Paper for Jornadas del Discurso Político, UPF, Barcelona
Second draft, April 29, 2001.

POLITICAL DISCOURSE AND IDEOLOGY

Teun A. van Dijk

University of Amsterdam

Universitat Pompeu Fabra, Barcelona

Introduction

Probably more than any other kind of discourse, political discourse is eminently ideological. In this paper we examine some general properties of ideologies as forms of social cognition and their relations to political text and talk.

The very formulation of this goal implies that my approach needs to be multidisciplinary: We are dealing (a) with political cognition, (b) with discourse structures and of course (c) with the socio-political context in which such cognitions and discourses have their meanings and functions. This framework is very complex, and this paper can only address some of its issues.

Thus, in earlier work we have examined the general nature of ideologies, the ways these may be expressed in discourse, generally, on the one hand (Van Dijk, 1998a), and with political discourse, on the other hand (Van Dijk, 1998b, 2000; Wodak & Van Dijk, 2000).

This paper tries to combine these different strands of research, by asking ourselves whether specific discourse genres or domains, such as those of politics, have specific ways of 'formulating ideology'. Indeed, is the expression of ideology in various forms of political discourse very different from the expression of ideology in educational, academic, corporate or legal discourse, for instance?

We define ideologies as the basis of the social representations of groups (Van Dijk, 1998), where the notion of social representation is broader than that used by Moscovici and his followers (see e.g., Farr & Moscovici, 1984), namely any kind of socially shared mental representation.

This means that if we focus on politicians, we shall usually have at least two ideologies as expressed in their text and talk: First professional ideologies that underly their functioning as politicians. And second, the socio-political ideologies they adhere to, e.g., as members of political parties or social groups. We thus have conservative and progressive politicians, socialists and neoliberals, Christian-democrats, greens, nationalists and racists, and so on. And it is likely that these ideologies will show up and combine in the discourse of politicians.

This interaction between several ideologies in the discourse of politicians is not simply accumulative, however. It may be the case, for instance, that the basic principles of professional ideologies sometimes clash with those of the social ideologies politicians represent. Thus, one may profess to democratic principles because that is what the dominant consensus requires, but socially one may represent ideologies based on principles of inequality, as is the case for racist ideologies. Or, if politicians represent religious parties, their allegiance to God may often conflict with their allegiance to democratic ideologies. In other words, ideology, discourse and politics form a triangle that poses interesting theoretical and analytical questions.

Ideology

I shall be brief about ideology, and refer to my other work for more extensive treatments (Van Dijk, 1998a). I just list the major tenets of my approach:

a. Among many other things, ideologies are systems of beliefs.

b. These systems of beliefs are shared by members of a social group.

c. Groups also share other beliefs, such as knowledge and attitudes.

d. The beliefs shared by a group will be called 'social representations' (SRs).

e. Ideologies are the organizing, 'basic' beliefs of these SRs.

f. Groups not only have their 'own', ideologically based, "knowledge" (often called "beliefs" by other groups), but also share in more general, consensual, culturally shared knowledge, which may be called (cultural) 'common ground'.

g. This cultural common ground may be seen as the foundation of all cognition, across and between different groups, and thus is also presupposed by different ideologies.

h. Common ground may be empirically assessed as all beliefs that are presupposed in public discourse. This means that, for a given culture, such common ground is non-controversial, commonsensical, and hence non-ideological.

i. Part of the common ground are also the general norms and values shared by the members of a culture.

j. Groups select some of these cultural values and organize them in their own ideologies, e.g., freedom, equality, justice or objectivity.

k. Ideologies probably have a canonical structure that facilities their acquisition, use and change.

m. Although we don't know yet what this structure might be, it is probably related to the basic social properties of a group, such as the criteria of group membership, activities, aims, norms and values, relations to other groups, and specific group resources (or lack thereof) -- or 'capital'.

n. Ideologies and their structures may also be seen as the cognitive core of the identity of a group and its members, that is, as a social self-schema of a group.

o. Ideologies and the social representations organized by them control the social practices of actors as group members.
 

This summary of my concept of ideology stresses its relations to other forms of socially shared mental representations, such as specific group knowledge and cultural common ground. My formulation implies a specific architecture of social cognition, in such a way that differential group beliefs presuppose more general, cultural beliefs. Thus feminist ideologies as well as male chauvinist ones both presuppose that group members know what men and women (and much more) are. The same is true for discourse, which also presupposes this common ground, so that people, even between groups, can understand each other, and cooperate. In the same way that such common ground serves as the basis for group cognitions, the latter are in turn based on the fundamental beliefs, the ideology, of a group. For instance, socially shared opinions about immigration may be based on racist or antiracist ideologies. Note incidentally that the notion of 'common ground' used here is broad and cultural, and not interpersonal, as defined for successful conversation by Herbert Clark in his work on common ground (Clark, 1996).

Social representations such as group beliefs, attitudes and ideologies are by definition general and fairly abstract: They need to be used by group members in different situations. They are, for instance, about what WE think about THEM, in general, and not what you and I now think about some specific person. Sometimes, such general beliefs are directly expressed in discourse, for instance as generic expressions in didactic discourse and propaganda -- which typicall have the function to explain to young people or potential new members, what OUR general beliefs are about an issue. Sometimes such general beliefs also are expressed as legitimating premises in arguments, for instance to argue a more specific point: In order to argue that Mary should have the same salary as John, we may express the general, social opinion that men and women should be paid the same salary for the same work.

