You are on page 1of 16

The Debate

Ideologies and Other Disciplines


Angel Gordo Lopez (University of Bradford): Not coming from a linguistic background, I would like to talk about the politics of the scheme you used. I find your use of some concepts from cognitive psychology to try to explain the structuring of attitudes, opinions etc. interesting since cognitive theory does not yet have a model to represent the internal ordering of the mind. Teun van Dijk: When I think about structures of cognition, my interpretation is very far away from current debates in cognitive psychology. I am not concerned with whether memory is structured in the computer metaphor of interface, because I am not aware of any other alternatives that I could use in my analysis. I am aware of the fact that when you opt for certain cognitive concepts, you may, to a certain extent, be adopting socio-cultural or political notions that you do not want in your analysis. But to be able to speak about representations of reality, I need a terminology and a theory for this, and, at the moment, I have no alternative than the one offered by cognitive psychology. I am open to other things, but however this theory is formulated, I need a theory of structures of the mind which accounts for thoughts, interpretations, opinions, ideologies, beliefs etc. I will not accept a theory that explains them away by saying Its all discourse, I dont know what the mind is, the only evidence we have is discourse, so why dont we talk about opinions etc. in terms of discursive configurations? I dont want to reduce the mind to discourse, or vice versa. Alexandra Korol (University of Manchester): Could we expand a bit on the reflexivity of attitudes and their connection to ideology. Because you did show the attitude to ideology link, but not the link the other way around, and for me, the relationship is reflexive. The social environment where these intercourses are going on, and where this ideology exists must be taken into account. Teun van Dijk: Yes, of course, and one of the things I try to fight against is the discussion of ideology as here and society or social context as there. So, one of my pleas is that the social context should be an integral part of the analysis. Ideologies are acquired, used, negotiated, changed in specific social contexts and these contexts may be local in the sense of talk, interpersonal communication etc. or they may be the broader social context in the sense of social structure. Sue Wright (Aston University): Im interested in seeing how ideology affects structures of vocabulary and how ideology is related to the way we formulate ideas in speech. You seem to see this as a linear process from the ideology to the way youre using the language. I see this rather as a circular process. Since you can only express it in language, the structures you are using will affect your ideology. I suppose its very weak Sapir-Whorf really but you cant say what you dont have the words to say, can you? For example, there are certain words for foreigner which carry with them a value judgement and this will affect how you think and talk and feel. You cant frame the concept without words and you cant really stand outside the concept and find a different way of framing it,
146

Debate

147

because youre still using the same language that framed the concept! So, would you agree that it is perhaps more circular than linear? Teun van Dijk: Yes. People acquire ideologies, attitudes, opinions etc. through speaking and communicating so, of course, the process is two-way. But you can also acquire and change attitudes and ideologies on the basis of discourse. That is why discourse is so fundamental, because it is part of the whole acquisition process. But then you have the problem of the chicken and the egg! Which comes first? How can people understand language in the first place, if they dont have all the knowledge needed to understand it, and how can they acquire this knowledge without the language? This brings us to the field of psycholinguistics and language acquisition theories. The second issue is more problematic. Do the concepts people have necessarily depend on the words they use? As regards the Sapir-Whorf hypothesis, I would say, yes, in a way everything we know is determined by language, but in another way it is not. People can conceptualise outside of natural languages. Sue Wright: And you can always invent. Teun van Dijk: Yes, exactly. This is why black people in the United States are continuously inventing new words to describe themselves, because of the associations that words like negro have. So, to answer the major question which you asked: yes, the process is circular and social cognitions are based on the discourse in which we participate and they are the conditions for our knowledge in general and for attitudes and ideologies in particular, and vice versa. Once you have acquired this knowledge and these social representations, once you have specific models and attitudes, much more of your reading of texts will be determined by these things.

Ideology as Hypernym
Ulrike Meinhof (University of Bradford): Id like to address the link between discourse structures and attitudes, opinions and ideologies. My primary interest is how to deduce opinions, attitudes and ideologies through discourse analysis? If we are not convinced of this interconnection, questions such as how to define attitudes, values, and opinions, or their internal structuring and social functions do not make much sense. So, if we work from this assumption, then everything else will fall into place. I can follow you through the mental models, I can follow you as far as attitudes, and I can see that you can theorise about the mental models and attitudes and opinions coming from and being linked with discourse. What I have a problem with, is when you go one step further back and postulate that these attitudes form a particular ideology which monitors various other mental models. Im no longer sure about this. Id like to be because it would solve some of the perennial problems of our research. Teun van Dijk: Lets assume someone is talking about affirmative action or abortion. There will always be some kind of coherence, allowing for personal interpretation and individual inconsistencies. So whether the event being discussed is abortion or affirmative action or whatever, what someone says will

