Radical Mass Media Criticism

On delineating 'reasonable' and 'unreasonable' criticisms of Muslims

John E. Richardson, Loughborough University, UK.

Abstract: This paper is concerned with anti-Muslim racism and, specifically, with distinguishing between legitimate criticisms of (people who happen to be) Muslims and illegitimate (in other words prejudiced or racist) attacks. Critical Discourse Analysis (CDA) is introduced and discussed as a productive way to make such a distinction. CDA views discourse on and about Muslims as social practice, and discourse analysis as the analysis of what people do with text and talk. I argue that in order to examine the ways that discourse may (re)produce relationships of prejudice, dominance and/or discrimination, analysis needs to 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. From such a perspective, the dominant modes of representing British Muslims are both a reflection of and a contributing factor in the continued racist discrimination suffered by Muslims both in the UK and internationally.

Keywords: Critical Discourse Analysis, argumentation, Islamophobia, anti-Muslimism, racism, readers' letters, Runnymede Trust, essentialisation, stratification, reasonableness.

Introduction

The validity and utility of 'Islamophobia' to analyse and account for the discrimination suffered by Muslims is a matter of some debate. Detractors, from both the political left and right, have attacked the concept arguing that its use panders to what they call political correctness and that it stifles 'legitimate' criticism of Islam and Muslims. An archetypal example of such argumentation, headlined 'In defence of Islamophobia' and written by Polly Toynbee, was printed a day after the public launch of the Runnymede Trust's report on British Islamophobia. The opening paragraph read:

'I am an Islamophobe. I judge Islam not by its words - the teachings of the Koran as interpreted by those Thought-for-the-Day moderate Islamic theologians. I judge Islam by the religion's deeds in the societies where it dominates. Does that make me a racist?' (Toynbee, 1997).

The column adopted liberal argumentation - a rights-based discourse with a clear antipathy to religious expression in the public sphere - and applied the stock subjects used by 'liberals' when arguing against Islam: women and inequality (specifically the charge that Islam is 'a religion that describes women as of inferior status, placing them one step behind in the divine order of things'); Rushdie and the fatwa; free speech and the idea that Muslims cannot abide 'criticism or mockery'; an alleged lack of Muslims denouncing atrocities in 'Islamic states'; the injustices of Saudi 'sharia courts'; and a presupposed opposition of religion and rationality. By the terms of the Runnymede Trust's eight binary argumentative positions characterising 'closed' and 'open' views of Islam, the column was highly Islamophobic - Islam and Muslims were represented as separate; inferior; the enemy; manipulative; criticisms of 'the West' by Muslim sources were not considered; 'Islam' and 'the West' were regarded as incompatible and in conflict; and Islamophobia was, as the title suggests, defended.

However, Toynbee (1997) dodges the charge that she is being racist or even explicitly anti-Muslim by including criticisms of other religions in the column: she proudly states she is 'also a Christophobe' because 'Christianity remains a lethal weapon in Northern Ireland' and that if she 'lived in Israel, [she'd] feel the same way about Judaism'. The inclusion of sideswipes against Christianity and Judaism makes categorising her argument as 'Islamophobic' a little more complicated. Should this, or other texts criticising organised religion or the wrongdoing of (people who happen to be) Muslims, be classified 'Islamophobic'? This paper offers one response to this problem. I will suggest that the Runnymede Trust's emphasis on textual representation, to the detriment of an analysis of the interaction between discourse and discrimination, is unhelpful in distinguishing between prejudicial or bigoted attitudes towards Islam and not only the undifferentiated - and therefore not purely 'anti-Islamic' - criticisms of atheists but also more general criticisms of the wrong-doing of (people who happen to be) Muslims.

The discussion of sampled texts - drawing mainly on the letters' pages of broadsheet newspapers - is informed by a discourse analytic methodology, which views discourse on and about Muslims as social practice, and discourse analysis as the analysis of what people do with text and talk. Letters to the editor were selected over alternative genres because letter pages serve as forums for opinion, dialogue and debate; they allow the readership of a paper 'to express their opinions, their fears, their hopes - and, just as important, air their grievances' (Jackson, 1971: 152). Letters pages aim to include readers - both symbolically and literally - who, should their letters be selected for publication, 'feel a sense of importance and special communication' (Wober, 2004: 50). They also play a significant role in communicating the identity of a newspaper given that they attempt to recreate the preoccupations and discourse of their readership, as expressed in their daily postbag. As such, the letters page is a very popular feature of most newspapers. Ian Mayes (2001), the readers' editor for the Guardian, describes the letters page as 'among the more important parts of the paper' and 'the paper's principal forum of reader opinion'. On any normal day, the Guardian receives around three hundred letters competing for a place in this forum. In times of political controversy, this can increase considerably - for example, it doubled during the first week of the 2003 invasion of Iraq (Mayes, 2003). However, certain opinions are included more frequently than others. According to Wober (2004: 52-3), 36.9 per cent of letters included in The Times are written by members of the social elite - in other words, MPs, Professors, Officers of Associations (e.g. the Christian Socialists Movement, Dorset Natural History and Archaeological Society, etc.) and 'people in organizations which speak for and try to keep order'. The inclusion of prejudice, and 'everyday racism' in such letters therefore stands as an indication of the extent to which racist views have 'become part of what is seen as 'normal' by the dominant group' (Essed 1991: 288).