Between ideology and discourse: mental models

Ideologies and other social representations control discourse and other social practices of group members. They are able to do that only when their natural and abstract properties also apply in specific situations, to specific actors, actions and events. That is, besides the social psychology of group beliefs, we also need an individual psychology of personal beliefs and experiences, namely those of individual social actors and speakers. Political discourse may express group ideologies and other beliefs, especially in collective forms of text and talk such as party programs. But many forms of political discourse are produced by individual speakers, and the ways they 'personalize' the group beliefs underly the more particular properties of political discourse. That is, between social beliefs and discourse we need a cognitive interface that represents personal beliefs, opinions or experiences.

Fortunately, current cognitive psychology has a powerful theoretical notion to account precisely for this interface: mental models (Johnson-Laird, 1983; Van Dijk & Kintsch, 1983; Van Oostendorp & Goldman, 1999). These models are personal representations (in what is called 'episodic memory') of specific events people witness, participate in, or hear/read about. Such representations not only feature personal knowledge about such an event, but also opinions. That is, they are at the same time personal interpretations and evaluations of an event, and thus represent what we usually call 'experiences'. Much of the content of these models derives from 'applied' general, social beliefs, including ideologies. That is, our personal models may be (socially) biased, as when a politician has a conservative interpretation of a political event.

When we speak, write or read about such an event, we use our mental models as the cognitive basis of the discourse production or comprehension process. This is why mental models were called an 'interface' between social beliefs and discourse. This means, for instance, that in the production of the semantic representation of a discourse, a speaker includes (contextually relevant -- see below) propositions from her or his mental model. And the structure of this model may in turn also influence the structure of the discourse: High level propositions may be selected as macropropositions (topics) of the text, and the same is true for the Setting (Time, Location), Participants and their attributes, Actions, etc. that define the model.

Context models

There is still one piece of theory missing before we can apply this framework to political discourse. Speech participants not only have mental models of events they talk about (and that are the basis of the meaning of their discourse), but also mental models of the event in which they participate as speakers and recipients. That is, they also have a personal (subjective, possibly ideologically biased) representation of the communicative situation. This representation will be called a context model, or simply a context. That is, contexts are not 'out there', in 'reality', but personal mental constructs and interpretations of a communicative event. They may hence be partly different for different participants, which often gives rise to communicative conflicts (Van Dijk, 1999).

Context models exercise the ongoing control of discourse production and understanding. They define what speech acts are being performed, what aims, goals and functions a discourse has, what knowledge and other beliefs the participants have (also about the others' beliefs), and in general how the structures of discourse are adapted to the social situation. Context models define what (for a language user) is the whole of the relevant information of a communicative event. That is, The theory of context models provides a (cognitive) theory of relevance (for a related but different approach, see Sperber & Wilson, 1986).

Context models are like models of other events, but with specific categories tuned to communicative events, such as the overall social domain (e.g., politics, education), the overall social action being accomplished (e.g., legislation, teaching), setting (time, location), participant roles (communicative, relational or social roles), and cognitions (beliefs, goals, aims, opinions). Their control of discourse operates at all levels: They tell the language users which relevant (interesting, new, etc.) propositions to select in event models, what speech act to realize and what the conditions of the speech act are, what politeness forms to choose, what style characteristics to select, and so on until the very intonation of spoken discourse.

Like other mental models, also context models may be ideologically biased. That is, when speakers of different political parties speak with each other, then they not only have ideologically based mental models of the events (e.g., some recent immigration event) talked about, but also about each other, or even about the current setting, aims, or ongoing interaction. Thus, one speaker may find the other a racist, an opinion that in turn may be based on an anti-racist ideology. It goes without saying that such ideologically biased context models influence discourse in many ways. For instance, they may cause speakers to select a less polite style than would be appropriate in this situation. Indeed, sexism or racism may also be implemented in the sexist or racist ways people speak to members of other groups.

It needs little argument that the same may be true in political argument, and that thus political ideologies may influence political discourse not only at the level of 'content' but also at the various levels of 'form' and 'interaction', as we shall see below.

Political discourse

Defining 'political discourse'

The first observation that needs to be made about political discourse is that it is not a genre, but a class of genres defined by a social domain, namely that of politics (Van Dijk, 1998b). In the same way, scientific discourse, educational discourse and legal discourse represent the classes of discourse genres of the domains of science, education and law, respectively. Thus, government deliberations, parliamentary debates, party programs, and speeches by politicians, are among the many genres that belong to the domain of politics. It can't be the task of this paper to precisely define that domain, which --like other social domains-- has of course fuzzy boundaries. Political science offers as many definitions of 'politics' as there are political scientists, ranging from very general characterizations of politics in terms of power or collective decision making, to the much more specific definition of politics as the set of activities politicians engage in.

In this paper, I prefer to start from this latter, more narrowly circumscribed characterization of politics, and shall for simplicity's sake assume that political discourse is the discourse of politicians. This rules out, for the moment, even those discourse genres at the boundaries of the domain of politics with other domains, such as the discourse of a student demonstration, the messages of an anti-abortion campaign, corporate talk intended to influence tax or investment legislation, or an everyday conversation about politics. That is, these discourses belong to other social domains, even if their intention may be to influence political decision making. Conversely, a bill about education policies is a genre of political discourse, even if it has or intends to have influence in the domain of education.

