prague School (Mathesius 1982 [1942]; Firbas 1992).

When dealing with the differences between spoken and written discourse, Tannen defines the concepts of high and low involvement. It is often stated that a high degree of involvement is connected only with informal discourse. One of the proposed hypotheses in this thesis is that formal interaction belongs to low-involved styles is also valid in the genre of political interview:

hypothesis I

The genre of political interview is detached and impersonal and it shows features of a low-involved style as is typical of any other type of formal interaction.

A typical feature associated with a low-involved style is maintaining social distance between interactants. Language which is used by participants in political interviews is depersonalized, detached and less immediate. However, these features of a low-involved style were not found in the genre of political interview to a great extent, by contrast, there are linguistic means, namely, boosting devices, hedging devices and modal expressions, which contribute to a high degree of speaker’s involvement.

Politicians use phrases relating to the speaker, such as I think, I don’t think, I mean, I believe, I know, and I agree, which express subjectivity of the speaker and in this way, they contribute to a higher degree of speaker’s involvement. All these means function pragmatically as speaker-oriented boosters or speaker-oriented hedges, however, their frequency of occurrence is much higher when used as boosters, which means that they stress subjectivity and self-confidence of the speaker, rather than his/her uncertainty and indeterminacy. Subjectivity is not, of course, the only means showing involvement. These phrases also promote trustworthiness and confidence of the speaker in front of their audience, which also strengthens the involvement of the speaker. Apart from performing pragmatic functions, I think, I don’t think and I mean are also modal means, in concrete terms, they express epistemic possibility. According to the pragmatic function of a particular linguistic means, epistemic possibility expresses either subjectivity (when the expression functions as a booster) or indeterminacy (when the expression functions as a hedge).

Politicians do not produce only speaker-oriented statements that emphasise or attenuate their subjective views but they also show involvement with the content of the message. For expressing this type of involvement, speakers use either discourse-organizing boosters or content-oriented hedges. These devices may accentuate (discourse-organizing boosters) or attenuate (content-oriented hedges) the propositions expressed.

When politicians want to show involvement relating to the hearer, they use either hearer-oriented boosters or hearer-oriented hedges. These expressions again, depending on the context, accentuate or attenuate the illocutionary force of speech acts. Politicians do not employ them as frequently as speaker- and content-oriented expressions, which means that they predominantly focus on themselves and on the content of their messages and not so much on the hearer.

Boosters in the corpus outnumber hedges. There appear 3,449 boosters in the whole corpus (see Table 3, page 94) and 1,320 hedges (see Table 14, page 133). As already pointed out, one linguistic means may function as a booster or as a hedge. The decisive factor for classifying an expression as a booster or a hedge is the context of the utterance, otherwise, their pragmatic function cannot be determined properly because these expressions are context-sensitive.

The frequency of occurrence of boosting devices in the corpus was examined in categories delimited by their relationship to discourse meaning, which means that boosters were divided into three groups: speaker-oriented, discourse-organizing and hearer-oriented (see Table 4, page 95). Speaker-oriented boosters are the most numerous category (1,968 instances), which indicates that politicians concentrate on emphasising their attitudes and opinions and try to persuade the public that they are the right persons they should vote for. Speaker-oriented boosters show a high degree of speaker’s involvement. Discourse-organizing boosters emphasise the content of the message, which may be viewed as helping the listeners to orientate in the message of the speaker better since they put emphasis on the parts of discourse which are important. There appear 1,078 occurrences. The least frequent category of boosters are hearer-oriented boosters, with 403 instances. This means that politicians prefer concentrating on strengthening their position in front of their audience and on the content of the message.