Islamophobia or anti-Muslimism?

As stated, the Trust's use of the term 'Islamophobia' has not escaped criticism, not least for its seeming ability to conflate prejudicial, insensitive or bigoted representations of Islam with both the undifferentiated - and therefore not purely 'anti-Islamic' - criticisms of atheists, or the more general criticisms of the wrong-doing of (people who happen to be) Muslims. In one of the more sympathetic criticisms, Halliday (1999) offers a critique of selected writings on the concept of 'Islamophobia', which he feels do not sufficiently account for the derogation and prejudice experienced by Muslims. At the outset, it should be stated that Halliday agrees with the Runnymede Trust in as much as 'there is such a thing as denoted by the term "Islamophobia" […] Recent examples in the British press are not hard to find' (Halliday, 1999: 898). Thus, Halliday's objection is centred on the use of the neologism 'Islamophobia' when describing and/or accounting for the (racist) exclusion of Muslims, feeling that the term 'anti-Muslimism' should be used in preference (also see Halliday 1996). Four criticisms of the concept are offered: first, that 'Islamophobia' is somewhat of a misnomer, and that it:

'misses the point about what it is that is being attacked: "Islam" as a religion was the enemy in the past: in the crusades or the reconquista. It is not the enemy now […] The attack now is not against Islam as a faith but against Muslims as a people, the latter grouping together all, especially immigrants, who might be covered by the term' (Halliday, 1999: 898).

This point is echoed by Reisigl and Wodak's (2001: 6) more general criticism of 'phobias': that they 'neglect the active and aggressive part of discrimination' by focusing on racism as a collection of (pathological, pseudo-logical or illogical) beliefs. In contrast to the thrust of the 'Islamophobia' concept, Halliday (1999) argues that the stereotypical enemy 'is not a faith or a culture, but a people' (ibid: 898) and therefore its use is misleading, shifting analysis away from the 'real' targets of prejudice: Muslims.

Second, Halliday argues that the concept 'Islamophobia' is undesirable on the grounds that it reproduces a particularly unhelpful distortion - 'that there is one Islam: that there is something out there against which the phobia can be directed' (Ibid.). While 'Islam, like all cultural systems, is a contested field of meaning' (Beinin & Stork, 1997: 21), Halliday argues the term 'Islamophobia' suggests otherwise, invoking a unified, singular and perhaps essentialised 'thing' at which such prejudice can be directed. This section of the essay is particularly strong, drawing heavily on Halliday's cogent critique of the constructed nature of the 'clash' between 'Islam' and 'the West'. The emergent, fluid and contestable nature of both 'Islam' and 'the West' are discussed (albeit with the stress upon the former) and, echoing the claims of Rodinson (1979), Said (1997), al-Azmeh (1993) and others, Halliday suggests that 'what is presented as "Islam" may well be one, but by no means the only possible interpretation' (1999: 897). What this focus does of course, is not only expose the discursive potential of 'Islam', but also the politically motivated actions of those who claim access to the one true Islam and therefore what is at stake in their maintaining such privilege.

Third, the combination of these two previous inadequacies with the concept 'Islamophobia' leads to 'confusing' practical results. In particular Halliday claims, 'issues of immigration, housing, employment, racial prejudice, anti-immigrant violence are not specifically religious' in either the UK or 'the West' as a whole (Halliday, 1999: 899). Here, my views differ from Halliday's. A great many British Muslims - predominantly women - report heightened discrimination and abuse when they appear 'conspicuously Muslim' than when they do not. Further, the increase in personal abuse and street racism since September 11th 2001 in which the perceived 'Muslim-ness' of the victims was the central reason for abuse, regardless of veracity of this presumption (resulting in Sikhs and others being attacked for 'looking like bin Laden'), suggests that racial and religious bigotry and discrimination are more interlinked than Halliday suggests.

Fourth, in the criticism most pertinent to the current discussion, Halliday states that the use of the term 'Islamophobia':

'challenges the possibility of dialogue based on universal principles […since it] inevitably runs the risk of denying the right, or possibility, of criticisms of the practices of those with whom one is having the dialogue. Not only those who, on universal human rights grounds, object to elements in Islamic or other traditions and current rhetoric, but also those who challenge conservative readings from within, can more easily be classed as Islamophobes' (Ibid.).

Halliday argues that this effect of 'Islamophobia' (curtailing dialogue on the basis that the 'invocation of universal principles violates tradition') is felt at both the national and international levels. However it is abuse occurring 'within Muslim societies themselves' (original emphasis) which receives the most attention, where 'horrendous violations of human rights […] being committed, against Muslims, in the name of religion' are shielded from criticism (Halliday, 1999: 900). Few people would argue that human rights abuse or the inhumane or criminal activities of (people who happen to be) Muslims should be 'off limits' to criticism; and, for the record, I would disagree with anyone who suggested that they should be off limits. However, when does warranted criticism such as this slip into prejudiced, derogatory, and anti-Muslim discourse? To a certain extent this fourth problem with 'Islamophobia' arises out of the Runnymede Trust's own definition of the concept, wherein they suggest that Islamophobia is characterised by an 'unfounded hostility towards Islam' (1997: 4). This clearly entails an interpretative problem: in short, how do we establish that such hostility is unfounded, as opposed to justified? How do we differentiate between the defensible, the benign and the prejudicial (aside from simply adapting Justice Potter Stewart's now infamous phrase: I know prejudice when I see it)? While I do not claim to have a definitive or incontrovertible resolution to this question, I believe the answer must lie in the relationships between text and context.