Having thus limited the range of political discourse to the 'professional' realm of the activities of politicians, the next observation is that such discourse is by the same token a form of institutional discourse. That is, only those discourses of politicians are considered that are produced in institutional settings, such as governments, parliaments or political parties. This means that an informal conversation of a politician with her friends does not count as a political discourse: the discourse must be produced by the speaker in her professional role of a politician and in an institutional setting. In a more action-oriented way, we may also say that discourse is political when it accomplishes a political act in a political institution, such as governing, legislation, electoral campaigning, and so on (for institutional discourse, see e.g., Drew & Wootton, 1989; Drew & Heritage, 1992; Sarangi & Roberts, 1999).

Describing political discourse

Once confined political discourse to the institutionally bound text and talk of politicians, our next task is to systematically describe the genres that belong to that domain. This is not a straightforward enterprise, comparable to the construction of any other genre. Take debates in parliaments, for instance. What specific and unique discursive properties define such debates (Van Dijk, 2000; for introduction to the properties of discourse in general, see van Dijk, 1997 )?

Their topic? Unlikely, because parliamentary debates may be about virtually any topic in society that politicians deem interesting and relevant to talk about. Perhaps the only (fuzzy) restriction one may come up with is that the topics are usually about events in the public sphere, in particular those events that require collective decision making, policies, regulation or legislation.

Their style, perhaps, such as their lexical choice? Hardly, although there is often a general formality constraint, as for any other kind of institutional discourse. There are very few words that are only used by politicians, although there may be some jargon politicians share with the bureaucracies (ministries, government agencies, etc.). Also, there are a few ritual formulas for addressing or speaking about other MPs, or addressing the chair, but of course these cannot alone define the genre.

The overall form or format of the debate? It would be surprising if politicians would follow their own interactional and argumentational categories and rules. Thus, a parliamentary debate is not very different from any other debate, for instance in the annual stockholders meetings of big corporations. As in some other institutional or organizational meetings, there is a strict time schedule and turn allocation by the Chair.

A contextual description

We must conclude that although there are perhaps a few topical, stylistic and interactional constraints that are fairly typical of parliamentary debates, we do not find obvious evidence that the typology of political discourse is based on ('verbal') properties of text or talk. The same is probably true for government meetings, party programs or election campaigns, which are most likely very similar to any other executive meeting, program or campaign.

If we then recall that the definition of the (class of) political discourse genres was defined in terms of professional politicians and political institutions, we may conclude that the genre description of political discourse should not so much take place at the levels of text, but rather at the level of context. That is, to begin with, a parliamentary debate is primarily (and nearly trivially) defined by the fact that the participants are MPs, and that it takes place in the institution of (a) parliament. Secondly, it is political act of legislation or policy-making, among others, that is accomplished by such a debate. Thirdly, the consequences of the debate are defined in the specific institutional terms of political decision making: laws are enacted, policies decided, elections held, and so on. These and many other characteristics of political discourse obviously are defined in terms of contextual --and not 'textual' --categories, such as:

(a) The global domain: politics

(b) The global act(s) being implemented: legislation, policy making, etc.

(c) The global setting (House of parliament, session of parliament, etc.)

(d) The local political acts being accomplished: Tabling a motion, 'doing' opposition, etc.

(e) The political roles of the participants: MP, representative, party member, member of the opposition, etc.

(f) The political cognitions of the participants: Political beliefs and ideologies; aims and objectives, etc.

These (and some other) categories form a schema that defines the structure of the communicative events represented by participants in their context models, as defined above. In other words, the 'political' nature of debates, speeches, meetings, campaigns, advertising, etc. is defined in terms of their context characteristics, and not primarily by structural properties of the text itself.

Of course, although the distinguishing properties of political discourse may be largely contextual, this does not mean that we should no longer study the structures of political discourse: A study of the topics, topoi, coherence, arguments, lexical style, disclaimers and many rhetorical features (metaphors, euphemisms, hyperbolas, etc) of a political discourse may of course reveal much about the unique character of such a discourse, and also allows inferences about the cognitive, social and especially political functions of such discourse (among the many studies of political discourse, see, e.g., Atkinson, 1984; Blommaert & Bulcaen, 1998; Chilton, 1985, 1988, 1995; Geis, 1987; Wodak & Menz, 1990; Wodak & Van Dijk, 2000).

Politics and ideology

We already suggested that if there is one domain of society where ideologies are rife it is of course politics. Indeed, the commonsense meaning of ideology often simply identifies ideologies as political ideologies: socialism, communism, (neo-)liberalism, and more recently green politics. Our definition of ideology is more general, so it may be expected that apart from such political ideologies also other ideologies may be expressed in political discourse, such as ecological, feminist or racist ones. Since ideologies are defined in terms of basic beliefs shared by the members of groups, this also means that political discourse is the site where politicians' multiple ideological identities are enacted: By definition they speak as politicians, but also as conservatives or liberals, men or women, feminists or anti-feminists, racists or anti-racists, and so on.