Boosters were also divided into groups according to their pragmatic functions. They are summarized in Table 6 (page 101) and Figure 2 (page 102). The most frequent function in the corpus is content-oriented emphasis with 923 occurrences in total, followed by subjectivity with 823 occurrences. Male politicians produced more instances of content-oriented emphasis than female politicians. However, this function is not the most frequent one in females, who produced more instances of subjectivity. It means that female speakers concentrate more on presentation of their opinions and thus on influencing the audience, while male politicians focus on the content of their messages in a greater extent. It is generally thought that politics is a “male issue” and even if females assert themselves in this area, it does not usually last for a long time. Therefore, to be perceived by the audience as competent for the position of a politician, female politicians must show confidence and certainty. Male politicians do not have to defend their position in front of their audience to such a great extent, thus they concentrate on accentuation of content of their utterances, which is significant for the listeners. The third most frequent pragmatic function of boosters is the emphasis of the degree of a certain quality, which stresses a positive of negative quality of the following expression. It occurred 425 times in the corpus. Assurance appeared only 421 times in the corpus, which indicates that politicians attempt to save their face in front of their viewers. They do not want to show a high degree of assurance since their claims could be regarded as untrue and politicians would sound unreliable. Intensification by repetition, which occurred 378 times in the corpus, contributes to greater emphasis put on politicians’ claims. A frequency of occurrence similar to that of intensification by repetition appears in hearer-oriented emphasis, which was produced in 370 instances. Linguistic means expressing this function turn to the hearer and stress important parts of the speaker’s message, which provides the hearer with a better orientation. These means may also ask indirectly for confirmation of speaker’s messages. The least frequent function appearing in the corpus is agreement with 109 occurrences. This function is more typical of informal interaction than of political interview because interviewers ask questions which they suppose the politicians will oppose. They want to make the discussion more attractive for the audience, therefore they ask challenging questions which are difficult to answer and the politicians have to defend their views in front of the public.

As is evident, the high number of occurrence of these pragmatic functions of boosting devices contributes to a higher degree of speaker’s involvement in this genre. This fact contradicts hypothesis I, which says that political interview shows features of a low-involved style.

If the speaker wants to attenuate the illocutionary force, s/he uses hedging devices. Even though the number of hedges in the corpus is significantly lower than that of boosters, they are important linguistic devices that modify the illocutionary force of utterances. Politicians aim at showing responsibility and confidence, they want to show themselves in a positive light. When using hedges, they express assumption and uncertainty, which weakens the illocutionary force of utterances. As with boosters, the number of occurrences of hedges according to their discourse meaning was investigated since it is convenient for determining their pragmatic functions. There are three categories of hedges: speaker-oriented, hearer-oriented and content-oriented. As shown in Table 15 (page 134), the most frequent is the category of content-oriented hedges with 857 instances of occurrence. This type of hedges decrease responsibility of the speaker for his/her assertions since when using maybe, probably or possibly, the speaker says that something may or may not be the case. Therefore, they may be used as a face-saving strategy of the speaker. The number of speaker-oriented hedges is substantially lower than the number of content-oriented hedges; they were produces in 304 instances. These hedges express hesitation of the speaker, which is a sign of his/her uncertainty or withholding some information. However, without sufficient context and background information, it cannot be specified properly if they indicate uncertainty or withholding information. Hearer-oriented hedges are the least frequent category of hedges in the corpus. There are altogether 159 instances. They express uncertainty related to the hearer. As with hearer-oriented boosters, this category does not appear too often since speakers concentrate more on the content of their utterances, in this case on hedging this content.

Hedges occur in nine pragmatic functions in the corpus, all of them are summarized in Table 17 and Figure 5, page 137. The most numerous function is attenuation of the forthcoming message, which was produced altogether in 735 instances. Politicians weaken the illocutionary force of their utterances not only when they are hesitant or uncertain but also for the reason that they do not want to sound too reserved and detached from the audience. They attempt to be closer to their voters. Politicians also try to soften negative facts they have to convey to their listeners. Assumption is the second most frequent function of hedges in the corpus, however, its occurrence is considerably lower than that of attenuation. It appears 243 times in the corpus. Assumption expresses uncertainty of the speaker about the proposition. In some cases, the speaker does not want to reveal the information to the public and wants to be evasive, but sometimes, s/he cannot predict the development of the situation so s/he must be vague and make only assumptions. Hearer-oriented uncertainty occurs 107 times in the corpus. It is connected with the uncertainty of the speaker concerning the hearer’s reaction and attitude towards the speaker’s proposition. Unspecified reference belongs with 50 instances of occurrence to less frequent pragmatic functions of hedges in the corpus. It is expressed by kind of and sort of therefore it is connected mostly with informal interaction. The fact that it also appears in formal interaction to a certain extent may be related to conversationalization of media discourse. In terms of frequency in the corpus, hesitation is similar to unspecified reference because it appears in 48 instances. However, there is a substantial difference between male and female speakers as regards the distribution of this function. Females used it to a lower extent than males, which relates to the effort of female politicians to persuade the voters that they are the right persons for the political position they exercise. They do not want to be evasive too much when answering the interviewer’s questions. Content-oriented uncertainty with a total of 41 appearances but only 14 appearances in females confirms the claim that females feel the need to show responsibility for their claims in order to assert themselves in front of the public, therefore, they avoid using expressions showing their uncertainty. Negative politeness is connected with indirectness of speaker’s expression in the corpus. Politicians do not say directly that they are opposed to something, instead, they use phrases like I would disagree or I don’t think to hedge the utterance and save his/her face in front of the audience. This function occurred 37 times in the corpus. Detachment, another pragmatic function performed by hedges, appeared in only 32 instances in the corpus. It relates to the fact that expression of speaker’s involvement is more typical of the genre of political interview than detachment. Politicians attempt to influence and persuade their audience, therefore, they cannot be detached. They are very well aware of the fact that being closer to the audience is better if they want to be successful. Therefore, their language is not so formal but, by contrast, a certain level of informality is apparent in the language of politicians. As already stated, they attempt to establish a closer relationship with their viewers. The least numerous function of hedges is evasiveness. It appeared in 27 instances in total, out of this number only twice in female speakers. This is a somewhat surprising fact, since evasiveness is typically connected with political discourse. It relates to the tendency of speakers to avoid responsibility for their statements. Mere two appearances in females indicate that they attempt to control their language and speak to the point. It is again connected with defending their position in front of the public.