Critical Discourse Analysis

The relationships between text and context form a principal focus of Critical Discourse Analysis. Discourse analysts emphasise, first, that discourse should be studied as language in use - we are interested in 'what and how language communicates when it is used purposefully in particular instances and contexts' (Cameron, 2001: 13; emphases added). We assume that language is used to mean something and to do something and that this 'meaning' and 'doing' are linked to the context of its use - that is, the immediate context of rhetor-text-audience and also the wider socio-political, cultural and historic contexts which bound the communicative act. Second, discourse analysts assume that language exists in a dialogue with society: that 'language simultaneously reflects reality ('the way things are') and constructs (construes) it to be a certain way' (Gee, 1999: 82). Thus, language represents and contributes to the production and reproduction (which discourse analysts usually label 'the (re)production') of social reality.

Critical Discourse Analysis (henceforth, CDA) represents a growing body of work within this general approach to language use, 'whose overall aim has been to link linguistic analysis to social analysis' (Woods & Kroger, 2000: 206). Critical discourse analysts argue that if we accept the second general principle of discourse analysis - that language use contributes the (re)production of social life - then, logically, discourse must play a part in producing and reproducing social inequalities. In response, 'CDA sees itself as politically involved research with an emancipatory requirement: it seeks to have an effect on social practice and social relationships' (Titscher et al, 2000: 147), particularly relationships of disempowerment, dominance, prejudice and/or discrimination. To examine how discourse may (re)produce such iniquitous relations, analysis needs to 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. Taking each in turn: first, the analysis of texts involves looking at the form, content and function of the text and usually covers:

'traditional forms of linguistic analysis - analysis of vocabulary and semantics, the grammar of sentences and smaller units, and the sound system ("phonology") and writing system. But it also includes analysis of textual organisation above the sentence, including the ways in which sentences are connected together ("cohesion") and things like the organisation of turn-taking in interviews or the overall structure of a newspaper article' (Fairclough, 1995: 57).

Second, one needs to consider the discursive practices of the communicative event, which usually involves an examination of:

'various aspects of the processes of text production and text consumption. Some of these have a more institutional character [e.g. the editorial procedures of the Independent compared to The Times] whereas others are discourse processes in a narrower sense [the "decoding" of texts by the reader/viewer]' (Fairclough, 1995: 58).

This level of analysis is a particular strength of good CDA, elaborating on the observations of texts offered by content analysis. While content analysis can offer a useful first stage of analysis, enabling the identification of broad patterns of representation in which to place the analysis of specific texts, too often content analysis - quantitative or otherwise - has skated over 'complex and varied processes of meaning-making within texts' (Deacon et al, 2000: 117). Content analysis presupposes that 'the meaning which can be recovered from particular content corresponds to the meanings that the speakers or writers intended in their texts and to those that the receivers read or hear' (Titscher et al, 2000: 10). More critically, Cottle (2000: 12) argues that such an approach to representation assumes 'that meanings are "contained" […] are not dependent on (differentiated) audience interpretations […or] are assumed to be confined to, embodied within, and "read off", depicted characters'. CDA attempts to resolve these problems of relying solely on content analysis in two ways: first, by acknowledging contestation and conflict of meaning within texts themselves. Van Dijk's (1988) work on textual macrostructures, for example, illustrates how, in understanding news reporting, audience members necessarily delete dissenting or counter-factual information, or else generalise it out of the picture. The point of (critical) discourse analysis is to examine how this is managed within texts. And second, CDA assumes (and argues) that textual meaning cannot be divorced from the context of social and discursive practice.

Finally, and developing the point made above, Fairclough (1995: 57) suggests that a fully rounded critical discourse analysis should involve an analysis of the text's 'sociocultural practice', or 'the social and cultural goings-on which the communicative event is part of'. This level of analysis, 'may be at different levels of abstraction from the particular event: it may involve its more immediate situational context, the wider context of institutional practices the event is embedded within, or the yet wider frame of the society and the culture' (Fairclough, 1995: 62). Titscher et al (2000) suggest that when tackling this level of analysis, '[q]uestions of power are of central interest; power and ideologies may have an effect on each of the contextual levels' (ibid: 151).

The methods of CDA are much more, diffuse than its shared principles suggest (see Phillips & Jørgensen, 2002; Wodak & Meyer, 2003). The approach I choose to adopt (Richardson, 2001, 2004; Richardson & Franklin, 2003) emphasises the argumentation of texts. Van Eemeren et al (1996) define argumentation as:

'a verbal and social activity of reason aiming at increasing (or decreasing) the acceptability of a controversial standpoint for the listener or reader, by putting forward a constellation of propositions intended to justify (or refute) the standpoint before a rational judge' (ibid: 5).