148

Current Issues in Language and Society

inevitably be controlled by a few basic views of gender relationships. This overall view will be the ideology. And I am claiming that you will see evidence for this in the discourse itself. Tom Bloor (Aston University): Isnt ideology in fact a collection of those things? You say that it directs attitudes, but it seems to me that attitudes are rather instantiations of ideologies. Teun van Dijk: I think that is a good alternative which Ive thought about myself. Its always wise in any kind of scholarly activity to ask whether we need a particular term and what purpose it serves. So we could say that ideology is just the collection of those particular attitudes nothing more and assume that there isnt some kind of underlying structure or force. I use some discursive metaphors to show that in the same way that you have sequences of sentences, local coherence and so on, we also have an overall organisation which controls the unity and continuity of the whole text. We assume that these attitudes are not an arbitrary collection, they follow a certain direction, they have some kind of continuity. And I want to explain that. Why is it that if someone acquires an attitude about abortion and one about class ceilings, that these dont just go in different directions? Why can we say that someone has a sexist or a conservative or a liberal view of these things? Somehow, I want to capture this kind of everyday observation and that comes very close to the traditional notion of ideology. I want something which defines this unity and that I would like to call ideology. Tom Bloor: Isnt it just a superordinate term? I wouldnt disagree with what you said. Im not suggesting that feminism is just a random collection of attitudes that might happen to fall together, I think there is some kind of overall connection. But, you seem to suggest that ideology is something independent of a set of attitudes, whereas I see it as a superordinate term for a set of attitudes. I think there is a substantive difference here. Teun van Dijk: Maybe there is, but this separation is a result of the particular analysis one carries out. As linguists differentiate, for analytical reasons, between syntax, semantics and pragmatics, even though all aspects of discourse are interrelated and happen simultaneously, I separate ideologies from attitudes for analytical reasons. When I speak about the mind, we all have these social representations in the mind, depending on how we talk about and represent these things, and it may very well be that in the structure of the mind there is some overall unity however this is defined and this particular unity can be called the ideology. The only thing I do is I take this structure out and represent it separately and say this is, provisionally, the ideology. It can be compared to what I do with macro-structures in texts. They are not there, they dont exist separately, but they are an abstract structure I need in order to explain something. And if you say that this is a superordinate structure, thats fine with me.

Ideology: Personal or Socially Shared?


Robin Warner (University of Sheffield): But when youre working with something as contradictory and self-disguising as ideology, dont you think that to use tiered structures and linear processes might actually prejudice the way we

Debate

149

think about it? If you see ideology as an organising level of attitudes, this is the same as saying that the function of an ideology for a person is to achieve coherence of attitudes. And attitudes again are something above the level of personal opinion. It did occur to me that we could start with the person. Im sure there is such a thing as an ideology of the person, of the self, which does include ones own sense of coherence of ones own attitudes and beliefs. Do you see what Im getting at? The whole thing can become completely interpenetrated. It seems perhaps a bit dangerous to start to consider a higher level of organisation and a lower level of organisation or a more social and a more personal level, where something like ideology is concerned. This is the first point I wanted to make. Secondly, Im not sure about the way youre defining and describing ideologies: you call both the organisation of attitudes effected by ultra-nationalists and the organisation of attitudes effected by socio-lingual advocates of affirmative action ideologies. I think in the latter case its very likely that you have a reflexive awareness of attitudes and the process of reaching points of view and decisions. This reflexive awareness makes the organisation of these attitudes qualitatively different to someone who simply works in terms of getting their prejudices into some kind of coherent order. Teun van Dijk: I have no fixed answers to these questions, and that is why this is one of the most complex projects I have ever worked on. So many things are interacting which I cannot bring into some kind of solution. Thats why I welcome this type of forum and the suggestions you can make. But in a tentative response to your question, I have decided, for myself, to try to explicate a notion of ideology which is useful for all kinds of purposes, for example, for critical discourse analysis. This is why I am not going to search for ideology everywhere and why, for me, discourse and ideology or discourse and action are not identical. There are so many offerings in the literature on definitions that you can take your pick from them. So I have decided to define ideology exclusively as a socially shared thing, just as language is thats my most persuasive analogy. In my definition, there is no such thing as a personal ideology. Thats a claim. What you mention about personal ideology is, however, interesting because for an individual there may very well be some kind of coherence in his/her thoughts. I have two explanations for this: first, the individual makes these thoughts coherent as a member of a group, and by definition s/he knows the language of the group and therefore s/he knows the ideology in which s/he is participating. So, where each member of the group accepts the particular ideology, s/he is doing so freely for him/herself and feels personal coherence. But, secondly, as individuals, people have their own principles and these may have their own consistency what traditional psychology calls personality. Both of these will control how individuals think, speak, read, understand, represent etc. and this accounts for the differences between individuals. Political scientists argue that this unity doesnt exist because of individual variation. This is true in a sense, but there is also an overriding coherence. For example, ultra-nationalists in various countries come up with the same arguments. How can I explain this is in a purely individual way? I cant.

150

Current Issues in Language and Society

Robin Warner: I was thinking more of the ideology of the self the perception of the relationship between (in crude terms) the mind and the body. Teun van Dijk: But dont you think that part of that is also socially shared? Robin Warner: I think its reciprocal. It is the other half of the group ideology. David Graddol (Open University): And it goes back to politics and discourse within which you constructed the problem. The kind of terms within which you have constructed the model itself the individual, the social group are very realist and humanist terms. You are suggesting this is an ideological discourse, the Western, humanist ideological discourse within which we all try to rationalise our behaviour, so that we can conceive of ourselves as being autonomous, rational selves. But that is causing me a bit of a problem, because youre starting within that place where you constructed the problem, i.e. that people are rational, autonomous selves, that things like opinions which you objectify have structures, systems that youre trying to get at this youve constructed as your research problem. But some of us in this room have been trying to work in an alternative discourse about language and social identity. For instance, you cant talk about someone having an intention, instantiating it, encoding it somehow in language, producing utterances and taking individual responsibility for the meaning of that utterance, without looking at how much of language is jointly and collectively constructed, even at the clause level. We know how problematic things like meanings of individual utterances are. Now that is obviously problematising the notion of the autonomous self in relation to things like opinions, intentions and language. I dont think its throwing that model out completely, because I wouldnt feel happy about abandoning the discourse in which youre working and its necessary for talking about these issues empirically. But I cant help feeling that some of the issues youre looking for, you cant really find entirely within the individual. Integrity and structure wont be found just within the individual mind. Some of these things do appear to be the result of collective action from which they may emerge in a very messy way. The notion of social is not something we have all internalised in an identical way, but something that arises out of interaction because a bit of it is inside you and a bit of it is inside me, and when the two of us get together to interact, something different but almost predictable comes out. Teun van Dijk: Yes, I agree totally with that. These cognitions are crucially not exclusively, because I dont know for sure constructed during interaction and discourse with other people of the group or the culture. The question now is: what is the consequence of that particular observation? Once these people have construed their meanings and so on through this on-line talking, or joint reduction, what decides the direction this meaning takes in our minds? Some people say that its all locally produced, it is contextual to the limit these people are radical contextualists. Others, like me, will say yes, thats how things are actually being done, thats how things are manifested. But at the same time there is a kind of result, consequence, emerging property which goes beyond the context. So, you learn something. You store something from this interaction in the mind a category, a concept or whatever and later you will use this again. So, it is not the case that you have to reinvent language, the world, ideologies

Debate

151

every time you interact, you use fragments which you already have and that have to be stored and represented somewhere and I have no other way than to say that this takes place in the mind. But although you have this stored knowledge, sometimes the context is so complex and inconsistent that there is no evidence of the social thing which interacting partners share. My objective is, on the one hand, to account for that kind of unique context, both for the uniqueness and for the ad hoc construction of meaning in this situation, and on the other, to investigate whether we all have a particular context that will still be there the next time, that is permanent, continuous. In fact, most people dont remember individual interactions at all, what they do remember is for example, how to speak to a professor, or how to express ones sympathy. This is evidenced by repetition, recognition etc., which show that this is permanent. There is a conflict between saying everything is locally constructed or saying everything is controlled by things that are already there. For example, some people say: Ill tell you a story, and there are schemata for stories, and thats what controls storytelling. Other people say: no, storytelling is not based on ready-made schemata, its based on interaction now. I think its both. Alexandra Korol: Id like to pursue this idea of ideology as a socially-shared thing. I always prefer to locate a particular phenomenon in its context and if we are talking about ideology as shared by a group of people, the question arises, what group of people? We shouldnt forget about the larger context call it society. The ideology is not just shared by the group, but also constructed by society and then reconstructed by the group. There is this interrelationship all the time, this negotiation about what is being accepted or rejected by this particular group, or by this particular society. If we talk about ideology, we always talk about the ideology of some particular society at some particular time. So, on the one hand, there is the continuity, but, on the other, there also is the specificity of the particular time. I would think that ideology is social, and also societal constructed by and for some people, some class or group that comes to power. To talk about the group only is not enough. Teun van Dijk: I think this is important theoretically: are ideologies for the society as a whole, or must they, necessarily, be linked to specific groups? Is there for instance an ideology for one nation; is there something like a North American ideology, for example? Alexandra Korol: Yes, in the North American context, we could say that this is the ideology of a society that tries to build capitalism and that this is an ideology for the whole of that society. We could also speak about the ideology of communist society. In both cases, to what extent does the ideology of the society correspond with the personal views and beliefs of the individual members of that society? I also have a problem with the notion of identity as part of group ideology. To what extent do personal views express group ideology? To what extent do I make the group ideology my own? Who does the group ideology belong to? Teun van Dijk: I take your point, but the link between group identity and ideology helps to explain why people will defend it in a certain situation, if they