Even if the number of occurrences of hedges is lower than that of boosters, they also contribute to a higher degree of speaker’s involvement since they attenuate the illocutionary force of utterances. This attenuation enables politicians to show uncertainty, assumption, detachment, or evasiveness. However, politicians do not want to be too detached or evasive since they would not have confidence of their voters. In sum, it is another proof that contradicts hypothesis I.

This thesis applies a wider approach to speaker’s involvement, therefore, the concept of modality was also included. This study follows the traditional division of modality into epistemic and deontic types. However, this division could not be applied to all instances of modality found in the corpus so the category of circumstantial possibility had to be included. This category has been defined by Huddleston and Pullum (2002). I have proposed a new class of modality, namely, epistemic attitudinal modality, since there were instances in the corpus which could not be included in any other type.

There appear altogether 2,203 modal expressions in the corpus (see Table 27 and Figure 7, page 178). The types of modality and number of their occurrence are summarized in Table 28, page 179. The most numerous category of modality is epistemic possibility with 1,061 instances. It is expressed by modal means that, according to the context, function pragmatically either as boosters or as hedges. When these modal means accentuate the illocutionary force of the propositions, they express subjectivity of the speaker, which should, again, influence the voters. By contrast, if the modal means hedge the propositions, politicians express uncertainty, assumption and doubt. Deontic necessity is the second most numerous modality in the corpus, however, compared with the frequency of epistemic possibility, it appeared to a much lesser extent. It was produced 565 times in the corpus. This type of modality is connected with showing detachment of the speaker. Politicians do not want to be responsible for their claims, therefore they prefer using the forms have to or have got to to the modal verb must. Using have to/have got to, the speaker is not involved in the proposition expressed and s/he is not responsible for it. Detachment is one of the pragmatic functions of hedging expressions and at the same time, it is connected with the modal functions of deontic necessity and circumstantial possibility. Epistemic attitudinal modality occurred in 369 instances in the corpus. I have proposed this type of modality since the means expressing it, namely, really, frankly and actually, cannot be included in any other type of epistemic modality. They express the speaker’s attitude to the proposition as other types of epistemic modality but they express neither epistemic necessity, possibility nor assumption. These modal means contribute to a higher degree of speaker’s involvement in that they attenuate or accentuate the illocutionary force, depending on the context. Thus, as with other epistemic types, this proves that pragmatic and modal functions are interrelated. Circumstantial possibility was produced 98 times in the corpus. This type of modality expresses possibility that may be fulfilled under certain circumstances. Circumstantial possibility has only been mentioned by Huddleston and Pullum (2002), to my knowledge, while Palmer does not refer to it at all. Epistemic necessity belongs with 87 occurrences to the least numerous modalities in the corpus. This relates to the fact that this function expresses assurance of the speaker and politicians do not want to express confidence about their propositions to a great extent but rather, they leave some space for modifying their assertions, which may also be regarded as a face-saving strategy. The function with the lowest number of occurrence in the corpus is deontic possibility, which appeared only 23 times in the whole corpus. It expresses permission, therefore this type of modality is not so frequent in political interviews. It is more typical of informal interaction.

As the present analysis shows, pragmatic and epistemic modal functions correlate. They express both epistemic stance and accentuation or attenuation of the illocutionary force. Thus, it may be claimed that these functions express very similar meanings.

From the quantitative analysis it follows that epistemic types of modality are used more frequently than deontic types to show speaker’s involvement. Epistemic modalities also show the modification of the illocutionary force more explicitly, and thus it is possible to interrelate these types of modality with the pragmatic functions of boosting and hedging devices.