Beyond this formal approach however 'it is also possible to conceive of argumentation from a broader perspective and to understand it as a process that aims at exerting an influence on one's opinion, attitude, even behavior. It is however important to insist on the fact that the means are discursive' (Grize, 1990: 41, cited in Amossy, 2003: 13). The remainder of this paper examines the argumentative discourse of Islamophobia in relation to the three levels of analysis constituting CDA.

Text, argument, interaction

Looking at 'text' first: as suggested above, the Runnymede Trust's 'closed views' of Islam are focused predominantly on textual representation. Clearly each of the 'closed views' are characterised by basic racist assumptions: first, they are founded on a belief in 'communities of descent', evidenced in the manner in which the actions Muslims are thought to be, and represented as, defined by their (almost) invariable, inflexible 'Muslim-ness' as opposed to a wide range of other explanatory factors. This I regard as an essentialisation strategy, similar to a synecdoche 'in the form of a part standing for the whole (pars pro toto) or a whole standing for the part (totum pro parte)' (Reisigl & Wodak, 2001: 45). A whole-standing-for-part synecdoche about Islam or Muslims usually takes one of two main argumentative structures: an individual ayah can be taken from the Qur'an and used to (mis)represent the whole of the Qur'an; or the (usually negative) actions of certain Muslims are taken and used to (mis)represent the whole of Islam: in both, the characteristics of the whole take on the characteristics of a part (fallacy of composition). Part-standing-for-whole synecdoches also adopt one of two structures: the Qur'an is taken and used to (mis)represent (and explain) all Muslim belief and practices; or the (usually negative) characteristics of a 'Muslim State' are used to (mis)represent the character of Muslims living there: in both, the characteristics of a part take on a characteristic of the whole (fallacy of division). Note: a text does not have to argue that Muslims are monolithic in order to commit either of these essentialising fallacies, it simply needs to make an unwarranted leap from part (Muslim; ayah) to whole (all Muslims; the whole Qur'an) or visa versa.

Second, the Trust's 'closed views' are based on a rhetorical stratification of groups in which Islam and Muslims are presented as (in extreme cases) inherently undesirable or, to varying degrees, inferior to 'Us' and 'Our' value system. When these two features - essentialisation and stratification - appear in combination, the result will necessarily be prejudiced. There are a number of argumentative strategies a rhetor may employ to achieve such essentialisation and stratification: as we know the Runnymede Trust (1997) suggest five; Karim (1997) suggests four others that he labels argumentative topio (Muslims are violent, lustful, greedy and barbaric). These argumentative strategies find expression in a central constellation of themes aimed at problematising the Muslim presence - threat of violence, numbers pathology (more Muslims equals bigger threat), sexism, 'fundamentalism' and (most recently) the al-Qaeda connection. Liberal, or more 'left-wing' themes tend to emphasise the threat that a Muslim presence assumedly presents to personal liberties - Our freedoms, Our privileges and Our tolerant way of life. For example, the column by Polly Toynbee referred to at the outset of this chapter provoked the following letter to the editor: 'How refreshing to see challenged in print the taboo which seems to protect anybody peddling sexist, homophobic, unreasonable or unlikely views cloaked in the language of religion' (Steve Morris, 1997).

By way of contrast, conservative, or more 'right-wing' themes tend to emphasise the threat that a Muslim presence assumedly represents to 'the nation' - national security, national culture and the national religion (Christianity, of course!). Take the following letter to the editor, which clearly articulates many of these themes:

'Sir, Ahmer Khokhar tells us that his relatives support al-Qaeda and regard its members as martyrs. He confirms my fears on the wisdom of letting large numbers of people of alien culture settle in this country. If we are to make a reasonable extrapolation from his article, many Muslims here support al-Qaeda; of these it is reasonable to suppose that some fraction will want to express that support actively, through violence, since that is what al-Qaeda exists for. […] I reserve an especial contempt for the obscene ingratitude of these people towards a country that affords them the freedom to destroy the freedom they supposedly came for' (John Campion, 2003).

This letter is an example of what Aristotle called a forensic rhetoric: argumentation focused on the past in which the rhetor accuses (or defends) someone or something, with reference to the alleged injustice or prudence of things already done. In this case, Mr Campion questions 'the wisdom of letting large numbers of people of alien culture settle in this country' but, interestingly, does so via importing several notable features of the second Aristotelian rhetorical genre epideictic rhetoric: argumentation concerned with the present and directed at praising or censuring somebody with a view to 'proving him worthy of honour or the reverse' (Rhetoric, 17). Epideictic rhetoric is intimately associated to the character (ethos) of those referred and, just as a eulogy is based on affirming noble deeds, invectives of the sort in this example 'are based on facts of the opposite kind: the orator looks to see what base deeds - real or imaginary - stand to the discredit of those he is attacking, such as treachery' (Rhetoric, 126). In keeping with Aristotle's advice, Mr Campion misrepresents British Muslims in general as 'alien' immigrants, hostile to the freedoms 'they supposedly came for', 'many' of whom support the violent and militant al-Qaeda. This characterisation is then used to support the conclusion of his higher-level forensic argument: we should have kept them out!