Indeed, one of the reasons why contemporary political science has often shown reluctance to accept the very notion of ideology is the fact that in actual discourse or social practices, social actors may have unique combinations of ideologies. That is, an MP may be at the same time economically a neo-liberal, radically progressive in social issues such as abortion or minority rights, but at the same time a staunch nationalist. Or more specifically, she may be a feminist but oppose liberal abortion policies, and so on.

Such personally unique combinations of ideologies, also in politics, are not at all inconsistent with a general theory of ideology, in the same way as people's idiosyncratic talk is not a counter-example against the existence of languages or grammars. Ideologies are defined for groups, and not for individuals. And of course, based on their personal experiences and life situations, and diverse group memberships, individuals may share in (sometimes 'incompatible') ideologies. Political and ideological discourse analysis is usually based on individual discourses, so it will not be strange at all to find influences of various ideologies. In the same way as discourses may exhibit intertextuality, they also may show 'interideologicality'. Indeed, one of the interesting challenges of ideological analyses of political discourse is not only to identify these various ideological sources, but also the unique way they interact in the production of specific discourses.

Before we move to the more specific question of ideological analysis of political discourse, we briefly need to attend to two further theoretical issues. Among the various types of ideologies that may be distinguished, professional group ideologies play an important role in everyday life, because they control much of our daily institutional or organizational activities, e.g., as professors, journalists, dentists or judges (Geison, 1983). The same may thus be true of politicians, so that the social identity of politicians will also be defined by such categories as membership devices, activities, aims and goals, norms and values, relations to other groups and resources or 'capital'. We may expect that politicians exhibiting their identity will tacitly or explicitly attend to such categories, also in discourse. Thus, the (ideological!) self-identity of politicians as professionals may be (roughly) defined as follows:

a. Identity criterion: Election to political office

b. Activities: 'Doing' politics (represent citizens, legislate, etc)

c. Aim: Govern country, state or city, etc.

d. Norms, values: Democratic values, honesty, etc.

e. Position, relation to other groups: Relation with constituents, etc.

f. Resource: Political power.

Obviously, we need empirical (discourse) data to fill in the detailed ideological contents of these categories, that is, analyses of political discourses in which politicians make reference to themselves and other politicians. Thus, if politicians regularly criticize other politicians for 'not listening to the voice of the people', as is often the case in populist political discourse, then we may assume that the basic activities and norms defining the ideology of politicians involves 'listening to the voice of the people'.

The second question that needs brief comment, and that has been referred to above, is whether it makes sense to speak of 'political ideologies', as a specific sub-type -- and different from the professional ideologies of politicians, which in principle apply to all of them, whether they are on the left or on the right. If there are political ideologies, then they must specifically apply in the domain of politics, and organize political attitudes and political practices.

The best-known candidates for this kind of ideological role are the broad system ideologies, such as those of socialism and liberalism. However, these do not apply specifically to politics, but organize attitudes about a host of social issues, also outside of politics, such as the labor market, corporate activities, and so on. The same is a fortiori true for typical social ideologies such as feminism, and (anti)racism, as well as for religious ideologies such as Christendom and Islam.

Perhaps closest to being specifically political, are those ideologies that specifically pertain to the organization of the state, such as democracy, oligarchy, and so on, of which democracy is consensually the currently dominant ideology. So much so, that many of its tenets probably coincide with politicians' ideological self-definition as being 'democratic'. The dominant value characterizing this ideology would be 'equality', and the main resource the power of the people. Indeed, democracy is (ideologically speaking -- which does not mean that it is so in real life!) so much dominant and consensual in politics, that it is hardly experienced as an ideology anymore, but as a general, undisputed value that is part of what we have called the cultural common ground. Indeed, lacking democracy is typically associated with other times, and "other" cultures. Relevant for us is that in the ideological analysis of political discourse, we are bound to find, perhaps largely implicitly, the contents of the ideology of democracy -- and more specifically, manifestations of the opinions that are part of the attitudes organized by this ideology, such as those about 'free elections', accountability to the voters, and so on.

Ideological analysis of political discourse

Against the background of the complex theoretical framework briefly summarized above, we are now able to focus on the ideological analysis of political discourse. The assumption has been that political discourse and its properties are controlled by one or more underlying ideologies, possibly through more specific (but still general) social attitudes, on the one hand, and more personal mental models of concrete events, and finally by context models of the communicative situation, on the other hand. So, we'll be interested in examining the following aspects of political discourse:

(a) The dimensions, levels, structures, strategies or moves in which ideologically based beliefs exhibit themselves in discourse.

(b) Discursive evidence of the interplay of several ideologies.

We shall conduct this investigation by analyzing a concrete text, namely a debate on asylum seekers in the British House of Commons (March, 5 1997, as published in Hansard).

Ideological structures in a parliamentary debate

Parliamentary debates constitute one of the most prominent genres of the domain of politics (for details, see Wodak & Van Dijk, 2000). Their prominence derives not so much from the fact that in democracies the 'power of the people' is being exercised by parliament, and that parliamentary debates thus --ideally-- constitute the most powerful type of discourse of the state. Rather, parliamentary debates are the site where the various ideological forces in society, in the form of the political parties that represent them, are confronting each other in the public sphere. Parliamentary debates are not only public, but also for the record, so that anything participants say is open for public inspection (usually through the mass media): Everything MPs say may indeed be used against them.