152

Current Issues in Language and Society

feel that they are members of a group. If that membership feeling is not fulfilled, if they dont identify with feminism or Dutch nationalism or socialism, then, by definition, the basis for this ideology is not there, according to that theory. You can of course be a Dutch person, but you may not identify with the Dutch at all, which means that you are unlikely to espouse a Dutch nationalist position. So, as an individual, of course you have the possibility of identifying with several groups and this identification is just one element of accepting the basic ideas of that particular group, which I call an ideology. Christina Schffner (Aston University): Does this mean that there is no limit to the number of ideologies you as an individual person can have? You speak of racist or feminist ideologies: where do these labels come from? How do we find the names for these groups? How are they identified compared with other groups? In MarxistLeninist teaching, we get this very simplified explanation: since ideology is defined as a class-based concept, and since in each social system there are only two main, antagonistic, classes, there are also only two ideologies, e.g. in the capitalist society, the bourgeois ideology of the ruling bourgeoisie as opposed to the socialist ideology of the working class. When we link ideologies to groups, which I think makes sense, we have the problem of specifying how many groups. Teun van Dijk: This is where I have run into problems myself and it is related to how we define groups. Do groups adopt ideologies or is the definition of a group related to its having an ideology? It may in fact be both. For example, Norman Fairclough would argue that there must necessarily be some kind of struggle and power relationship or dominance. But I doubt that. Helen Kelly-Holmes (Aston University): Would world-view be a better concept in terms of group identity? This would get us away from the associations which particular ideologies such as feminism and communism have and which, perhaps, limit the definition of ideology. Teun van Dijk: This is one of the many words that has been used, but it doesnt really matter what you call it. I would say that world-view is more complex and encompasses everything all social representations, knowledge, attitudes, plus basic orientation. Within that social cognition, there is one structuring, organising, monitoring element and thats what I call ideology. So its not the same thing as world-view. Other people suggest its the same as belief, but belief is much too general, too vague. We have all kinds of beliefs.

Inconsistencies in Ideologies
Kay Richardson (University of Liverpool): Id like to come back to what you said about a unique context of interaction that can be complex and inconsistent, on the one hand, and, on the other hand, can have permanent and coherent aspects. My question concerns the methodological interpretation of whats going on. Is there not a problem of projecting too much coherence, and then perhaps too much rationality, onto the underlying thing that is carried from one context to another? I think this connects with what David was saying in terms of the extent to which we expect the components of an ideology (within the individual or even in the social) to be tied together without contradiction, inconsistency and

Debate

153

irrationality. And if you are looking for coherence, arent you at least going to run the risk of seeing it where it isnt, or probably isnt, there? I want to connect this with another point: the variability of levels of coherence. If you want to call Keynesian economics or monetarist economics an ideology but also racism an ideology, what kind of fancy footwork are you going to have to do to account for the fact that people have sat down and thought about monetarist economics? What kind of careful unpicking of carefully thought-out constructs will you have to be involved in to prove that this is a self-serving ideology although I understand that this may not be your definition of what an ideology is? Whereas with racism, to come back to what you were saying, there is a level of thought production which isnt nearly as rational and thought through with as much intellectual energy as is the case for Keynesian economic theory. Teun van Dijk: There is a lot of debate going on about how complex, how consistent ideologies are. We can collect evidence from polls which shows that opinions are not consistent. Personal deviations seem to be one element of the evidence we do have. But my way of looking at things is not to say that ideologies are rational, because they may be very irrational. Secondly, I do not say that they are necessarily consistent; they may be construed from inconsistent parts and this is true for a lot of racist thinking. Therefore, I use the term coherence to say that there are relationships that ensure that there is some kind of orientation in the attitudes people choose once they have a racist outlook on the world. Another thing I do not want to say is that they are more or less complex, as it is discussed in political science. Philip Tetlock makes a distinction between ideologues and non-ideologues. The ideologues are the people who consciously think about all these things, in rational terms and concepts, the people who write the books and propaganda for a party. But one of the things that has been neglected is the notion of common sense. So, for example, everyday theorising about economics etc. would be seen by some people simply as common sense and would not be a consciously thought-out ideology for them. One way of explaining irrationalities, inconsistencies and incoherences in the way people speak about certain issues is to say that the ideologies or attitudes dont exist because it is all local, related only to particular issues and individuals. My suggestion is that you dont have to throw out the theory because of these inconsistencies. The relationship between ideologies and discourse is so complex. People as individuals are members of different groups and this will affect attitudes. Similarly, the individual has his/her own priorities which will also impinge on that. Sometimes this will take precedence and this causes variation, inconsistencies etc. I want to account for that variation and that continuity. Kay Richardson: Well, this is why I termed it a methodological question. Given all the ways in which the particular discourse could be produced, how do you make a claim in any particular case that its coming from one source and not another? Tom Bloor: Concerning inconsistencies, I would say that the very fact that we can socially recognise inconsistencies is proof that theres something there which we could call an ideology. But it doesnt exist independently. For example, if

154

Current Issues in Language and Society

someone whom we recognise as a progressive humanist suddenly comes out with a comment in favour of hanging, we will be shocked. Nevertheless we dont have hard and fast rules that say the ideology must determine the attitude. In fact there is a kind of prestige in our society attached to people who dont accept ideology-packages even though most people dont fit neatly into one particular ideology. So its a cloud, a fuzzy mass, rather than a rigid structure. Teun van Dijk: My view is that we should try to account for both, when you get these inconsistencies for example s/he is liberal on economics, but tough on punishment. These combinations and configurations exist. However, what is interesting here is not the inconsistencies, but the expectations we have, as listeners, that attitudes will be consistent. Ideally, someone who is for hanging is also a racist and a redneck and all these things that we expect them to be. That expectation is also a socially constructed thing. It is something we reproduce by our discourse. So, what I am talking about is not the description of an individual, but the description of groups. This takes us back to the problem of group definition mentioned above i.e. is a group defined by an ideology? Thats what I want to find out. Srikant Sarangi (University of Wales, Cardiff): I think weve been looking here at the positioning of the discourse analyst carrying out ideological analyses as opposed to discourse analyses where the discourse becomes the way through. Ideology is embedded in the text, it is presupposed and naturalised in the way the text is produced. So, in that sense, its there. But its not explicit and, therefore, we study speech acts or opinions or attitudes. In a sense, it seems to me, as a discourse analyst, that the problem is the ideology that you have as a discourse analyst and the way in which you use your ideology and language to carry out the analysis. Teun van Dijk: Its a theoretical concept, invented obviously to explain certain features of discourse. Srikant Sarangi: Ideology is also an analytical construct of our scholarship to explain social attitudes in discourse opinions. Teun van Dijk: When we read the text of the newspaper editorial for example, we are able to say that the ideology of the author is liberal. My question, then, is how do we explain the fact that we are able to make that kind of judgement. And when we read other texts in a particular paper on a specific topic, for example terrorism, we will also notice (sometimes explicitly, sometimes implicitly) a certain coherence in the way this topic is being reported and I call this ideology. Now, my question is: how do we explain that social competence which we obviously have that enables us to say a certain author has a certain attitude or position. And we share this competence with others. This is what Im trying to explain by introducing this theoretical notion. Alexandra Korol: We shouldnt forget that what we are reading, what is being reported is the reproduction of the ideology of society in general, the personal view of the individual author and the policy of the particular newspaper. It is a complex, ideological relationship. Its not only me as a reader, an actor trying to interpret this text: I have to try to deconstruct what is already there.

Debate

155

Teun van Dijk: This links back to your previous question: can a whole society have an overarching ideology which keeps everything together? An ideology that explains everything? However, not all individuals in the society will identify with this overarching ideology as was proved in the easy transition from communism to capitalism in Central and Eastern Europe. In socialist countries, those people who perhaps did identify with this ideology were the ideologues, the people who consciously thought about the ideology as something in its own right. Ulrike Meinhof: If newspapers and individuals are reproducing the ideology of society, and this would have been particularly the case in the more monolithic societies of these countries, where would dissent find its expression? As you say there is consistency and reproduction and coherence, but at the same time you allow for these fractures and fissures in peoples mental models. Where do you find traces of that?

What Evidence Can We Get from the Sample Text?


Kay Richardson: As a response to this, could we try and link this with the text you have given us, this newspaper editorial. Can I ask about the question of voice in all of this? Because it seems to me that this is absolutely fundamental. You could do a straightforward, Machiavellian analysis, even with a clause-by-clause approach. Or we could do a global analysis of the text, an overall analysis of how ethical those in power need to be. The problem with using voice is that it wont let you do that, its really difficult textually to see whether that Machiavellian analysis is the authors, or whether the author is trying to persuade the politicians to do whats in their interest, or just ventriloquising the politicians and inviting his readers to do a cynical reading. I cant tell which is the right way of reading it and who is speaking or who is speaking for whom? The textual traces arent strong enough for me to know what game the author is playing. Robin Warner: A factor in this text that is interesting for me is the use of what I call reproduced discourse. I use this term because none of the other terms, such as intertextuality or indirect speech embrace the whole range I want to talk about Im using discourse in a wide sense. There are fairly unproblematic things you can do with a text like this: for example, you can look at the definite descriptions and presuppositions. When the author refers to the West for instance, there is an assumption that it is a coherent political entity. Are we to accept this? This article is also about taking control of the discourses of others, bending them to a certain end. To what extent can we try and think what was actually said, given the way in which this reproduction has coloured and shaped the original discourse? Angel Gordo Lopez: We should be careful with interpretations. There is a danger in such a grand project that we might be mapping out ideologies based on texts. There is the political agenda which we must be aware of, a grand narrative that is controlling people, and there is also the danger that in trying to formalise and generalise things in a model, we are not listening to minor voices, or that these will be ignored these traces of dissent as Ulrike said. Even from our formal perspective, we should try to recognise variety. Analyses based on Bakhtin, on

156

Current Issues in Language and Society

postmodernism and post-structuralism would come up with quite different methodologies and interpretations. Teun van Dijk: One or two brief reactions. First of all Im not a formalist in that sense, Im not trying to formalise for the sake of formalisation Im theorising at this moment. The second thing is that Im trying to approach one big question, namely that of ideological relationships : the way ideologies show up in discourse; how discourses can be understood or analysed as ideological. There may be other ways to do it, but for me it is by using the scholarly instruments which I understand, which I can handle and manipulate. Psychoanalysts and literary theorists would probably approach this in other ways. However, my experience of literary theory has taught me the limits of this approach and Ive moved away from more intuitive ways of talking about this. This is why I dont have much affinity with postmodernist and post-structuralist approaches, because they do not allow me to explain things, to analyse things. With regard to the political agenda and what you term the narrative controlling what people say, Im aware of the postmodernist approach that we dont have evidence. I think its nice as a hypothesis, but it cannot be confirmed. There is evidence that what people say is controlled by what you call narratives, but what are these but ideologies? All the evidence I have comes from the area of racism and it may be that this is easier, in the sense that in other fields it might not be so clear that people are controlled by these things. It may well be that many of the articles in a newspaper do not have these clear voices that you can hear or this grand narrative behind them that you can see. The other general definitions of ideology, for example the Marxist notion of false consciousness, are too vague. And then there are the endless debates about what false consciousness is. And none of this helps me. I want to analyse it and take it apart. I want to know what is involved in this complex theory of ideology, therefore I go to more concrete things like opinions, or a model people have of specific structures of discourse. But on the other hand, Im not an empiricist who looks at discourse structures and hopes that a theory will emerge. Because for me its obvious that this happens the other way around. Paul Chilton (University of Warwick): I also dont think that postmodernism is helpful. I would like to respond to Kays question and return to the text. It seems to me to be very important to look at this kind of text not in isolation but in the whole historical context that might include recent history. The scripts, the programmes that have been articulated are in some sense ideologically driven. In order to make sense of the text, you dont just have to bear recent American foreign policy in mind, you have to look at something more long-standing than that, discourses if thats the right word that were being constructed in the period of the Cold War and in particular just after World War II. The discourse or script that is being invoked here and on which the reading depends is the American obsession with the dangers of appeasement. Thats not mentioned explicitly, but then an American reader would be aware of that. The second thing has to do with more recent debates about American policy in the former Yugoslavia. The paragraph beginning President Clinton says

Debate

157

starts with a straightforward instance of reported speech, but it finishes with the sentence The excuse is feeble, which guides our interpretation and how he wants us to read it. From this and knowing the context, we infer his position. There are some more examples where he is using other peoples words. Were dealing with a very complex set of political and linguistic interactions. The reading depends on a lot of prior, stored knowledge, scripts, discourses, scenarios which I suppose one could call ideology. But if one did, I would want to bring in the notion of power, because ultimately it has to do with power. Srikant Sarangi: I think in your analysis of the text you refer to what is right and wrong in terms of ideologies. I wonder to what extent the moral dimension and ideologies collide or connect? Teun van Dijk: Its a summarising term to refer to the whole value basis of ideology. So, ideologies are constructed from value components. In relating this to texts, are we able to tease apart the various voices to try to identify the ideologies and morals? For instance in this particular text, is Anthony Lewis speaking for himself, or as an American columnist? Also, we need to consider the selection of particular words, such as the use of the word dismemberment. This is certainly not a neutral word and its certainly not an innocent term. And many other terms and structures are not either. So we should ask: is the choice of this word controlled by a specific opinion and, if so, what kind of opinion is it positive or negative? Do all the choices, lexical and otherwise, point to a particular perspective? Christina Schffner: I think this question of choice is very important, and it is also linked to intertextuality. Terms such as dismemberment have been used before in the context of international relations, they have been established as political concepts, and have thus acquired a certain meaning and value. Thats why they can readily be taken up in a new context, to be confirmed or challenged. Teun van Dijk: When we speak of intertextuality in this context, it is the textual manifestation of what I see as an ideology. The continuity expressed through intertextuality simply means that not only can the mental structures be the same, but the formulations too. So we have a very strange kind of intertextuality where people in different countries talking about the topic, not only have the same attitudes, but also use the same formulations. They could not have heard them from each other. Kay Richardson: But you have to work with the voice that they offer you. The journalistic voice in this article is very hot under the collar. Hes got righteousness on his side, most of the time. He doesnt really need this approach. Paul Chilton: No. This hot-under-the collar approach is necessary for legitimising his argument. That is part of the discourse of foreign policy. Kay Richardson: To me, the hot-under-the-collar approach is just the journalistic voice. In the rest of the text he is talking to these politicians who are constructed as being completely cynical. Robin Warner: There are a lot of contradictions in the text. This sort of contradiction and uncertainty I find quite characteristic of ideologically condi-