Thus, when essentialisation and stratification appear in combination, the result will necessarily be prejudiced. However, it is where argumentative discourse foregrounds one of these strategies and omits the other that the problematic issue of distinguishing between 'prejudiced' and 'defensible' opinion arises. In other words: an essentialisation occurs when a text assumes a 'timeless' Islam exists, but does not derogate or vilify Muslims. A variety of radical Muslim political organisations adopt such an argumentative strategy, resulting in seemingly paradoxical similarities between their representation of a timeless Islam and the discourses of individuals and groups who aim to derogate Islam and Muslims. Stratification occurs when a text critiques Islam and argues (implicitly or explicitly) that an alternative religion/value system is superior, but does not essentialise (and hence misrepresent) Islamic beliefs or practices. The distinction between the acceptable (though contentious) and the discriminatory in these cases must lie in the reasonableness of the critique. To date, the theory of reasonableness in argumentation, which I consider to be the most sophisticated and practicable has been offered by van Eemeren and Grootendorst (1987, 1992, 1994a). According to their pragma-dialectical theory:

'Argumentation is seen as part of a procedure aimed at resolving a difference of opinion concerning the acceptability of a view or a standpoint. The moves made by the protagonist of the standpoint and those made by - or ascribed to - the real or imaginary antagonist in the discourse are regarded reasonable only if they can be considered as a contribution to the resolution of the difference of opinion. […] In order to comply with the dialectical norms of reasonableness […] the speech acts performed in the discourse have to be in agreement with the rules for critical discussion' (van Eemeren, Garssen & Meuffels, 2003: 275).

These rules for critical discussion 'represent ten different norms which are to be observed for resolving the difference' (van Eemeren & Grootendorst, 1994b: 63) and are as follows:

1. the freedom rule: participants must not prevent each other from putting forward standpoints or casting doubt on standpoints;
2. the burden of proof rule: whoever who puts forward a standpoint is obliged to defend it if asked to do so;
3. the standpoint rule: an attack on a standpoint must relate to the standpoint that has indeed been advanced by the protagonist;
4. the relevance rule: a participant may defend his or her standpoint only by advancing argumentation related to that standpoint;
5. the unexpressed premise rule: a participant can be held to the premises he leaves implicit; equally, an antagonist may not falsely suggest that a premise has been left unexpressed by the other participant;
6. the starting point rule: no participant may falsely present a premise as an accepted starting point, or deny a premise representing an accepted starting point;
7. the argument scheme rule: a standpoint may not be regarded as conclusively defended if the defence does not take place by means of an appropriate argument scheme that is correctly applied;
8. the validity, or logic rule: the reasoning in the argumentation must be logically valid or must be capable of being made valid by making explicit one or more unexpressed premises;
9. the closure rule: the failed defence of a standpoint must result in a protagonist retracting the standpoint, and a successful defence of a standpoint must result in an antagonist retracting his or her doubts;
10. the ambiguity, or usage rule: participants must not use any formations that are not sufficiently clear or confusingly ambiguous, and they must interpret the formations of the other participant as carefully and accurately as possible;
(from van Eemeren & Grootendorst, 1992: 208-209).

Violations of these rules produce unreasonable argumentation of various kinds in the form of fallacies. The strategies of essentialisation cited above for example, in which the negative actions of (people who happen to be) Muslims are presented as representative of either all Muslims or of Islam as a religion, are fallacies of composition because they contravene rule 8. These are by no means the only fallacies - indeed 'it is possible with each of the formulated discussion rules to indicate precisely which classical fallacies can be controlled through these rules' (van Eemeren & Grootendorst, 1994a: 23). Take the following example:

'I shall save my rejoicing over the recent announcement about state funding of two Islamic schools in Britain until I hear that a stronghold of Islam - Saudi Arabia - has put an end to its unashamed persecution of Christians' (Pastor Graham Horsnell, 1998).

This letter to the editor, in which the granting of voluntary aided status to two British Muslim schools is used as a cover to enable the author to insert a rather incoherent pseudo-analogous argument, is fallacious because it contravenes rule 4 - the relevance rule. The irrelevance of the argument to the conclusion offered illustrates the lengths to which the author is willing to go in order to argue for continuing anti-Muslim discrimination.

By way of a third and final example, the argumentum ad hominem is another traditional fallacy frequently committed by rhetors attempting to disparage Islam or exclude Muslims. To explain this fallacy it is necessary to quote van Eemeren and Grootendorst (1994b) at length:

'An implication of the first discussion rule is that a party may not improperly harm his collocutor's position as a serious discussion partner in any way. And this is precisely what happens in the various variants of the argumentum ad hominem. In the abusive variant one party denigrates the other party's expertise, intelligence or good faith; in the circumstantial variant, he suggests that the other party is driven by personal interests and is therefore not capable of making an impartial judgment; and in the tu quoque variant, he denounces an inconsistency in the other party's opinions or behaviour. In effect all three variants amount to one party claiming that the other party has no right to speak, thus violating the first rule for critical discussion' (ibid: 64).

Richardson (2001) examines an example of this fallacy in a letter to the editor written by Ray Honeyford, a long-time antagonist of Muslims. In the relevant section of the letter Mr Honeyford argues:

'the well known anti-racist pressure group The Runnymede Trust […] had made up its mind about the subject before the survey was carried out. This is made clear in a consultation paper issued in February entitled Islamophobia - its features and dangers. This paper is highly critical of British public opinion and represents British Muslims as an oppressed and persecuted group - views which the survey obligingly confirms' (Ray Honeyford, 1997).

Here Honeyford commits the circumstantial variant ad hominem - he alleges that the Trust were motivated by an ideological stake in achieving certain results and that the findings of their report can be rejected on the basis of this alleged bias. However, even if it were true that the Islamophobia Report confirmed the opinions of the Runnymede Trust's consultation paper (and, I hasten to add, this is not really true), it does not mean that the Trust had 'made up its mind about the subject before the survey was carried out' as Honeyford suggests, merely that consultation paper and survey produced the same results - results which could be true.

To summarise this approach to reasonableness in argumentation: although the critique of Islam and the implicit stratification - or even the explicit judgment - of Islam are legitimate (just as they are for Muslim or secular critiques of Christianity), this critique:

1. Should be temperate and not descend into hostility or attack.

2. Critical views should be accurate and relate to the actual beliefs, standpoints and
practices of Islam rather than distorted or perverse interpretations;

3. Argumentation should be reasoned, logical and non-fallacial.

4. Fundamentally, while individuals are free to disagree with Islam (provided they do not violate the above rules for critical discussion), such disagreement should always be prefaced by the principle of freedom of religion, and the right of Muslims to practice the religion of their own choice.

Violation of these rules will produce discourse, which is, to a greater or lesser extent, 'prejudiced against Islam'.

Discursive practices, discursive context

Next, we need to consider the context in which argumentative/textual representation of Islam takes place, since it is in social contexts that prejudicial arguments are endowed with racist effects. Despite each argumentative context being constituted by specific (and in some ways unique) complexes of social, political, historical and other extra-linguistic (e.g. institutional) discursive characteristics, for our purposes context may be considered in terms of three basic elements: the arguer; the audience; and the broader socio-political locality. Each of these contextual elements have a clear bearing on the opaque and transparent, implicit and explicit features of argumentation, and hence must be considered when distinguishing between unacceptable (racist) and acceptable (though debatable or contestable) argumentative representations of Islam.

Starting with the arguer, in the case of journalism we should ask ourselves: who is the journalist and for what purpose is s/he writing? The motivations of certain journalists writing about Islam are demonstrably questionable - take Peregrine Worsthorne for example, whose career (predominantly writing and editing the Sunday Telegraph) has been characterised by an antipathy towards non-white immigration and non-white British communities. In one column (Guardian, 18 January 1997) he openly admitted to his dread and dislike of Black people, and apologised for the 'shameful', 'wicked nonsense' he'd written about Britain's African Caribbean communities over the last thirty five years. However, he was not sorry for his treatment of Britain's Muslim communities who, he argued, still represent a problem because they fail to assimilate with White British values:

'[…] they believe too much in family values rather than too little; take religion too seriously rather than not seriously enough and insist on defying the new mores of permissive society rather than caving in to them' (cited in Law, 2002: 82).

Weeks later, while he explicitly denied any dislike of Muslims, it still formed an implicit undercurrent to his 'anti-PC' column 'I believe in Islamophobia' (Daily Telegraph, March 1st 1997). Once Worsthorne's journalistic writing is viewed in relation to his admitted anti-Muslim prejudice, it makes little difference how many times he claims: 'One can find in the Mohammedan religion [sic!] as many appeals for universal love and compassion as in the Christian religion'; or how many times he argues that blaming 'Mohammedanism for Islam's militancy […] makes no more sense than to blame the Sermon On The Mount for the atrocities of the IRA'. - His claimed openness and goodwill towards Islam is reduced to an apparent or 'show' concession by his unshakable loathing of Muslims. Similar accusations can be levelled at 'left-liberal' journalists such as Will Hutton, or more 'radical' journalists like Julie Burchill, who have regularly justified their open antipathy towards Islam using examples of Muslim wrong-doing (a whole-for-part synecdoche). In summary therefore, we need to consider the character and conduct of the arguer.

Second, the audience or readers of a text should be considered. After all, argument is not simply text for its own sake, but is practical, even political discourse aimed at persuading an audience of the acceptability of a thesis and thereby provoking them into an immediate or future course of action. Echoing this perspective, Perelman points out that 'it must not be forgotten that all argumentation aims somehow at modifying an existing state of affairs' (1979: 11), whether this be mental, social or political. Elsewhere, Perelman argues that 'argumentation does not aim solely at gaining a purely intellectual adherence. Argumentation very often aims at inciting action, or at least creating a disposition to act' (1982: 12). Thus, the (potential) effects of argument on their audiences also need to be considered: what is the audience/readership being convinced of and, since people 'do things with words' beyond their immediate communicative context, how are a text's argumentative conclusions likely to be used? For example: while we may believe that Ariel Sharon is a war criminal, responsible for the deaths of 2-3,000 men, women and children in Sabra and Shatila refugee camp massacres, I nonetheless feel uncomfortable reading arguments such as this on the website of the British National Party, given the well known anti-Semitism of the Party and its supporters. Would those reading this argument on the BNP's website (2003) be motivated into campaigning for Palestinian rights, or is it more likely that they would see it as (more) evidence of 'the Zionist conspiracy' which they should fight against? This is particularly important to consider when analysing argumentation, which an audience considers to be informed and authoritative - like the elite broadsheet press.