For our purposes, as suggested, parliamentary debates are especially interesting because they exhibit, by definition, the social cognitions of political parties and their members. We already argued that it is not so much the text of these debates as rather various aspects of their context that define such debates as a genre: Who is speaking and what political acts being accomplished by such speech inherently defines what legislation is. Part of that context are the relevant social cognitions of the participants. Indeed, party membership, and whether or not one supports or opposes the current government, is first of all a question of socially shared opinions. And debate in parliament is thus the embodiment of a clash of opposed opinions.

Ideological analysis in its most straightforward guise involves detecting in text and talk the expression of such ideologically based opinions. We already have shown that, somewhat less trivially, the relation between ideology and discourse may be quite indirect. Specific attitudes (say about immigration), and personal models (personal beliefs about a recent event), may be the kind of representations that are the interface between ideologies and discourse structures. This means that ideologies are not always very explicit. They may be disguised, hidden or implicit. They may come in the form of opinions about specific events, or in the way such an event is described, more or less prominently.

Let us now examine some of the ideological properties of the debate we have selected as our example. Note though that many of the properties of the debate are ignored here, because they are not an expression of underlying ideologies of MPs.

Note that Mrs. Gorman is a supporter of the Conservative government in power, and that her attack is directed against a Labour opposition aiming at making current legislation about the reception of reception less harsh.

The debate is opened by --and the initiative of-- Mrs. Teresa Gorman, Conservative MP for Billrecay. Mrs. Gorman is a supporter of the Conservative government in power, and that her attack is directed against a Labour opposition aiming at making current legislation about the reception of reception less harsh.

In her speech, which is briefly interrupted by two other conservatives, and then replied to by Labour immigration specialist Corbyn, Mrs. Gorman specifically focuses on the financial consequences of what she defines as "bogus" asylum seekers. Little analysis is necessary to conclude that Mrs. Gorman is indeed conservative, tough on immigration, and racist. That is, her speech is imbued with many expressions of these ideologies. It is our special task here not only to identify these traces of the ideologies she identifies with, but especially also to examine how such ideologies are actually formulated (or left implicit) in parliamentary debates.

Mrs. Gorman begins her speech as follows,
 

(1) Mrs. Teresa Gorman (Billericay): I want to bring to the attention of the House the particular difficulties faced by the London boroughs because of the problems of asylum seekers.

Before we say something about underlying ideologies, note that part of this first fragment is not so much directly about asylum seekers, as rather about Mrs. Gorman and the House, that is, on properties of the context. In other words, the speech begins with some deictic expressions that need to be interpreted in terms of the contexts (or rather: context models) as ongoingly constructed by herself and other MP. And when she speaks about the difficulties of the London boroughs, this is not merely a summarizing reference to the topic of her speech, but also implicitly an expression of her role as MP: to represent the people (in this case London boroughs) and their problems -- even when these are not her constituents, as is the case here. She thereby conveys the impression that she is doing her job, and cares for the people -- an important part of a democratic conception of the task of representatives. Note that, at the deepest level of these context models (Mrs Gorman being a democratic MP defending the people) we do have the ideology of democracy. This and other ideologies and attitudes will coordinate the social representations of all participants.

As to the "content" of this fragment, Mrs. Gorman not only speaks about the difficulties of the London Boroughs, but also of the "problems of asylum seekers". This ambiguous expression may refer to the problems of asylum seekers, or to the problems caused by asylum seekers. The ambiguity and hence vagueness may be a device of positive self-presentation, which requires that anything negative said about Others be formulated in an indirect, vague or mitigated way. The rest of her speech, however, does not leave any doubt about the interpretation that attributes the problems to (be caused by) the refugees.

Defining refugees as a problem is classic topos of anti-immigrant discourse -- and probably a dominant category in a socially shared anti-foreigner attitude, which in turn is based on a racist ideology. Most of the cognitive representations derived from such an ideology describe Us in positive terms, and Them in negative terms. One way of doing this is to represent the Others in terms of a Problem-for-Us at all social levels: jobs, housing, welfare, crime, attitudes, and so on.

The same polarization may also be implemented and expressed in a stronger form: They are a Threat-for-Us. Thus, we see how underlying belief structures are implemented at the semantic level of the text by the choice of a specific lexical item. The ideological significance of such lexicalization comes to the fore when we substitute it by possible other items expressing a similar situation: The situation of the reception refugees may also be defined as a challenge, and as a cultural, financial and laboral bonus for the country: After all, many refugees are highly educated and very motivated, and will probably contribute a lot to their new homeland.

Note finally a question of form: Defining refugees as a Problem -- as people who cause (our)difficulties -- is being done right from the beginning. In other words, there is an order effect here -- in which initial definitions have an important impact on the understanding of the rest of the speech. Indeed, since the notion is part of the thematic first sentence, it would be part of the macroproposition controlling this speech -- as it does in the mental model Mrs. Gorman has about the immigration of refugees and the consequences for some London boroughs.

In sum, in this example we witness the expression of an underlying ideology in terms of a negative definition of a situation as a 'problem', and an (still implicit) attribution of such problems to refugees.

Mrs. Gorman is not very vague for a long time. Already the next paragraph of her speech opens the usual registers of xenophobic talk, though still in a subdued tone:
 

(2) There are, of course, asylum seekers and asylum seekers. I entirely support the policy of the Government to help genuine asylum seekers, but to discourage the growing number of people from abroad who come to Britain on holiday, as students or in some other capacity and, when the time comes for them to leave, declare themselves to be in need of asylum.