To sum up, hypothesis I concerning detachment of the genre of political interview is not valid. Although political interview belongs to formal types of discourse, there are many features of speaker’s involvement. Using boosting and hedging devices and modal expressions, politicians modify the illocutionary force of their utterances. They attempt to strengthen their position in front of their audience, they also attenuate the force of their utterances when they are not certain about their propositions. A formal type of discourse is characterised by more complex syntax of sentences. This issue was not the subject of the present analysis, but a tendency to use shorter and loosely structured sentences may be clearly observed. The language of politicians contains informal vocabulary. The reason is that they aim at being closer to their audience. If they were detached, the voters would not trust them too much. This is also connected with a certain amount of affectiveness, which is apparent in political interviews. Even if matter-of-factness prevails since this genre focuses mainly on conveying information, showing attitude and relationship towards the audience is also observable. This feature contributes to expressing speaker’s involvement as well.

Thus, it may be concluded: The material under investigation reveals that the genre of political interview shows a high degree of speaker’s involvement, which manifests itself in the frequent use of boosting and hedging devices and their pragmatic functions and in the use of modal expressions.

The second hypothesis was formulated as follows:

Female politicians are more indeterminate and vague in their expression and they do not speak to the point. The expression of male politicians is matter-of-fact and they express themselves more precisely than female politicians.

Pragmatic functions of boosters reveal that female speakers focus on showing subjectivity and confidence and on emphasising the content of their propositions, rather than on expressing detachment or uncertainty. Indeterminacy and vagueness do not manifest themselves to a great extent in the corpus. This is connected with the position of women in politics. The sphere of politics is usually regarded as an area typical of men, in which women have a difficult position for asserting themselves. The voters may consider them “weak”, they do not believe in their abilities, therefore, female politicians must persuade them about the opposite and show them that they are the right persons for the political functions they hold or want to hold. By contrast, male politicians have an easier position in this sense. They do not have to defend their position in front of their viewers to such a great extent, which is why they can afford to be more indeterminate, hesitant or vague. Males are also more evasive, which means that they try to avoid responsibility for their statements. Responsibility for their claims is more apparent in female speakers in the corpus since they produced only two instances of evasiveness and only 14 instances of content-oriented uncertainty. Epistemic attitudinal modality is also very common in female politicians, especially the use of frankly is very frequent. My explanation is that they attempt to show a positive and responsible attitude to their viewers. They want to be frank, trustworthy and more cooperative, and in this way to gain the confidence of the voters.

In sum, although a certain degree of indeterminacy may be found in female politicians, it does not necessarily have to be a manifestation of uncertainty of the speaker. It may be a sign of female sensitivity towards the propositional content. Women are well aware of the fact that the explanation of some issues is not so easy and that things are not always unambiguous. This may be related to the fact that women do not tend to make “clear” or “non-hypothetical” claims. For these reasons, when evaluating the quantitative results of modality and making gender distinctions, it is necessary to focus not only on the types of modality but also on the linguistic means used by males and females respectively to express these types of modality.

Since male politicians produced more hedging devices showing uncertainty, assumption, and evasiveness, it may be stated that they do not express themselves more precisely than females. As already pointed out, they do not have to vindicate their position in front of their audience to such a great extent as female politicians do. They therefore do not pay attention to the linguistic means they use and express themselves less precisely. To conclude: the present analysis of political interviews shows that hypothesis II is not valid in this genre. A certain amount of indeterminacy may be indicated by female politicians but it relates to their sensitivity towards the propositional content of their utterances. Male politicians used more hedging functions expressing uncertainty, assumption and evasiveness, which shows imprecision and a lower degree of matter-of-factness.

This thesis has shown that even if political interview belongs to formal types of discourse, it nevertheless contains many features of speaker’s involvement. This analysis focused on the modification of the illocutionary force by the use of boosting and hedging devices and modal expressions. The research into syntactic and prosodic devices which also contribute to expressing speaker’s involvement was beyond the scope of the present study. It would also be interesting to analyse whether and to what extent the political position exercised influences the degree of speaker’s involvement. This was not the subject of this study, but certain tendencies supporting this claim may be observed in the corpus. Another topic for further investigation might be the difference between TV and radio interviews. Speakers on the radio are aware of the fact that their gestures and facial expressions are ineffective, whereas these means are significant when the interviews are broadcast on TV. It would be interesting to analyse whether the speakers on the radio somehow compensate the absence of visual means linguistically.

Download 1.35 Mb.

Share with your friends: