This article was downloaded by: [PERI Pakistan] On: 21 April 2010 Access details: Access Details: [subscription

number 778684090] Publisher Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 3741 Mortimer Street, London W1T 3JH, UK

Journalism Studies

Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713393939

CAN DISCOURSE ANALYSIS SUCCESSFULLY EXPLAIN THE CONTENT OF MEDIA AND JOURNALISTIC PRACTICE?
Greg Philo

To cite this Article Philo, Greg(2007) 'CAN DISCOURSE ANALYSIS SUCCESSFULLY EXPLAIN THE CONTENT OF

MEDIA AND JOURNALISTIC PRACTICE?', Journalism Studies, 8: 2, 175 196 To link to this Article: DOI: 10.1080/14616700601148804 URL: http://dx.doi.org/10.1080/14616700601148804

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

CAN DISCOURSE ANALYSIS SUCCESSFULLY EXPLAIN THE CONTENT OF MEDIA AND JOURNALISTIC PRACTICE?
Greg Philo

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

It is argued here that textual analysis of media accounts requires the study of the social structures from which competing ideological explanations develop. A comparison of the methods of the Glasgow University Media Group with the work of Norman Fairclough and Teun van Dijk shows that discourse analysis which remains text-based has problems in its ability to show: (1) the origins of competing discourses and how they relate to different social interests, (2) the diversity of social accounts compared to what is present (and absent) in a specific text, (3) the impact of external factors such as professional media practice on the manner in which the discourses are represented, and (4) what the text actually means to different parts of the audience. There are other problems with text only analyses in relation to (1) the accuracy of representations, (2) the significance of texts to our own audience, and (3) the question of how rhetoric belongs to or is used by different social interests. To overcome these problems requires a method which analyses processes of production, content, reception and circulation of social meaning simultaneously. KEYWORDS critical discourse analysis; Glasgow University Media Group; media content methods; Norman Fairclough; Teun van Dijk

Introduction
This article outlines the methods developed by the Glasgow University Media Group and compares them to discourse analysis in the work of two theorists, Norman Fairclough and Teun van Dijk. These authors do not represent the whole of discourse analysis but they are prominent in the area and their studies provide useful points of similarity and difference with the methods which we employ.1 The main issue I will raise is that their textbased studies are limited in the conclusions which can be drawn, since their approach does not include the study of key production factors in journalism or the analysis of audience understanding. Finally, I will show through a case study how it is possible to study simultaneously the three processes of production, content and reception of news messages.

Van Dijk and Fairclough: Discourse and Ideology

Teun van Dijk is well known for his work in analysing racism in news accounts. He points to the differences between traditional content studies in this area and between that of his discourse analysis which focuses on a systematic description of semantic and syntactic features of text: Journalism Studies, Vol. 8, No 2, 2007
ISSN 1461-670X print/1469-9699 online/07/020175-22 2007 Taylor & Francis DOI: 10.1080/14616700601148804

176

GREG PHILO Traditional approaches to the role of the media in the reproduction of racism were largely content analytical: quantitative studies of stereotypical words or images representing minorities . . . Discourse analytical approaches, systematically describe the various structures and strategies of text or talk, and relate these to the social or political context. For instance, they may focus on overall topics, or more local meanings (such as coherence or implications) in a semantic analysis. But also the syntactic form of sentences, or the overall organisation of a news report may be examined in detail. (2000, p. 35)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

There are many different theoretical strands in discourse analysis and the word discourse is used in varying ways. It is used abstractly to mean statements in general or to refer to a particular group or type of statements (as in a discourse). John Fiske refers to it as a language or system of representation that has developed socially in order to make and circulate a coherent set of meanings, which serve the interests of a section of society (1987, p. 14). The important point here is that for critical discourse analysts such as van Dijk and Fairclough, discourse is linked to power and social interests. From such relationships, there emerge different perspectives on the world. As Fairclough writes:
I see discourses as ways of representing aspects of the world * the processes, relations and structures of the material world, the mental world of thoughts, feelings, beliefs and so forth and, the social world . . . different discourses are different perspectives on the world, and they are associated with the different relations people have to the world. (2003, p. 124)

Fairclough and van Dijk are particularly concerned with ideological effects of discourse. As Fairclough notes:
One of the causal effects of texts which has been of major concern for critical discourse analysis is ideological effects . . . ideologies are representations of aspects of the world which can be shown to contribute to establishing, maintaining and changing social relations of power, domination and exploitation. (2003, p. 9)

Van Dijk notes that a key function of ideologies is to promote and coordinate the interests of a group and comments that dominated groups also need ideologies as a basis for resistance. That said, he reaches the conclusion that:
It is of course true that many ideologies develop precisely in order to sustain, legitimate or manage group conflicts, as well as relationships of power and dominance. (1998, p. 24)

The intellectual origins of many approaches to ideology (and notably Faircloughs) are in structuralism * in left/Marxist variants such as in Althussers work (1969) for whom ideology was ultimately a function of class power in capitalism, to the development of it by Foucault, who shifted the source of power to language itself. In his work discourse is a social force which has a central role in what is constructed as real and therefore what is possible. It determines how the world can be seen and what can be known and done within it. Discourse is thus crucial in explaining how the social subject is positioned and limited. A key question which he highlights is: how are we constituted as subjects who exercise or submit to power relations? (1994, p. 318). What emerges from these theoretical developments is a concern with how language embodies systems of thought which structure what can be understood. For example, in the earliest work of the Media Group we showed how news language was organised

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE

177

around very limited ways of understanding economic and political activity. In the 1970s, trade unions and their wage claims were blamed for economic problems such as inflation. There was alternative evidence to suggest other causes but the government view dominated the news. As we wrote:
What there is here is an illusion of balance, whereby reports are included from what appear as different sides. But the reported views have a totally different status, legitimacy and meaning in the text. In a very real way, only one set of statements makes sense in that we are systematically given the information necessary to understand the explanations and policies which they represent. (Philo et al., 1977, p. 13)

In our conclusion to More Bad News , we were critical of broadcasters claims to objectivity and impartiality when the news was actually reproducing the assumptions of the powerful about what was necessary and possible in our society:
The bland assertion of objectivity and impartiality . . . serves only to obfuscate what is in fact the reproduction of the dominant assumptions about our society * the assumptions of the powerful about what is important, necessary and possible within it. (Glasgow University Media Group, 1980, p. 115)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

In his recent work on textual analysis, Fairclough writes on ideology in terms which have some resonance with this:
A particular discourse includes assumptions about what there is, what is the case, what is possible, what is necessary, what will be the case, and so forth. In some instances, one might argue that such assumptions, and indeed the discourses they are associated with, are ideological. (2003, p. 58)

In illustrating what is seen to be necessary, Fairclough takes the example of global economic change. He notes that this may be presented as an inevitable process, without human agency and comments on a text published by the European Union:
It is similar to many other contemporary texts in representing global economic change as a process without human agents . . . a process in a general and ill-defined present and without a history (it is just what is) which is universal (or, precisely, global) in terms of place, and an inevitable process which must be responded to in particular ways * an is which imposes an ought, or rather a must. (2003, p. 45)

We offered a similar analysis in our Really Bad News , and noted how the harmful effects of movements in the world economy were likely to be treated on the television news as a form of natural disaster, rather than as the result of human decisions:
Recession, inflation and unemployment, if they are not being blamed on wage claims, were in the period of our study most likely to be treated as natural disasters. The world economy is presented as an omnipresent force, and movements in it . . . are the problem, but these movements are rarely explained for what they actually are . . . a multinational firm may be reported as regrettably being forced to close a factory in the north of England because it is uneconomic, but will not usually be spoken of as having made a decision to move its capital somewhere else because it can make more money there. (Glasgow University Media Group, 1982, p. 130)

178

GREG PHILO

Ideology and the manner in which some perspectives are legitimised and achieve dominance remains a central issue in textual analysis. The methods which are employed are the focus of the next section.

Three Methods in Textual Analysis

I will look first at the thematic analysis developed by the Media Group, then at Faircloughs work on the semantic and grammatical features of texts, and finally at van Dijks use of what he terms the ideological square. The Media Groups work began with studies of television news. We focused on major thematic areas such as industrial struggles or international conflicts and then examined the explanatory frameworks or perspectives which underpinned the descriptions which were given. In any contentious area there will be competing ways of describing events and their history. Ideas are linked to interests and these competing interests will seek to explain the world in ways which justify their own position. So ideology (which we defined as an interest-linked perspective) and the struggle for legitimacy go hand in hand. Much of our work focused on the role of the media in these ideological struggles and how the reporting of events can embody different ways of understanding. We were interested in how language was linked to wider social processes and how individual meanings and communications related to conflict and divisions within the society as a whole. The language and definitions used were at one level the battleground for competing groups. The issue then was not to look simply at the descriptions which were offered of the world in a specific text, but to look at the social relations which underpinned the generation of these descriptions. Thus in our recent work on television and the Israeli Palestinian conflict we gave an extended historical account of the dispute and showed how each phase in it had generated competing histories of what had occurred (Philo and Berry, 2004). When the conflict is reported in media on a daily basis, both sides struggle to assert the validity of their own accounts. This is so for each event in terms of descriptions about what has happened and who is to blame. But it also relates to the more general frameworks of understanding and interpretation which underpin the public relations of each side. For example, when we analysed news reports of the intifada between 2000 and 2002, we were puzzled as to why the Israelis were not at that time stressing the issue of anti-semitism as part of making their case. There was evidence of anti-semitism in the speeches of some Muslim clerics, so we asked experts in public relations why there had not been a great emphasis on this. Nachman Shai had been chief spokesperson for the IDF (Israeli Army) and he described the decision to focus on the war on terror rather than anti-semitism:
We selected the first [war on terror] instead of the second [anti-semitism] because we are part of the Western world. We very much played the first argument. It worked better with governments, they gave us more support. Its like if youve run out of arguments, youre stuck with anti-semitism. The first one is based on common interests. (Interview, in Philo and Berry, 2004, p. 249)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

It is also the case that for Israel to present itself as part of the general war on terror against those who dislike Western values also has the advantage of drawing attention away from specific actions by Israel which have contributed to the origins and development of the Middle East conflict. The Palestinians would for their part see the

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE

179

Israelis as state terrorists and would point to what they see as Israeli violations of international law in imposing a military occupation in the Palestinian territories. The important theoretical point is that the interests behind an ideology may remain the same, but its immediate content does not. The parties in a conflict deploy different arguments in relation to constantly changing circumstances. To understand this process necessarily involves going beyond the immediate text. Our approach thus began from the assumption that different ways of explaining the world emerged from such conflicts and from other social divisions such as those between classes and subgroups within these, from interests based on gender or from competing institutions. Such an approach necessarily involved us in the detailed analysis of public debate and its origins in political and social conflict. This analysis was undertaken before work could begin on examining specific news texts to show how various perspectives were represented (or not) within them. So an important first step in our method was to identify the range of arguments which existed on an issue. This showed what was available for journalists to choose from, as well as which arguments belonged to different interests and also made it possible to explore the conditions under which they were deployed. In our work on TV news coverage of industry and the economy, we examined the public debate which existed at that time about the failings of the British economy (Glasgow University Media Group, 1980). In the 1970s and 1980s, this was a matter of great concern as Britain was perceived to be falling behind the rest of the worlds economies. In this public controversy the trade unions pointed to management mistakes in the organisation of industry and to low levels of investment which meant that machines often broke down and production was much less efficient than that of competitors in other countries. In contrast, right-wing commentators (including the Conservative Party) preferred to point to the actions of the workforce and blamed strikes for the failings of the economy. This became a favourite theme of the Conservatives in the 1980s, in the early years of Margaret Thatchers government (see Philo et al., 1995). We were interested in how the TV news reported such arguments and the potential influence on public belief. We began by noting each of the explanations and ways of understanding which were put forward in public debate and the range of available evidence which could underpin different positions. We identified these from public materials such as books, reports, the press and TV, and any other relevant sources. From these we constructed what was in effect a conceptual map of the different beliefs which were available in the political and economic debate of that time. Using this, we could then analyse how different parts of the debate were featured in the news. A key issue was the absence or presence of explanations and the manner in which some accounts were highlighted or preferred in the text. We distinguished between statements that were simply reported by journalists as being from a specific source and those which they directly endorsed. This can be seen in the difference between a reported statement such as the government says that strikes are a major problem as compared with a direct statement as in this example:
Its the kind of strike that has contributed significantly to the dire economic problems. (ITV, main news, 4 January 1975, cited in Philo et al., 1982, p. 29)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

If Fairclough was analysing these types of statement, he would refer to them as indicating differences in modality in as much as they relate to differences in commitments, stances, and therefore identification (2003, p. 166). We did not always use the

180

GREG PHILO

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

categories of formal linguistics and developed our own descriptions of what was occurring.2 But it is clear that there are some parallels between these semantic and syntactical features of text which we examined and the later work developed by Fairclough. For example, he uses the concepts of hyponymy and synonymy to mean a pre-supposed semantic relationship. The first implies an inclusion and the second an identity between words. For example in the case of hyponymy, a neo-liberal might treat globalisation and economic progress as hyponyms. The point is, that to make such a relationship can favour a particular perspective. In hyponymy one word is strengthened by its suggested closeness with the other. An even stronger relationship is in synonymy, where words are treated as being so close that they are interchangeable. We illustrated such a use of language in our study of strikes in the car industry. We had shown how the lack of investment in new plants meant that machines broke down very often and that much production was lost. This, however, was ignored on the news and when destruction of output occurred it was related very largely to strikes. So when the strike ended, the plant was referred to as returning to full production, even though such normal periods included a loss of output which was greater than that caused by strikes in the factory. For example, a journalist commented that:
With the engine tuners back at work . . . the plant was also back in full production. (ITV, lunch-time news, 7 January 1975)

We noted in our commentary on this that normal production and full production are treated as synonymous and are equated with being strike free (Philo et al., 1982, p. 36). Fairclough also uses the concept of collocation. This means a regular or habitual pattern of co-occurrence between words. He gives the example of poor old (as in poor old man) as being a more habitual combination than poor young (2003, p. 213). We showed this process in our account of the treatment of wage claims in the news in the 1970s. At the time, inflation was high and was routinely linked in reporting to wage demands by trade unions. We noted at the time that there were many other causes of inflation, such as speculation on property or increasing oil prices. The point about focusing on wages was that the government believed it could reduce inflation by forcing down wages and controlling wage claims. The view that wages had actually caused inflation was contested in public debate but the TV news nonetheless very extensively featured the government view, as in these examples:
[The Chancellor] has warned again of excessive wage increases as the miners start negotiating. (BBC1, main news, 11 February 1975, our italics, cited in Philo et al., 1982, p. 61) [The Chancellors] warnings about wage-led inflation and pay rises well in excess of the cost of living. (BBC1, main news, 16 April 1975, our italics, cited in Glasgow University Media Group, 1980, p. 46)

So excessive occurs in collocation with wage increases. The link became routinised as the news monitored each new wage claim and commented on its potential effect on inflation. We developed this conception of how meaning was established by noting that such a relationship could become routinised to the point that journalists might actually dispense with the emotive word excessive. As we wrote:

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE When this economic view is pursued, the logic of who is to blame is inescapable. It seems perfectly natural to monitor wage claims . . . this becomes so routine that journalists could dispense with apparently emotive terms such as excessive. They have only to say and tonight another wage claim, for everyone to know what they mean and at whom the finger is being pointed. (Philo et al., 1982, p. 60)

181

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

We also developed at this time the concept of the explanatory theme. This is an assumed explanation which gives a pattern or structure to an area of coverage. For example, the explanatory theme that strikes were to blame underpinned whole processes of news reporting. This might include going to a factory, interviewing workers, asking them about strikes and crucially not asking the management about investment policies or their own mistakes and then perhaps listing in the bulletin other strikes which had occurred that week. The crucial point is that the pattern of coverage and the subjects that it highlights can assume the explanation even without it being directly stated. Not all news is as closed as this suggests. There are a range of factors which can influence news output, some of which produce texts which are organised very tightly around such explanatory themes. But other factors can generate a greater diversity of explanation. To analyse these requires an approach which necessarily goes beyond the content of specific news texts and looks more broadly at the processes which underpin their production. I will suggest later that the absence of production studies in the work of Fairclough and van Dijk means that the conclusions which they can draw are limited and sometimes problematic. Part of our work does focus exactly on this link between production and content. We can look briefly at this now and the issues which it raises for the development of appropriate methods.

Production Processes and the Analysis of News Texts

The first issue to consider in studying production processes is the professional ideology of journalists and the institutions which they represent. Some news is dominated by specific explanations and ways of understanding but such assumptions are overlaid by other beliefs and practices. They include the need to feature some form of apparent balance between views * if only at the level of interviewing opposing sides. The credibility of television news and the legitimacy which it seeks for itself depends upon its claim to be even-handed and fair in controversial areas. Our research suggests that it is skewed in favour of the powerful, but the broadcasting institutions are intensely reluctant to be seen as simply the mouthpiece of the state or other major interests. There is sometimes a real substance to their claims to be featuring a range of views. How balanced they can be depends in part on the area of news. On issues where the state is very sensitive, such as in coverage of Northern Ireland in the period of the troubles, the news could become almost one-dimensional * alternatives were reduced to fragments or disappeared altogether (see Miller, 1994). But even in such a closed area, there are still cases where journalists have jeopardised their own careers and become involved in intense controversies in order to release information and produce stories which they thought should be told. This is done from a sense of personal and professional commitment. In 1985, BBC journalists went on strike to protest about the banning of a Real Lives programme featuring elected representatives from Northern Ireland. The government had asked that the programme be not shown and the Board of Governors of the BBC had concurred with this (for other examples from Northern Ireland see Glasgow University

182

GREG PHILO

Media Group, 1982, pp. 140 3). We also encountered instances where journalists had deliberately undermined what they saw as the official line of their own news organisation. For example, during the Falklands War, the British military had ordered the bombing of the air field at Port Stanley, the capital of the Falklands, which was occupied and being used by Argentine forces. The British Ministry of Defence (MoD) claimed initially that this attack was successful and that the air field could not be used. There was, however, evidence that it was still in use, including film of it in operation which was being flown out by the Argentines. But the TV news stayed with the official government view that it was out of action. The MoD was then forced to make a concession, so it issued a statement that the air field could be used by light air craft. An ITN journalist told us that when he received this, he deliberately went to find pictures of a very large Hercules transport plane going along the run way. He then placed these as the visual background to the MoD statement. We noticed this in our analysis as a contradiction in the text between the images and his report that:
Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

The MoD now concede that light aircraft can now use the runway. (ITV, main news, 14 May 1982, cited in Glasgow University Media Group, 1985, p. 86)

The point is that textual analysis alone could not reveal what had occurred and we had no way of understanding this until we spoke with the journalist. There are other factors which should be considered in the news production process including the use of sources, the organisation and logistics of news gathering and market pressures. The relation to outside sources is crucial and we have studied the manner in which external interests such as governments, corporations, non-governmental organisations and other lobby groups seek to promote their own perspectives. The state has some absolute powers in that it can impose censorship though it rarely acts in such an overt fashion. Perhaps more importantly it is the key supplier of routine information to media institutions in areas such as employment or health or in relation to new policy development. Journalists depend on the access granted to them, and this becomes particularly acute in circumstances such as war or conflict where only a limited number may be allowed in to the key zones for reporting. Politicians in power are in a strong position to insist that their views are featured * especially in the publicly regulated media. We describe the media in our work as a contested space. Not all of those in the contest are of equal power and journalists are subject to intense pressures which can effect the climate in which they work. In Bad News from Israel , we noted a number of key factors which influenced broadcast output including the political link between Britain and the United States and the strong support in America for Israel, plus well-organised lobbying and public relations, together with the intense criticism of journalists who were seen to be putting out negative reports on Israel. As we concluded:
The pressures of organised public relations, lobbying and systematic criticism together with the privileging of Israeli perspectives by political and public figures, can affect the climate within which journalists operate. There is no total control and there are areas of the media where the debate is relatively open. But these factors go some way to explaining why journalists sometimes have difficulty in giving a clear account of the Palestinian perspective, while they can apparently more easily facilitate that of the Israelis. (Philo and Berry, 2004, p. 256)

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE

183

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

Overall, the extent to which media are open to alternatives and may feature many contesting views is likely to vary in relation to conjunctions of political, economic and institutional factors * including differences between publicly regulated and privately owned sectors.3 We might expect a more open expression of varied perspectives when there is no clear account emerging from the most powerful groups or when they are divided. The divisions in the Conservative Party over Europe in the 1990s would be an example of this or the controversy in the Labour Party in the latter stages of Tony Blairs leadership. The privately owned media have often been used to promote political views, which has a limiting effect on the range of arguments and information which is featured within them. We noted this as a factor in the representation of the Israeli Palestinian conflict (see Philo and Berry, 2004, pp. 252 6). The economic interests of the media have strong effects on their political preferences. Not surprisingly the right-wing press in Britain has normally supported the Conservative Party. But the Murdoch organisation has also a tradition of reaching agreements with various political parties in order to gain commercial advantage. In Britain, the Murdoch press had for many years supported the Conservatives and Margaret Thatcher, but in 1997 they switched support to New Labour, following a meeting and an unspecified agreement between Tony Blair and Rupert Murdoch. As Michael White commented in The Guardian :
In every country in which Mr. Murdoch operates (and minimizes his tax bill) he is a power-broker, speaking power, not truth, unto power through his diverse media outlets. The Blairites have . . . made a Faustian bargain with Rupert. They think they have a good bargain. (30 January 1998, cited in Philo, 1999, p. xi)

Most of the media in the UK are commercial institutions in their own right, so the need for market share * to gain viewers and readers * is a paramount concern. We have shown how this can override potential links to government or public interests. For example, media coverage of mental health in the 1990s went largely against government policies of reducing long-term institutional care in special hospitals. The policy of community care was attacked as the media focused on audience-grabbing stories of the dangers posed by the mentally ill * what we referred to as the mad axe murderer syndrome (Philo, 1996). In this area, the desire for market share was crucial and news values which place a priority on fear, drama or spectacular events become central in the choice of stories and the angle taken. The key theoretical point is that all these pressures exert major influences on the content of texts. For example, where there is an intense and controversial debate as in the Israeli Palestinian conflict, journalists might simply seek to avoid giving explanations because whatever they might say will draw criticism from one side or the other. There is also a limit to the amount of time that they will spend giving an account of the views of each side. As Lindsey Hilsum from Channel 4 News commented to us:
With a conflict like this, nearly every single fact is disputed . . . I think, Oh God, the Palestinians say this and the Israelis say that . . . I have to say what both sides think and I think that sometimes stops us from giving the background we should be giving because I think Well bloody hell Ive only 3 minutes to do this piece in it and Im going to spend a minute going through the arguments. (Philo and Berry, 2004, p. 245)

184

GREG PHILO

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

The history of the conflict and explanations of its causes are intensely controversial. We noted how rarely they were referred to in news programmes. It was also the case that when journalists did include references to them, they sometimes spoke in a form of short hand or code which noted an issue but was very inexplicit. Thus an ITV journalist commented The basic raw disagreements remain * the future, for example of this city Jerusalem (ITV, early evening news, 16 October 2001, quote cited in Philo and Berry, 2004, p. 110). Some viewers would already have an understanding of the historical events which gives significance to this comment, for example that Jerusalem is sacred to both Muslims and Jews and East Jerusalem has been occupied (illegally, in the eyes of most countries) by Israel since 1967. But as we showed in our audience research, most of the population simply do not have this knowledge. So the journalists comment has little meaning, in terms of the potential to understand what the disagreement is about. The meaning cannot always simply be assumed using the cultural knowledge of the investigator. It also requires some knowledge of the audience. The key conclusion which we drew, in terms of methods, was that it was not possible to analyse individual texts in isolation from the study of the wider systems of ideologies which informed them and the production processes which structured their representation. Other theorists in the field of media research drew the same conclusion and there were very significant studies undertaken which examined production and the dynamic contention between media and their sources (Hallin, 1989; Schlesinger, 1978). It was also necessary to simultaneously study processes of audience reception before making judgements about social meaning and the potential impacts of texts on public understanding. This linking of production, content and reception became the basis of our methodological approach. With this in mind I will now go on to consider the methods of text analysis employed by Fairclough and van Dijk.

Faircloughs Critical Discourse Analysis

As we have seen, Fairclough is centrally concerned with issues of power and legitimacy and the common sense assumptions which are implicit in the conventions according to which people interact linguistically (2001, p. 2). He terms his approach critical discourse analysis and he challenges social theorists and researchers to show the effects of language in contemporary social life. As he writes, these effects include making the socio-economic transformations of new capitalism and the policies of governments to facilitate them seem inevitable (2003, p. 204). Drawing on Bourdieu and Wacquant, he notes that what is at issue is the process of classification, by which naturalised preconstructions generate a particular vision of the world (Fairclough, 2003, p. 130; Bourdieu and Wacquant, 2001). Thus he notes that discourses can be differentiated:
In terms of semantic relations (synonymy, hyponymy, antonymy) between words * how they classify parts of the world * as well as collocations, assumptions, and various grammatical features. (Fairclough, 2003, p. 133)

I have discussed collocations, synonymy and hyponymy above (antonymy is the opposite of hyponymy).4 My own view is that the master category here has to be assumptions, since these underlie the deployment of the various textual strategies * as in treating economic progress as a hyponym of globalisation. Put another way, the textual strategies are used to frame a description in accordance with the underlying

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE

185

assumption. Fairclough also analyses other dimensions of the rhetoric and promotional messages of politicians such as Tony Blair. In a close analysis of political speeches, he shows how desires are typically represented by Blair as actually existing. For example, globalisation is spoken of as a fact and at the same time as a project or plan (2003, p. 114). He also shows how Blair positions himself with his own audience in terms of what Fairclough refers to as the process of conversationalisation. Blair presents himself as speaking as an individual, saying for example I realise why people protest against globalisation (2003, p. 76). This simulates person-to-person conversation, thus reducing distance and explicit hierarchies (rather than saying for example, the government believes that . . . ). This is an interesting account of rhetorical strategies but there are difficulties with Faircloughs approach in as much as it remains text-based. Fairclough does note the limits of textual analysis and the need to link the micro analysis of texts to the macro analysis of how power relations work across networks of practices and structures (2003, p. 16). But my own view is that text analysis cannot be simply bolted on to other work. There are problems in that it is difficult to explain the construction and meaning of a text without simultaneously examining other factors such as production and reception processes. In essence, I have suggested that discourse analysis which remains text-based encounters a series of problems specifically in its ability to show: (1) the origins of competing discourses and how they relate to different social interests; (2) the diversity of social accounts compared to what is present (and absent) in a specific text; (3) the impact of external factors such as professional ideologies on the manner in which the discourses are represented; and (4) what the text actually means to different parts of the audience. There are three other problems for what we might call text only analyses to which I want to draw attention. The first is the issue of the accuracy of representations and the need to go beyond the text to check these, the second is the question of the significance of the text to our own audience (i.e. the readers of our critical work) and the third is what I will term whose rhetoric is it anyway? We can look first at the issue of the accuracy of representations. A key function of texts is to represent other entities * such as ideas, beliefs or actions (Fairclough, 2003, p. 27). A news text often represents these in terms of the reported statements of whoever is being featured. Fairclough seeks to criticise such representations by showing how they are sometimes framed to give a favourable view of one side rather than another. We would do the same thing, but we would first examine the external context from which the statements were derived. An example from Faircloughs work would be his comments on a BBC radio news report of September 1993 on the extradition of two Libyans accused of responsibility for the Lockerbie bombing in 1988. He points to the manner in which the UN Secretary-General is reported as taking a tough line with Libya and is said to be using the word demand:
Part of the framing here is the choice of demand as the reporting verb * it is highly improbable that the Secretary-General said I demand that . . . , so demand rather than, for example, ask would seem to be a framing conducive to an interpretation which casts the Libyans in an unfavourable light. (Fairclough, 2003, p. 53)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

The question we would put is, did he say it or didnt he? The research would be stronger without the guess about what the Secretary-General probably said. But to find out requires going beyond the immediate text, to examine the original speech and to assess the significance of what it actually included.

186

GREG PHILO

The second issue is, what gives power to critical analysis for our own readers. By this I mean, what makes people sit up and take notice? It might not come as much of a surprise to know that politicians exaggerate, that they use rhetoric, or that they speak of things they want to happen as if they are already happening. A critique is given more force by conveying a sense of what is excluded * of what we are not being told. This is especially so when it is possible to show that the excluded accounts have a strong evidential base. In our work when we indicated that trade unions were blamed on television news for industrial failures, we were able to give more credence to alternative accounts by quoting from figures produced by management in the car industry. These revealed the role of their own mistakes in lost production which had not been featured in television news. We also quoted from other authentic sources beyond the immediate news texts. We included this report from the Financial Times which very clearly stated the trade union view:
Shop stewards tell hair-raising stories about managerial failings, and point at the moment to constant assembly-track hold-ups caused by non-availability of supplier component parts. (6 January 1975)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

Critical discourse analysis would be more powerful if it routinely included a developed account of alternatives. The final point is what I have called whose rhetoric and also relates to the need for an account of the social and political structures which underpin the content of texts. Fairclough does sometimes go beyond the immediate text in order, for example, to contrast the claims of neo-liberalism with its actual effects. In this, he shows a concern with the gap between rhetoric and reality and between what people say and what they do. In his New Labour, New Language? , he also contrasts Blairs relaxed and inclusive style against evidence of control freakery (2000, p. 156). But much of Faircloughs work is textbased in the sense that it focuses on texts to show how dominant perspectives are legitimised through various strategies in the use of language. There is another dimension to the issue of the difference between what politicians say and what they do which can create problems for such textual analysis. The point is that politicians sometimes speak in favour of policies in which they do not actually believe and which they have no interest in pursuing. The rhetorical strategies employed would not therefore be seriously intended to win support for the policy but may exist simply to gain personal support for the speaker. For example, in September 2002, Tony Blair made a speech about the need to continue to redistribute wealth:
It must be a Britain in which we continue to redistribute power, wealth and opportunity to the many, not the few. (Tony Blair, 17 September 2002)

It was an unusual statement since the governments own figures on the distribution of wealth show that Britain has one of the most extreme divisions between rich and poor in Europe. Blair and New Labour have rejected any suggestion of increasing income or property taxes to dent the wealth of the super rich. On the TV news, the speech was attributed to the need for Blair to gather support in his own party for the coming war with Iraq. As this journalist comments:
He [Tony Blair] has spent the past few weeks preaching a message about war and backing America that makes many Labour members nervous and some positively sick at the soul. So ahead of the party conference [Blair is making] promises more to their taste.

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE He even used an old socialist word, one banned from the New Labour lexicon. (BBC1, early evening news, 17 September 2002)

187

So Blairs use of the old socialist word redistribution might be seen as rhetoric designed simply to increase his support in the Labour Party. This implies the existence of different types of rhetoric. One which we have seen earlier is designed to legitimate and develop a new economic order for the benefit of the powerful, and the other which might be called mere rhetoric designed for more limited personal or political advantage. There are gaps between rhetoric and reality in both types * between for example the claims made for the new order and what it actually delivers. But this is a different order of gap from when a politician is merely stealing the clothes of another group and is not seeking to legitimise fundamental changes. It also raises the question of whether there is much point in analysing the textual strategies of a political speech in such a case as if it represents a fundamental commitment to an ideological programme, when by other (external) criteria we would know that it does not. In considering types of rhetoric, another possibility is of course that Tony Blair and New Labour do actually believe some of the progressive comments which they make, for example on the need to abolish child poverty or debt in the developing world. So to distinguish these from mere rhetoric and the legitimising of economic power in neo-liberal speech requires a detailed exposition of current political debate, the perspectives within it and their relation to policy and real change. A discourse analysis which focuses on political rhetoric as legitimising a new economic order and analyses texts on this basis would miss such differences in political meaning and potential impact. Crucially, if the analysis remains within the text it is not possible to explain the social relationships which underpin the presentation of the descriptions and accounts which appear. If we look, for example at an analysis by Fairclough of another speech by Blair, we can see this more clearly. This speech was about the war on terror and globalisation. Fairclough shows how the structure of the speech dramatically constructs a dialogue with imagined opponents (as in: people say we should do this, we have tried it, it didnt work, so now we must do our policy). The point, which Faircloughs discourse analysis research reveals, is that in these parts of the speech Blair constructs the supposed opposition in such a way as to affirm the correctness of his own decisions. But in another part of the speech which Fairclough does not discuss, Blair actually endorses the views of his opponents. He states:
The demonstrators are right to say there is injustice, poverty, environmental degradation. (Tony Blair, 13 October 2001, in Fairclough, 2003, pp. 47, 238)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

The problem is that when Blair does this, his opinions do not sound much different from those of Faircloughs. So if we stay within the text, it is not immediately clear what critical discourse analysis is being critical of. At such a point, it would be helpful to consider the political relationships which led to the use of such rhetoric and also to have a detailed account of the potential gap between it and the reality of what New Labour is actually doing on, for example, the environment. The key point is that to distinguish between types of rhetoric necessitates an analysis of political structures, purposes and strategies. It requires an account of the social and political system and conflicting interest within it, beyond what can be seen from an immediate text. Without this we cannot comment on the difference between rhetoric and reality in terms of the intentions of the speaker, the validity of representations and the relation between accounts that are featured and alternative versions of truth. Textual

188

GREG PHILO

analysis should extend its methods to include the study of the structures and relationships which shape the content of texts. We will see this again in considering the work of Teun van Dijk.

Van Dijk and the Ideological Square

As we have seen, van Dijk is also concerned with questions of ideology and particularly with the issue of the reproduction of racism in discourse. He notes that traditional studies of ideology and language have focused on lexical items, i.e. individual words which may imply a value judgement (e.g. terrorist or racist). He suggests that a discourse analytical approach should go beyond this. He notes that opinions and the assumptions which they contain may be expressed in many complex ways in text and talk, in for example:
Headlines, story structures, arguments, graphical arrangements, syntactic structures, semantic structures of coherence, overall topics and so on. (van Dijk, 1998, p. 31)
Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

Some of his work parallels our analysis of texts. He highlights for example the importance of agency, responsibility and blame for actions and the perspective from which events are described and evaluated (1998, p. 44). Thus in Bad News from Israel , we showed how the Palestinians were often represented on the news as initiating conflict in the sense that they started a problem by attacking the Israelis who then retaliated or responded * as in this example:
Five Palestinians have been killed when the Israeli army launched new attacks on the Gaza strip in retaliation for recent acts of terrorism . (BBC Radio 4, 7.30 a.m., 6 March 2002, our italics)

So the agency and responsibility for initiating the violence is presented as being with the Palestinians. We showed in our reception studies how audience members could infer blame from such accounts. As one young woman commented in a focus group:
You always think of the Palestinians as being really aggressive because of the stories you hear on the news. I always put the blame on them in my own head . . . I always think the Israelis are fighting back against the bombings that have been done to them. (In Philo and Berry, 2004, p. 222)

The Palestinians do not see the conflict in this way. From their perspective, its origins lie in the loss of their homes and land when Israel was established, which made them refugees and in their situation as living under Israeli military occupation since 1967. A senior journalist from the BBC commented to us on the absence of this perspective on the news. What was missing, he said, was the view that this was an uneven war and that:
It is a war of national liberation * a periodic guerrilla war, sometimes using violent means, in which a population is trying to throw off an occupying force. (Paul Adams, interview, in Philo and Berry, 2004, p. 260)

We also noted how the Israeli perspective, which as we have seen focused on the war on terror was sometimes endorsed by journalists in their commentaries * as in this description of an Israeli action:

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE The [Palestinian] attack only reinforced Israeli determination to drive further into the towns and camps where Palestinians live * ripping up roads around Bethlehem as part of the ongoing fight against terror . (ITV, early evening news, 8 March 2002, our italics, cited in Philo and Berry, 2004, pp. 187 8)

189

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

Another dimension of method which van Dijk notes is the value of quantitative demonstrations. These are important to establish whether overall strategies exist in establishing differential descriptions of social groups. In our study we used quantitative analysis to show the relative prominence of casual linkages explaining action and motive. For example, we showed in a major sample of news content that Israelis were said to be responding or retaliating to something that had been done to them about six times as often as the Palestinians (Philo and Berry, 2004, p. 160). In his own work, van Dijk develops the concept of the ideological square which is intended to highlight key functional moves in developing an ideological strategy. For van Dijk, the heart of this strategy is a polarisation between us and them, which involves positive in-group description and negative out-group description (1998, p. 33). So an ideological account would:
1. 2. 3. 4. Emphasise our good properties/actions Emphasise their bad properties/actions Mitigate our bad properties/actions Mitigate their good properties/actions. (1998, p. 33)

Using this theoretical model, van Dijk analyses media texts, notably of British tabloids such as The Sun . He concludes that various levels of discourse may be involved in the enactment, expression or inculcation of negative beliefs about immigrants or minorities (van Dijk, 2000, p. 42). We certainly found in our work that media coverage could stigmatise out-groups having the effect, as we wrote, of re-enforcing our identity and their exclusion (Philo and Beattie, 1999). We did not, however, use a concept such as the ideological square. The reason is that the four points of a square can easily become four boxes into which language is fitted (especially so in the hands of students who usually make it go in one way or another). My own view is that it is better to avoid such a priori categories when beginning the analysis of a text. Van Dijk is certainly right that there are many elements of media accounts which correspond to his schema. The difficultly is that some do not, so the question is raised, how can we develop a method which can explain contradictions and variations? We can pursue this by looking at the case of The Sun newspaper. This has a differentiated readership and a complex marketing strategy which produces variations in its news coverage, editorials and features. The following editorial is on a woman who has built a mud hut in her garden and would fit without difficulty into the ideological square:
Mrs. Desiree Ntolos 20-ton mud hut is being demolished. She built it in the back garden of her council house in Dagenham, Essex because she was homesick for Cameroon. Just a thought: why doesnt she build the next monstrosity in the African homeland shes pining for? The neighbours would no doubt club together to pay her airfare. (The Sun , 26 September 1992)

190

GREG PHILO

In van Dijks words, this would favour the in-group and derogate or problematize the out-group (2000, p. 42), with the neighbours being the in-group who would pay the airfare for the out-group African woman to go. But there are other cases of some news and editorial coverage which clearly contradict the model. How do we account for a commentary which appeared in The Sun criticising the abuse hurled by some supporters at black football players? It had the headline Racist Morons Ruin the Game (1 September 1990). In another editorial The Sun explicitly attacked the sacking of a black chambermaid:
Jennifer Millingtons job as a chambermaid lasted one day. She was sacked, according to the owners of [a hotel] in Newquay, Cornwall, for one reason. She was black and some of the guests objected to her. We hope this disgraceful story has one certain result. Any decent person going to Newquay will stay in any hotel but [this one]. (The Sun , 2 September 1991)

Here, the out-group are clearly the management and the guests but the story is described as disgraceful and there are no mitigating factors given for their behaviour. The Sun would defend itself against criticisms of being racist by pointing to such stories and also by noting its large number of black readers. This offers some potential insight into these apparent contradictions. The Sun has always been faced with the issue of selling to diverse readerships and with potential variations between its content and the views of those who buy it. It has a history of being a right-wing populist newspaper which sells to a predominantly working-class audience, many of whom vote Labour. When it supported Margaret Thatcher at the beginning of the 1980s, one of its targets was the left-wing head of the Greater London Council, Ken Livingstone. By the end of the 1980s, when the Thatcher government was in serious decline, The Sun actually gave a regular column to Ken Livingstone under the title A View from Labour. He then used it to attack other columnists in the same paper for being too right-wing. The Sun s approach to issues with racial overtones shows a similar concern to appeal to a diverse readership and specifically not to offend key minority groups. For example, in 2006 there was a major controversy over the publication by European newspapers of cartoons featuring the Prophet Muhammad. The Sun did not publish these. In an editorial it stated that:
The cartoons are intended to insult Muslims, and The Sun can see no justification for causing deliberate offence to our much-valued Muslim readers. (The Sun , 3 February 2006)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

We spoke to a Sun journalist who dealt with race issues. He expressed the view that his paper had deliberately avoided being in any way anti-Muslim in dealing with stories such as the Abu Hamza case, where a cleric was arrested and tried for inciting violence:
I think we handled the Abu Hamza story well * concentrated just on what he was, it didnt spill over into being anti-Muslim. Being cynical I could say its because a lot of Sun readers are from ethnic minorities. (Interview, 21 June 2006)

To explain the apparent contradictions in The Sun s coverage really requires a production study and an analysis of the conflicting pressures which affect content. Van Dijks method does not include this and when confronted by such variations, he can

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE

191

simply pass over them. Consider, for example, his commentary on newspaper coverage of the views of Enoch Powell, a right-wing politician who called for the repatriation of migrants:
Despite its formal rejection of Powells ideas, the conservative Press seldom misses the opportunity to publicise his racist views, so that millions of readers will know them. The Times even published another recent diatribe of Powell against migrants, thereby legitimating his racist views as part of the public debate, even when it distances itself from such views. A reassuring Sun poll shows that the majority of the British people do not support Powells astonishing blacks go home call. That white public opinion (or the methods by which it is assessed) is fickle, is shown a few weeks later, after the Brixton and Tottenham disturbances, when The Mail reports that most white Britons want to stop further immigration and favour repatriation. (van Dijk, 1991, p. 97)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

My own view is that when The Sun refers to Powells opinion as an astonishing blacks go home call and publicises a study showing that most people disagree with him, then this is a significant variation from the hypothesis that The Sun is explicitly or implicitly racist. It really needs to be explained. But in this case, van Dijk simply passes over it and moves on to comment on the fickle character of white public opinion, which is a different issue altogether. To explain the position of The Sun and other media requires a study of production and a method which includes analysing the practices of journalists as well as newspaper marketing strategies. I am also not sure about van Dijks claim that for the media to present views while criticising them has a legitimating effect. There is a good deal of media coverage of Osama bin Laden and his video tapes when they are released but this does not legitimise his actions or those associated with him in British public opinion.5 The impact of specific media messages really has to be assessed using audience studies. In his more recent work, van Dijk analyses The Sun s coverage of illegal migrants. He shows how in a report they are presented as invading Britain and he indicates the extremely negative quality of much of what is written. The text from The Sun is as follows:
Britain is being swamped by a tide of illegal immigrants so desperate for a job that they s and nightclubs. will work for a pittance in our restaurants, cafe Immigration officers are being overwhelmed by work. Last year 2191 illegals were nabbed and sent back home. But there are tens of thousands more, slaving behind bars, cleaning hotel rooms, and working in kitchens. (The Sun , cited in van Dijk, 2000, p. 44)

But once again he is confronted by the issue of variations in the coverage, where as he notes, there is an element of empathy creeping into the article when the journalists describe the immigrants as slaving at their work (2000, p. 45). This relates to the poor conditions and salaries which the immigrant workers receive. So there is an ambivalence in The Sun report between the implied violence and threat of an invasion and the potential sympathy which might be invoked for the slave workers. To explain this requires as before, an investigation which goes beyond the text. When we interviewed television journalists, we found some who were intensely critical of media attacks on asylum seekers and migrants. One very senior editor from ITN told us that he believed some sections of the press should be prosecuted for their role in the promotion of violence against these groups.6 He and others with whom we spoke had attempted to produce news items which went against the dominant flow. One focus of

192

GREG PHILO

these alternatives was the exploitation and poverty of migrant workers. We can see this in the following headline from ITN, which is from a special extended report in a news programme:
They came in search of a better life * what they find is squalour and slavery. (ITV, late news, 22 May 2006)

In the case of The Sun report, it might be that the newspaper is attempting to appease different audiences as I noted above, or that the journalists are trying to introduce alternative ideas. Van Dijks approach is first to acknowledge the contradiction in the report and then to attempt to resolve it by reinterpreting the text. He notes that when the journalists describe the immigrants as slaving: This totally converts (and subverts) the earlier characterisation of the immigrants as active and evil, and not as victims (2000, p. 45). He then attempts to find potential meanings which could put the text back into his ideological square. He goes on to say:
On the other hand, the use of slave presupposes slave holders, and instead of mere empathy, this may suggest an accusation of restaurant owners who exploit their illegal workers. (2000, p. 46)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

I do not see how a connotation of slave holders could reduce sympathy for the slaves in this context. The difficulty is that because van Dijks methods remain text-based, there is no way in which it is possible to explain such contradictions other than by speculating on further potential meanings which can be derived from the text. Without the analysis of production and reception processes, discourse analysis is limited in the conclusions that it can draw. There is a need to develop methods which can trace the communication of messages from their inception in contested perspectives, through the structures by which they are supplied to and processed by the media, then to their eventual appearance as text and finally to their reception by audiences. There follows a brief example from our own work to show how this might be done.

Production, Content and Reception of a Message

This is a practical example based on news coverage of the Israeli Palestinian conflict and specifically on reports about the shooting early in the Intifada of a young Palestinian boy, Mohammed al-Durrah. The images of him and his father crouched against a wall were widely shown and became a potent symbol of the Palestinian intifada. The circumstances of this killing were highly contested and became the focus of an extensive propaganda struggle. We have seen earlier in this chapter how Israeli public relations focused on the war on terror and sought to present Israel as threatened and essentially responding to attacks. This provided an overall framework, but each new event in a conflict requires a specific public relations response. In the case of Mohammed al-Durrah, the Israelis issued a statement saying that the boys death was unintentional. This was reported on TV news as follows:
Israel says the boy was caught unintentionally in crossfire . (ITV, lunch-time news, 2 October 2000, our italics)

The Palestinians rejected this account and stated that the targeting was deliberate. This view appears on the news in an interview from hospital with the boys father, who is reported as follows:

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE Miraculously his father survived but his body is punctured with eight bullet holes. They shot at us until they hit us, he told me, and I saw the man who did it * the Israeli soldier. (BBC1, main news, 1 October 2000)

193

The two accounts of the events are therefore opposed, but it is the Israeli view that became dominant on the news. Most significantly, it is endorsed by journalists as the normal account of events. It is referenced not simply as a viewpoint in the sense that the Israelis say that he was caught in crossfire, but rather as a direct statement, as in the boy was caught in the crossfire. There are a series of examples of this:
Newscaster: Palestinians have been mourning the death of a 12-year-old boy killed in the crossfire . Journalist: The Palestinian death toll is rising steadily, among them a 12-year-old boy, Mohammed al-Durrah, who with his father got caught in the crossfire . (ITV, early evening news, 1 October 2000, our italics) Journalist: Nearby I met the mother of 12-year-old Mohammed al-Durrah, the Palestinian boy killed on Saturday in the middle of a ferocious gun battle . (BBC1, main news, 3 October 2000, our italics) Journalist: The worst clashes have been in Jerusalem, the Gaza Strip, and the West Bank, where a 12-year-old boy was killed in the crossfire . (ITV, main news, 1 October 2000, our italics)

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

It is clear that the journalists are sympathetic and do say that it was the Israelis who killed the boy, but it is the Israeli explanation of this event which is most frequently referenced (for a full account see Philo and Berry, 2004, pp. 148 50, 225 31). In our audience studies, we were then able to show how news accounts of these events had a measurable influence on the understanding and memory of them amongst viewers. For this work we brought together members of the public in focus groups. We also invited journalists to attend these and they took an active part in the research, asking questions about specific responses to coverage in which they had been involved. One dimension of this work was a news writing exercise, in which we asked the audience group members to imagine that they were journalists and to write their own small news items. They were given 16 photographs which were taken from TV news footage of the conflict and asked to use these as a stimulus. They were not constrained to focus on these pictures but in practice could write anything they wished. As a method this was designed to show what audiences have retained from news programmes. We found that many participants had a remarkable ability to reproduce both the content and structure of news bulletins. One of the pictures in this exercise showed Mohammed al-Durrah with his father, others included the aftermath of a suicide bombing and the body of an Israeli solider being thrown from a window after he had been captured and later killed. There was no prior discussion of these and no attention was drawn to these or any of the pictures.These events were, however, referred to in some of the news stories written by the audience group members. In the case of Mohammed al-Durrah, there was no reference made to the Palestinian view that he had been deliberately killed. Some group members did, however, reproduce the language of the original Israeli statement:
A young boy was caught in the crossfire as Israeli troops opened fire in the West Bank. (Middle-class female, London, our italics)

194

GREG PHILO Israeli soldiers return fire and a father and son are caught in the crossfire * the boy is fatally wounded. (Middle-class male, Glasgow, our italics) The American flag has been publicly burned by the Palestinians following the death of a young child who had been cornered in the crossfire between the Jewish soldiers and Palestinians in Jerusalem. (Middle-class female, Paisley, our italics)

There was another very significant feature to the stories written by some members of the audience groups. They also reproduced the structure and sequence of accounts as they had most frequently occurred on TV news. This, as I noted above, included the presentation of Palestinian action as initiating a violent event, while the Israelis were then shown as responding. In the case of Mohammed al-Durrah, this is an unlikely scenario since his death was at the very beginning of the intifada. But in order to retain what became the normal sequence of action, some group members took events which had occurred after Mohammed al-Durrahs death and wrote about them as producing the Israeli response in which the child was killed. This historical reversal occurs as in the following examples:
Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

An Israeli soldier was taken hostage and thrown to his death by Palestinians on the rampage. The scene was witnessed live on TV by a shocked nation who took to the streets to protest . . . the Israeli people vowed to revenge this act and in the fighting that followed a 10-year-old Palestinian boy was shot dead in his fathers arms. (Female teacher, Paisley) A young boy was killed as his father helplessly tried to shield him from Israeli bullets. The Israeli onslaught came as a direct retaliation to a newly-wed Israeli couple being killed by a Palestinian suicide bomber in the latest Palestinian terrorist attack. (Middle-class female, London)