158

Current Issues in Language and Society

tioned discourse. What bothers me is why he has to be so indirect about expressing his opinion about American foreign policy. Sharon Imtiaz (University of Warwick): Has that not got to do with the medium in which the article appeared? Its the International Herald Tribune. Hes not talking solely to a domestic audience. Teun van Dijk: But all the articles are taken from domestic newspapers. So it is an American audience hes talking to first. Sharon Imtiaz: But its then redirected and so we have the phenomenon of global intertextuality. Even rhetoric becomes recontextualised. David Graddol: There is a multiple nature to the audience for a medium like this, even if the articles were originally written for a domestic audience. Going back to voice, what struck me at a deeper level in this text is how many of the propositions are assigned to other people. It is woven out of the voices of other people and what is problematic is knowing where the author identifies with those opinions. This notion of possession of opinions is problematic for me that someone has an opinion, that you want to find out that opinion and how the discourse emerges from the holding of that opinion. I was interested to hear people saying how they find literary theory useless for this purpose, because Bakhtinian theory, as I understand it, comes from literary theory and it seems to be one of the most useful ways of approaching this. You get simple formulations, because everybody is appropriating each others language, but that doesnt necessarily mean that everyone has the same opinions and beliefs. Its problematic because people are also putting in distancing devices, evaluations of these opinions or they might even be trying out certain opinions to see how they are working in a particular social context. We may have representations as you call them of these opinions and know what kind of opinions are appropriate to have in particular social contexts. But where we position ourselves in relation to those opinions is much more problematic. Teun van Dijk: As readers or as analysts? David Graddol: As individuals speaking and that makes it even worse to try to analyse the speech of other people. But why is the author so indirect here? Why is so much of the passage made up of other peoples words? Srikant Sarangi: Isnt it also a question of other voices being silenced? Its not just what voices are represented, but which arent. We need to ask, who is not quoted? Teun van Dijk: I agree. From the discourse analysis point of view, one of the most interesting things is not to find out whether Anthony Lewis is pro or contra American policy, but to see in which complex way that opinion is being managed in this text for a specific audience. In all the ways this is being done, theres a long list of complex strategies: moves, ways of mitigating, emphasising, avoiding opinions he agrees with but also finds reprehensible. Kay Richardson: What is interesting is that although values are invoked, they are being invoked so that they can be manipulated between a discourse that is

Debate

159

completely cynical about them and Lewis discourse that you should do moral things for the sake of morality. Isnt that the play-off? Julian Edge (Aston University): Not knowing anything else about the author or his views, if I want an interpretation without comments, I would say that the text is about attacking President Clinton. What Im hoping for now is some help with the idea of opinions grouping into attitudes and attitudes being governed by an underlying ideology and how this model is illuminating the text. Teun van Dijk: To do this, you need a large number of texts which Im currently compiling. Analysis should provide evidence of strategies and show how basic, ideological propositions show up in these texts. For example, in an argument, people will say this is my opinion. In some aspects of argumentation, people will resort to using something we all know or we all agree upon and this is where we get into the area of attitudes. Sometimes it can be more fundamental, involving a proper ideological proposition. In this kind of analysis, you get these various levels of generality and sharing. Lorna Milne (Aston University): I think we should also consider the notion of an ideology as having a vocation to transfer itself and we havent talked at all about propaganda or the transfer of ideology. I just wonder whether this catchall, have-your-cake-and-eat-it kind of discourse is one which is particularly useful when trying to persuade people. The more you can show two facets of an argument, the more people you can recruit to your way of thinking. Robin Warner: Im not sure about this. I find that most political discourse uses other peoples discourses. A lot of the time it seems to be about talk and how talk is evaluated, not about reality. Paul Chilton: I would like to comment on Davids question about the indirectness. I think you might get some answers to this by drawing on politeness theory. As Julian mentioned, the text is a criticism and indeed we do find some face-threatening acts in the text and these have to be justified. For example, the author has to explain why the USA should commit its forces to Yugoslavia. The persuasive content can be seen as that kind of active self-justification. To do this, he seems to trigger off existing discourses through lexical cues. For example, he refers to the notion of what a state is in international law. Another example is where he invokes directly a Hitler figure, which is not only World War II discourse, but also evokes the 1990 Gulf War discourse. This is the way historians have talked about it. So, there are many cues with various resonances in the text. Teun van Dijk: I think thats why so many people working in the area of discourse analysis find the notion of repertoire useful. They say its not just underlying ideas and structures but also collections of expressions and so on. For example, in the text, when Lewis speaks about the dismemberment of the country, he is invoking a basic argument that although it is not a nice process, it is necessary for human lives to be saved. His argument about what is allowed and what is not the humanitarian attitude of avoiding death, of valuing life these are very basic culturally-shared things. He has to concede that, because he cannot ignore this fundamental value of our culture. So, now he is no longer arguing superficial things, relating to what Clinton has to do, he is much more concerned with arguing basics of our society and therefore he has to say that this

160

Current Issues in Language and Society

argument has some force. But, having said this, there is the disclaimer, the but; But those who make it. Then comes his own basic principle the abandonment of principles. So, what I read from this as a suggestion for my theoretical construction is that apparently people have some very basic value principles and that they have some kind of hierarchy or priority that gives some order to them. In the authors hierarchy, justice is even more fundamental than saving lives. So, I read this kind of argument as a manifestation in the text of some kind of structure in the underlying system.