There are further reasons why an examination of the audience is key to understanding the form and function of (argumentative) discourse. We do well to remember that real-world arguments rarely take on a full and deductively valid form; in terms of argumentation theory, every day argumentation often deletes (or omits) the minor premise of an argument - in other words a statement which allows a move from a claim to a conclusion. Take an uncontroversial example:

John: I'm just going to put this cheque into the bank.
Albert: You can't do that, it's a Sunday.

Here Albert offers an argument as to why John cannot put a cheque into the bank, but omits the major premise providing the relevance of his argument. A reconstructed version of Albert's argument, with all premises present, could be:

A: It's a Sunday
[B: Banks are closed on Sundays]
C: Therefore, you can't deposit your cheque

The premise B is omitted and yet Albert's argument remains coherent. Why? Because the audience - John in the case of the exchange, or you the reader at present - fills in the missing premise in order to make the argument coherent. Albert assumes that the audience will know that banks are closed on Sundays and therefore this premise can go unstated. This argumentative structure is called an enthymeme - a kind of truncated syllogism, with a premise omitted or missing which the audience are themselves required to supply 'out of [their] stock of opinion and knowledge' (Bitzer, 1959: 407). Another way of explaining this type of argument is offered by Conley (1984): in an enthymematic argument, 'the rhetor argues A, so C, with the audience filling in the missing B to understand how the connection between A and C could be asserted' (ibid: 170). Enthymemes are a particularly powerful form of rhetorical argument: not only do they allow a rhetor to omit claims which, if stated explicitly, may lay him/her open to a charge of being prejudiced, but the audience usually completes the enthymeme with a premise which provides the 'best fit', in accordance with Grice's (1975) 'cooperative principle'. Thus, with enthymemes, audience members play an active role in convincing themselves.

Enthymemes therefore provide the analyst with a wealth of information about both the arguer and the audience, not least the kinds of opinion and knowledge about a subject, which the arguer assumes the audience, knows. While the structure of an enthymeme can be an appropriate one, equally (and in accordance with rule 5 cited above) a rhetor can be held to the premises s/he leaves implicit in such an argument scheme. Regarding enthymemes about Islam, Richardson (2004) illustrates that broadsheet newspapers often:

'presume their readers to hold prejudiced (racist?) opinions of Muslims, and require them to use this "knowledge" to fill in any implicit or missing premises, in order that their argumentative conclusions remain plausible. Without the essentialising and prejudiced "missing" premises of these enthymematic arguments, the information offered simply does not support [their] conclusions' (ibid: 78).

Take the following example, in which a journalist argues that Iran developing nuclear weaponry is, in and of itself, threatening:

'Some Foreign Office officials have warned that Iran is becoming a daunting military force in the region […] Experts fear that with the new technology, Iran may be less than two years from building the first Islamic nuclear bomb' (Sunday Times, Britain at heart of Iran nuclear web 18 January 1998).

Here, the verb phrase 'have warned' is highly ideological, placing a negative frame on the remainder of the predicate in a way, which alternative verbs ('have claimed'; 'have suggested'; 'believe') would not. This negativity is continued in the suggestion that Iran's military prowess is 'daunting' as opposed to (for example) 'consequential' or 'significant'. Further, by labelling Iran's (potential, suspected and unconfirmed) nuclear ordnance an 'Islamic nuclear bomb', the journalist imbues the device with a (Muslim) ideological agenda. And, in case the reader does not automatically associate this 'Muslim nuclear bomb' with 'threat', 'a Whitehall official' is quoted lower down claiming that Iran 'has become a far bigger threat' to 'the West' than even Iraq. Missing throughout this report is a premise, which explicitly links Iran's (alleged) weapons with the threat they pose to the region, or to 'Us', or to 'the West', or indeed to anyone.

Anti-Muslimism revisited: the war against terrorism

Finally, as stated above, we need to consider a text's 'words, images and themes in juxtaposition to […] the broader social, cultural and political context' (Daniels, 1997: 142). More specifically, we need to take into account the surrounding social and discursive locality (the occasion, the social, political and individual characteristics of participants and/or recipients) and to 'the broader socio-political and historical context which the discursive practices are embedded in and relate to' (Reisigl & Wodak, 2001: 41). Although this level of analysis necessarily involves a certain degree of speculation and conjecture, I suggest that a discourse may be considered anti-Muslim if it constructs, perpetuates or transforms racist social practices. Here I would like briefly to discuss 'the war against terror', instigated and evangelised by the British Government, fought by the USA and propped up by an assorted 'coalition of the willing'.