In this approach we can begin to show how TV news can shape not only the language which we use in ascribing meaning to events, but also the way in which we group and organise our memories. We developed our methods with the intention of investigating mass communications as a totality in which meanings were circulated through the key dimensions of production, content and reception. The concept of circulation is crucial since it allows for the possibility of interaction between elements of the process and does not imply a single one-way flow from the top to the bottom of the system, i.e. from production to reception. Those who supply information to the media certainly intend it to have an impact, but they are still aware of the contexts within which their messages will be received, So what is supplied will itself be shaped by an anticipation of the reception process as well as by an understanding of the likely response of different elements of the media. We have shown how messages can have powerful influences on audience beliefs and understanding but our work also demonstrates how some audience members critique and reject what they see and hear in media accounts. The growth of new technology and the Internet has also to some extent increased the potential of ordinary citizens to develop their own systems of communication and has added to the interactive possibilities of traditional media. In terms of methodology, the key point is that all these elements must be understood and studied as part of a total system * rather than in isolation as with studies which remain focused on texts. Many theorist in the sociology of media now accept the importance of such an approach. To analyse processes of content, production,

DISCOURSE ANALYSIS AND JOURNALISTIC PRACTICE

195

reception and circulation simultaneously is a complex task but it is the way forward if we are to come to terms adequately with the generation and reproduction of social meanings in media accounts.

NOTES
1. Another reason for using their work is that they stand in a tradition of critical analysis, in which they are concerned to study the consequences of the use of language in its ideological forms * in other words its effect in developing or reproducing social relations of power and exploitation. We share this concern with consequences in the real world of social relations and structures of power. Some other approaches in discourse theory avoid such issues by bracketing off what is real or true. Reality is seen as a product of discourse and of representation and is therefore indenitely negotiable. This produces an inability to comment on the relationships of our society and is a form of intellectual quiescence. For a more extended critique of this, see Philo and Miller (2001). There are also theorists from within discourse analysis who have suggested, as I do, that a focus on linguistics leads to a lack of proper attention to processes of production and consumption. John E. Richardson, for example, has argued that critical discourse analysis should be focused at three levels: on texts; on the discursive practices of production and consumption; and on the wider socio-cultural practices which discourse (re)produces (2006, p. 1). The principal reason for this was that it seemed to us that there were an indenite number of ways in which characteristics of language in use could be described. Rather than attempt to label all these, we focused on the specic textual features which were the heaviest carriers of meaning * which we could establish through a combination of textual analysis and our work with audience groups. We then analysed how these elements of the text worked to establish explanations * for example through descriptions of sequences of action which implied cause or responsibility and reected on the legitimacy of different parties (see e.g. Philo and Berry, 2004, pp. 160 5). For a fuller discussion of this and the impact of the release of the free market on broadcasting in Britain, see Philo (1995). Where hyponymy implies inclusion in a group of words, antonymy would suggest exclusion from the group. Social cohesion and organic community would be hyponyms, with antonyms as polarisation, or fragmentation. See, for example, forthcoming research by Sarah Oates and Mike Berry of Glasgow University on public attitudes to terrorism (ESRC New Security Challenges Programme) * The Framing of Terrorist Threat in British Elections, ESRC RES 228- 25-0048. The interviews were conducted in 2001 2 as part of our study of news coverage of the Israeli Palestinian conict. Since the issues of migrants and asylum seekers were very prominent at the time, several journalists made additional comments on coverage of these.

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

2.

3. 4.

5.

6.

REFERENCES
ALTHUSSER, LOUIS BOURDIEU, PIERRE

(1969) For Marx , London: Penguin. and WACQUANT, LOIC (2001) New-liberal Speak: notes on the new planetary vulgate, Radical Philosophy 105, pp. 2 5.

196

GREG PHILO (2000) New Labour, New Language? , London: Routledge. (2001) Language and Power , London: Pearson/Longman. FAIRCLOUGH, NORMAN (2003) Analyzing Discourse , London: Routledge. FISKE, JOHN (1987) Television Culture , London: Methuen. FOUCAULT, MICHEL (1994) What Is Enlightenment?, in: P. Rabinow (Ed.), Michel Foucault, Essential Works , Vol. 1, Ethics , Harmondsworth: Penguin. GLASGOW UNIVERSITY MEDIA GROUP (1980) More Bad News , London: Routledge. GLASGOW UNIVERSITY MEDIA GROUP (1982) Really Bad News , London: Writers and Readers. GLASGOW UNIVERSITY MEDIA GROUP (1985) War and Peace News , London: Writers and Readers. HALLIN, DANIEL (1989) The Uncensored War: the media and Vietnam , Berkeley: University of California Press. MILLER, DAVID (1994) Dont Mention the War: Northern Ireland, propaganda and the media , London: Pluto. PHILO, GREG (1995) Television, Politics and the Rise of the New Right, in: Greg Philo (Ed.), Glasgow Media Group Reader: industry, economy, war and politics , Vol. 2, London: Routledge. PHILO, GREG (1996) Media and Mental Distress , London: Longman. PHILO, GREG (Ed.) (1999) Message Received, Glasgow Media Group Research 1993 1998 , London: Longman. PHILO, GREG and BEATTIE, LISA (1999) Migration in Media, in: Greg Philo (Ed.), Message Received, Glasgow Media Group Research 1993 1998 , London: Longman. PHILO, GREG, BEHARRELL, PETER and HEWITT, JOHN (1977) One Dimensional News * television and the control of explanation, in: Peter Beharrell and Greg Philo (Eds), Trade Unions and the Media , London: Macmillan. PHILO, GREG and BERRY, MIKE (2004) Bad News from Israel , London: Pluto. PHILO, GREG, HEWITT, JOHN and BEHARRELL, PETER (1982) Industrial News, in: Really Bad News , London: Writers and Readers. PHILO, GREG, HEWITT, JOHN and BEHARRELL, PETER (1995) And Now Theyre Out Again: industrial news, in: Greg Philo (Ed.), Glasgow Media Group Reader: industry, economy, war and politics , Vol. 2, London: Routledge. PHILO, GREG and MILLER, DAVID (2001) Cultural Compliance, in: Greg Philo and David Miller (Eds), Market Killing , London: Pearson/Longman. RICHARDSON, JOHN E. (2006) On Delineating Reasonable and Unreasonable Criticisms of Muslims, Fifth-Estate-Online , August. SCHLESINGER, PHILIP (1978) Putting Reality Together , London: Constable. VAN DIJK, TEUN (1991) Racism and the Press , London: Routledge. VAN DIJK, TEUN (1998) Opinions and Ideologies in the Press, in: A. Bell and P. Garrett (Eds), Approaches to Media Discourse , Oxford: Blackwell. VAN DIJK, TEUN (2000) New(s) Racism: a discourse analytical approach, in: S. Cottle (Ed.), Ethnic Minorities and the Media: changing cultural boundaries , Buckingham: Open University Press.
FAIRCLOUGH NORMAN FAIRCLOUGH NORMAN

Downloaded By: [PERI Pakistan] At: 09:21 21 April 2010

Greg Philo, Glasgow University Media Group, University of Glasgow, Adam Smith Building, 40 Bute Gardens, Hillhead, Glasgow G12 8RT, UK. http://www.gla.ac.uk/centres/ media group