From Discourse Structures to Attitudes, Opinions and Ideologies?


Julian Edge: Arent they two different propositions? The one being that underlying values are used to justify opinions and positions. The claim that through identifying opinions and attitudes, one can come to an underlying ideology. One needs to find evidence for that tripartite organisation in discourse. Teun van Dijk: Yes, especially through clusterings of opinions that reveal attitudes. These can be identified through, for example, the choice of words, in his argumentation and legitimation strategies. Take, for example, political correctness. When you have a position like this, it is assumed that you will adopt a particular vocabulary. Here, in particular, we can see the relationship between discourse and ideology. The question for us as discourse analysts is, is it useful? Does it bring us anything interesting? I would say, yes. Otherwise I would not be able to explain continuity in lexical choice. Sue Wright: Isnt this all rather dangerous? You see some attitudes and then you surmise the ideology. You have to be aware that you are expecting clusters of opinion to occur. If you expect to find the whole set of attitudes that go with that ideology, isnt this a form of stereotyping? Teun van Dijk: Exactly. As you say thats the danger in our everyday life. We expect people to have consistent ideologies; when people say a, we expect them to say b, c, d as well I want to explain why this is so, why people do that. People are stereotyping and generalising, from hearing a, they conclude the rest which, for me, is evidence that this actually happens. The dangers you mentioned link in with ethics the ethics of everyday life and the ethics of scholarship. Sue Wright: Yes. I feel very insecure about the way we unpick texts. It seems to me that a lot is taken out with very little to go on and we bring so much of ourselves to the process as well. In this kind of analysis, we make so many assumptions on the basis of so little evidence, that we often reveal as much about ourselves as anything else. Paul Chilton: Yes, but what is ourselves? Surely its us as the reader informed by interaction with other discourses and cultural knowledge, ideologies which we bring to the text. Sue Wright: But its interesting that what were talking about is not what we have brought to the text, but what is coming from the text to us. Paul Chilton: Well, I think of myself as someone interacting with Lewis.

Debate

161

David Graddol: Do you? Or do you actually think of yourself as someone interacting with the text? Helen Kelly-Holmes: There may be a danger that we find in the text what we want to find. Where, as analysts, we are sure that the author has a very strong ideology, there is a temptation to find evidence to support this. Julian Edge: It seems less dangerous than speculation without any data. Ulrike Meinhof: Yes, and this text is riddled with presuppositions and you dont have to postulate anything, you can just identify them. This shows the advantage of having a discourse analysis toolkit. We dont say that this is what the text means; were just saying that these items exist in the text and that they can indicate a certain attitude. But my problem is that I still dont know whether I necessarily find evidence of something that I would be prepared to call an ideology. I like the word cluster or filter. For me, ideology is a much more fixed term, much more dominant and I would be wary of trying to discover this through examining statements. I think its very difficult to argue for any kind of fixed ideology, because this would be totally unshakeable, made up of discursive formulaic speech such as Thatcherism was. Teun van Dijk: This brings us back to the contradictions. People are exposed to the discourse of party political ideologies etc. in the public sphere, but they also interact with groups and in groups and this may give rise to another type of discourse. Ulrike Meinhof: But this is a superficial way of using the word ideology. I think from your paper youre trying to get at something much more complex. We use phrases like Tory ideology as a short cut, meaning a cluster or set of beliefs, that translate into the individual ticking a particular box. What is interesting for me is how these competing sets of clusters survive and how they are not in fact governed by this very monolithic set, which Ill call an ideology, but how they keep on reproducing such things as racist discourse not just in the way people speak, but in the way they act. Ideology doesnt help me with that. Teun van Dijk: But I find that it helps me to explain certain things and I dont have other instruments for this. Discourse analysis can highlight and select. But I want to go beyond this to another level of description and explanation, and social psychology offers things such as impression management and face-saving, that can help me. But these are just elements in a more complex system. Why, for instance, when people make a negative comment about somebody else, do they follow it with an apology or use face-saving strategies? Such strategies in connection with racist discourse reveal links to ethics and power, and this brings you to questions of basic norms and values. So in order to explain whats behind this, you dont just have one word or one structure you have all these layers of explanation. I can get a long way with notions or concepts from discourse analysis and they explain a lot but not everything. Now if I bring discourse and its structures and ideology together in a clear, systematic way, that explains what I am after. I would love to find another word rather than ideology: the term is always loaded and I keep having to defend the concept of ideology whenever I use it.

You might also like