The second move in Mrs. Gorman's speech is one of categorization, namely a distinction between genuine (good) and non-genuine (bad) asylum seekers. Although later in her speech she sometimes pays lip service to this distinction, very often she generalizes about all refugees. Categorization is one of the elementary mental aspects of actor and group description, as we know from many studies in social psychology (see, e.g., Abrams & Hogg, 1999; Fiske & Taylor, 1991).

In order for positive self-presentation and negative other-presentation to work, however, Mrs. Gorman cannot possibly ignore, in the current parliamentary context, that not all refugees are bad. This first categorization is thus an important move of positive self-presentation, namely a disclaimer of Apparent Concession: There are also "good" ones among them (for these disclaimers, see e.g., Van Dijk, 1984, 1987, 1993). The implication of this disclaimer is to emphasize that she cannot possible a racist, because she supports help for "real" refugees. Of course, after this quick disclaimer, she can then focus on the Others -- namely the "bad" immigrants.

Note that this whole passage is articulated by underlying norms, values and ideologies that sustain the polarization between genuine and non-genuine asylum seekers. There is not merely a rhetorically expressed categorization (using a repetitive standard phrase --usually pronounced with mounting intonation: there are X and X!), but also hyperbolic ("entirely") reference to support for the government's help. Similarly, instead of bluntly describing the non-genuine ones as "bogus" as she does later in her speech, she here refers to the others as people from abroad who overstay, ironically adding "declare themselves to be in need of asylum" -- a reference to a declarative verb that implies that the others only say they need asylum, but do not really need it. Note that even the lexical item "discourage" may be interpreted as a euphemism for the harsh immigration policies of the Conservative government -- and hence as form of avoiding a negative self-description.

In terms of the point of our analysis, we see that this passage indirectly expresses a xenophobic ideology by sub-categorizing refugees into good and bad ones, and at the same time "softens" the possibly negative impact of such a categorization by emphasizing the expression of a humanistic ideology of help to people who need it (limited of course to the few who are declared "genuine"). The actual description of what happens is like a scriptal model of what happens when non-genuine asylum seekers want to enter the country. The last sentence of this fragment thus becomes a mini-story expressing this model, following the well-known categories of everyday storytelling: Orientation ('they were just innocent holiday makers...') followed by the unexpected Complication ('...but suddenly they declare themselves..."). Although framed in more general terms (no specific persons are mentioned), such a model-based mini-story makes her general point much more concrete, and therefore persuasive.

Finally, this small passage can (trivially) be interpreted in light of a professional ideology -- namely that of the MP correctly doing her job. In the previous passage, she declares herself to worry about London boroughs, and in this passage she shows that she is a staunch defender of her own (Conservative) government. That is, she implements and sustains the ideologically based policy of her party, and thus behaves like an MP should do. In others words, we see that nearly at all levels of this small fragment, several ideologies appear in various structures.

The political upshot and function of the expression of these ideologies here is first of all clearly expressed in terms of Mrs. Gorman's explicit support for her government's policy, by associating this policy with a positive concern for genuine asylum-seekers, on the one hand, but to distance herself and her party, from non-genuine asylum-seekers. In this way, as will also be abundantly clear in the rest of her talk, she at the same time plays a strategic populist card, by preformulating the usual public rejection of "bogus" asylum-seekers.

It is important to emphasize the political specificity not only of the underlying ideologies (Conservative vs. Labour) but also that the way these are actually expressed, implemented or enacted always has political conditions, consequences of implications. Ideologies are only brought to bear in parliamentary debates when they are contextually -- and hence politically -- relevant. Mrs. Gorman may not feel or share any concern for genuine refugees at all, but it is important for political reasons that the Conservative party not be directly associated with xenophobic exclusion policies.

Similarly, political discourse and hence political action (such as supporting the government, attacking the opposition, representing voters and making populist claims) may derive part of their coherence, and hence their credibility and legitimacy when based on consistent ideological principles.

On the other hand, politics nearly always has priority over ideology. Thus, when the left in both the UK and France came to power -- after criticizing the immigration policies of the previous, conservative governments -- they hardly changed the draconian immigration laws, well aware that large numbers of their voters would not welcome a too soft policy on immigration. In other words, in political situations many politicians, especially of government parties, will rather look at their popularity with the voters than in their ideological heart of hearts.

This is also one of the reasons why the rest of Mrs. Gorman's talk is increasingly blatant about the category of "non-genuine" refugees, whom already in the next paragraph she not only describes as "economic refugees" but also labels as "benefit seekers on holidays". The style of these last words not only is a form of the familiar negative other description of xenophobic discourse, but also has a clear populist function: It is thus how her voters like to think and speak about refugees. This political orientation to the "ratepayers" becomes even clearer in the following passage:
 

(3) It is wrong that ratepayers in the London area should bear an undue proportion of the burden of expenditure that those people are causing.

The use of "wrong" presupposes a norm, and such a norm is implemented in the ideologically based political attitudes ratepayers are attributed by Mrs. Gordon in this blatant populist move. It may indeed be doubted whether a conservative and racist MP like Mrs. Gorman cares very much for the taxes her "ratepayers" have to pay. But again, her professional ideology prescribes concern for the voters, a conservative ideology professes less state intervention and less taxes, and her anti-foreigner ideology inspires the selection of precisely the refugees as a cause for financial concern.

It is not likely, for instance, that she will explain that virtually all the expenditures "caused" by the refugee are being paid to British welfare professionals and companies -- and hardly to the refugees themselves, and even then will eventually benefit British businesses, and that the taxpayer thus finances British professionals and businesses rather than asylum seekers.

This last remark is not only meant as a critique of Mrs. Gorman's speech, but also as an analytical remark about the kind of information that is not included in a speech, also for ideological reasons. In other words, also omissions are often ideologically based, and they can only analytically be recuperated if one knows about the details of immigration -- knowledge that few members of the public at large have, so that such incomplete and therefore misleading discourses meet very little critique, not even in the mass media.

Consider also the following fragments, in which Mrs. Gorman begins with a well-known disclaimer of Apparent Empathy (I understand..but) expressing the underlying humanistic ideology, but ends, in the next passage, by describing the refugees in terms of "bogus":
 

(4) I understand that many people want to come to Britain to work, but there is a procedure whereby people can legitimately become part of our community. People who come as economic migrants are sidestepping that.

(5) The Government, with cross-party backing, decided to do something about the matter. The Asylum and Immigration Act 1996 stated that people whose application to remain in Britain had been turned down could no longer receive the social security and housing benefit that they had previously enjoyed. That is estimated to have cut the number of bogus asylum seekers by about a half.
 

Politically speaking fragment (4) also expresses a legalistic ideology, sustaining attitudes about legal immigration. The use of ingroup possessive pronoun "our" in this context also signals fragments of a nationalist ideology, by which membership of "us" can only be legitimate under specific --legalistic-- conditions and procedures.

Fragment (5) shows another familiar strategy of political discourse, namely that of Apparent Consensus: The speaker claims that her or his opinion or policy is not just a partisan --and hence less credible-- position but "bipartisan". This argument is especially intended, of course, to prevent the opposition from protesting, and thus manages the interaction of this debate. At the same time, such a Consensus move is politically interesting because it implies that the proposed policy is ideologically neutral (if both Conservatives and Labour accept it). The final sentence then aims to enhance credibility with a powerful argumentative move, namely that of efficiency: Something is good because it works. A classical expression of a utilitarian ideology.

Whereas at one moment bipartisan consensus is being used as a powerful argument, the next fragment takes the usual confrontation road, and squarely attacks Labour:
 

(6) It is a great worry to me and many others that the Opposition spokesman for home affairs seems to want to scrap the legislation and return to the previous situation. I would consider that extremely irresponsible. It would open the floodgates again, and presumably the 200 million a year cost that was estimated when the legislation was introduced would again become part of the charge on the British taxpayer.

That the Opposition "worries" Mrs. Gorman is nothing special, and part of the political game. Therefore, if a politician signals a problem or complication, she'll have to emphasize that it is a problem for many, as she does here -- another well-known (Generalization) move of parliamentary semantics. The accusation to "scrap" legislation, to return to a (working) previous situation may be thought to be based on a professional ideology organizing attitudes about what can and should (not) be done in politics. That is, the Opposition is not only accused of being lax on refugees, but also to be unprofessional.

A crucial argument in all political argumentation is of course financial, and that is why --through the usual threatening-fluid metaphor used for the arrival of refugees ("opening the floodgates") -- she brings us the alleged costs, combining such reference to the standard populist phrase "charge on the British taxpayer". The accusation of irresponsibility addressed at Labour is another fragment of the underlying norms and values that are supposed to define the professionality of MPs, on the one hand. At the same time, accusing Labour of financial irresponsibility (spending) is a standard tactic of the Conservatives -- pleading for a less generous State -- in this case also for populist reasons (it is money spent on Them and not on Us).

Conclusion

These observations on a few fragments have shown how political discourse in general, and especially parliamentary debates, are replete with ideological expressions at all levels. We have found that in such debate various ideologies may be at play at the same time, and be made relevant in different contexts.

Politically relevant, we first find evidence of professional ideologies, controlling attitudes, practices and discourses of MPs and what these should and should not do. Representing the people, defending the taxpayer, being responsible with public finances, are all part of such an attitude, and as soon as the political opposition is seen to break these norms and values, they will be negatively valued and hence politically accused, as we have seen above.

Secondly, we obviously may expect the direct or indirect expression of the political ideologies that inspire the politics and policies of the parties involved, such as Conservative and Labour in our example. Thus, Conservatives will advocate a less generous state, will want to reduce benefits, and so on, whereas Labour may be less restrictive in this case (depending on whether they are in the opposition or form the government).

Closely related to these conservative and more liberal ideologies are the ideologies about who belong to Us, and who do not -- ideologies of nationalism, race and ethnicity. We have seen that much of the debate is controlled by the usual polarization, first between Us British and Them Foreigners and especially refugees, and secondly between genuine (good) and bogus (bad) refugees. Here and even more in the rest of her speech, Mrs. Gorman attributes all possible negative things (fraud, abuse of benefits, laziness, etc) to the bogus refugees, and the overall picture is clearly a racist definition of the others.

The political response of Labour in this case is to emphasize opinions inspired y a humanistic ideology, to advocate human rights and in the text to elaborate on the miserable situation and the terrible experiences of the refugees.

Again, all these ideologies do not express themselves neatly and explicitly. Sometimes they merely control an intonation, a lexical item or an argumentative fallacy. They combine with other ideologies, and sometimes this leads to apparent contradictions in the text, for instance in disclaimers that profess empathy with the plight of the others when in the rest of the speech no such empathy is visible.

Finally, we have found some evidence of the more specific political expressions and functions of ideological opinions in parliamentary debates. We have assumed that politics generally has priority over ideology. That is, even when a party's philosophy in principle would mean an open-door policy for immigrants, as is the case for Labour, the forces of politics (voters!) may make politicians much more " realistic" and often forget their principles, especially when their governing is at stake. Moreover, especially those ideologies will be preferentially selected that in the current political contexts pay most dividend. Thus, if it is relevant to be humanitarian, then we'll find that ideology, and when it is politically rather practical, then a more utilitarian ideology may become dominant.

It is this complex latter process of "ideological management" that detailed political discourse analysis is able to describe in more detail.

REFERENCES

Abrams, D., & Hogg, M. A. (1999). Social identity theory and social cognition. Oxford: Blackwell.

Atkinson, J. M. (1984). Our masters voices: the language and body language of politics. London: Methuen.

Blommaert, J., & Bulcaen, C. (Eds.). (1998). Political linguistics. Amsterdam: Benjamins.

Chilton, P. (1988). Orwellian language and the media. London: Pluto Press.

Chilton, P. (1995). Security metaphors. Cold war discourse from containment to common house. New York: Lang.

Chilton, P. (Ed.) (1985). Language and the Nuclear Arms Debate: Nukespeak Today. London and Dover, New Hampshire: Frances Printer.

Clark, H. H. (1996). Using Language. Cambridge, England: Cambridge University Press.

Drew, P., & Heritage, J. (Eds.). (1992). Talk at work. Interaction in institutional settings. Cambridge: Cambridge University Press.

Drew, P., & Wootton, A. J. (Eds.). (1988). Erving Goffman: Exploring the interaction order. Boston: Northeastern University Press.

Farr, R. M., & Moscovici, S. (Eds.). (1984). Social representations. Cambridge (Cambridgeshire New York Paris: Cambridge University Press Editions de la Maison des Sciences de l'Homme.

Fiske, S. T., & Taylor, S. E. (1991). Social cognition. New York: McGraw-Hill.

Geis, M. L. (1987). The language of politics. New York: Springer.

Geison, G. L. (Ed.). (1983). Professions and professional ideologies in America. Chapel Hill: University of North Carolina Press.

Johnson-Laird, P. N. (1983). Mental models. Cambridge: Cambridge University Press.

Van Oostendorp, H., & Goldman, S. R. (Eds.). (1999). The construction of mental representations during reading. Mahwah, NJ: Erlbaum.

Sperber, D., & Wilson, D. (1986). Relevance: Communication and cognition. Cambridge, Mass.: Harvard University Press.

Van Dijk, T. A. (1984). Prejudice in discourse: An analysis of ethnic prejudice in cognition and conversation. Amsterdam Philadelphia: J. Benjamins Co..

Van Dijk, T. A. (1987). Communicating racism: Ethnic prejudice in thought and talk. Newbury Park, CA: Sage Publications, Inc.

Van Dijk, T. A. (1993). Elite discourse and racism. Newbury Park, CA, USA: Sage Publications.

Van Dijk, T. A. (1998a). Ideology: A multidisciplinary approach. London, England UK: Sage Publications.

Van Dijk, T. A. (1998b). What is political discourse analysis? In: Jan Blommaert & Chris Bulcaen (Eds.), Political linguistics. (pp. 11-52). Amsterdam: Benjamins.

Van Dijk, T. A. (1999). Context models in discourse processing. In: van Oostendorp, Herre, & Goldman, Susan R. (Eds.), The construction of mental representations during reading. (pp. 123-148). Mahwah, NJ, USA: Lawrence Erlbaum Associates.

Van Dijk, T. A. (2000). Parliamentary Debates. In R. Wodak & T. A. van Dijk (Eds.), Racism at the Top. Parliamentary Discourses on Ethnic Issues in Six European States. (pp. 45-78). Klagenfurt, Austria: Drava Verlag.

Van Dijk, T. A. (Ed.). (1997). Discourse studies: A multidisciplinary introduction. 2 vols. London (England: Sage Publications.

Van Dijk, T. A., & Kintsch, W. (1983). Strategies of discourse comprehension. New York: Academic Press.

Wodak, R., & Menz, F. (Eds.). (1990). Sprache in der Politik - Politik in der Sprache. Analysen zum "ffentlichen Sprachgebrauch. (Language in politics - politics in language. Analyses of public language use). Klagenfurt: Drava.

Wodak, R., & Van Dijk, T. A. (Eds.). Racism at the Top. Parliamentary Discourses on Ethnic Issues in Six European States. Klagenfurt, Austria: Drava Verlag.

(Please refer to the orginal website for more . 请查阅原网站获得更多信息。此页面仅作推荐。页内连接为原网站连接)

来源：http://www.discourse-in-society.org 2003年7月4日17:46