In contrast to some other commentators, I do not consider 'the war against terror' to be a war against Islam. There are certain notable exceptions of course, for example the extreme fanaticism of Anne Coulter. In one particularly infamous article following September 11th 2001, Coulter wrote of Muslims that 'we should invade their countries, kill their leaders and convert them the Christianity', resulting in her being sacked from the National Review (cited in Sifry & Cerf, 2003: 333). In a more recent article Coulter argued that Muslims:

'should be worried. They hate us? We hate them. Americans don't want to make Islamic fanatics love us. We want them to die. There's nothing like horrendous physical pain to quell angry fanatics […] Japanese kamikaze pilots hated us once too. A couple of well-aimed nuclear weapons, and now they are gentle little lambs' (Why we hate them, September 26th 2002, cited in Sifry & Cerf, 2003: 333).

However, the vast majority of argumentation employed in support of 'the war on terror' has not been Islamophobic, nor have its proponents attempted to justify war in reaction to the Muslim-ness of enemy combatants - quite the opposite in fact. For example, reacting to a question about whether 'this could degenerate into a conflict between […] Islam and Christianity', Ari Fleischer, the US President's spokesman, forcefully stated: 'This attack had nothing to do with Islam. This attack was a perversion of Islam' (cited in Brown, 2003: 96).

The war is, however, a war against Muslims; it is racist in its assumptions and racist in its outcomes. In the domestic sphere, it is only Muslims who have been targeted through racial profiling, only Muslims who have been arrested and, in the case of the US and UK, only Muslims who are currently being detained without trial for an indefinite period. In the international sphere, it is only Muslims and Muslim countries that have been targeted in this 'war against terror' focusing on the 'retail terrorism' of minority groups and ignoring the structural terrorism enforced by 'Western' neo-liberalism and its client regimes. For this reason, the 'war against terror' provides future analysts with a paradigm example of a discourse whose true significance lies not in the propositional content of its texts and its arguments (though these are of obvious interest, not least because of what they conceal), but in the effects they have upon the material world.

The war on Afghanistan is a case in point. In contrast to the righteous indignation following the indefensible terrorist attacks on September 11th 2001, the thousands of Afghan civilians killed both directly and indirectly by US retaliatory bombings 'have received the barest of concern from political leaders and the mainstream media, who have essentially deemed Afghan lives expendable to avenge the attack on the US. Afghan people are Unpeople, whose deaths go unnoticed' (Curtis, 2003: 49). In order to 'win' in as short and as decisive a manner as possible (and, in the words of Geoff Hoon, to send a 'clear message' to others), the US used weapons and tactics in Afghanistan that made a mockery of their claimed desire to avoid civilian casualties through 'precision' bombing. Curtis (2003) reminds us that the:

'US war strategy gradually escalated from using medium sized bombs to cruise missiles to bunker-busting 2,000lb bombs, then to B-52 carpet-bombing [including cluster bombs] and finally to the devastating 'daisy cutter' bombs that destroy everything in a 600-yard radius' (ibid: 51).

The terrible destructive capabilities of these and other weapons used to raze Afghanistan, elicited only the usual discourse of pseudo-sexual weapons technophilia from journalists across the UK. The idea that it was homes and civilian electricity-generating facilities that were being 'busted' not 'bunkers', or that it was children being cut to size and not 'daisies', are notable by their absence from press descriptions of such weapons and tactics. Meanwhile, journalists in Afghanistan - particularly television journalists - stationed themselves on top of the roofs of hotels in order to provide 'exciting pictures and backdrops' for their reports, often getting no closer than seven miles from a frontline filled with unwelcome images of Muslims being killed (Ridley, 2003: 249). The media feeding frenzy that greeted Yvonne Ridley's arrest and detention by the Taliban neatly illustrates the hierarchy of death operating in the British press. It was the arrest and possible mistreatment of Yvonne Ridley - the journalist, the woman, the mother - that received press attention (attention she has since criticised, see Ridley 2003), not the deaths of whole Afghan families, killed in their homes, as they gathered together to break their Ramadan fast.

Second, the war on Afghanistan was not fought to 'free' Afghanistan or to restore justice and human rights. Both the low regard for Afghan life during the bombing campaign and the lack of interest in providing humanitarian aid to help reinstate the rule of law outside Kabul (despite repeated assurances to the contrary) illustrate that 'defending civilisation against barbarity' was not high on the US/UK agenda. Attacking Afghanistan was an act of retaliation against a country that wouldn't put up much of an opposition, in order to show 'who's boss, not only to Afghans but to anyone else' (Curtis, 2003: 66). The deaths of Afghani civilians were viewed as a regrettable sideshow in the face of the US achieving this goal. To be clear: the Muslim-ness of the Afghan victims was never cited, or even implied, as a motivating reason for the US/UK attack. Indeed if the Muslim-ness of Afghan civilians had been cited as a motivating factor for war, then the argument for our 'wondrous benevolence' (Curtis, 2003: 74) would have been exposed for the lie it so clearly was. However, the outcome of the attack, and indeed the underlying logic of the discourse, was racist nonetheless, as illustrated by the way in which the war was justified, discussed and later evaluated. For example, when the BBC reported the deaths of seven US servicemen, they stated that there would eventually be US victory in Afghanistan 'but that it would be at a price' (Stephen Sackur, BBC 10 o'clock News, March 4th 2002, cited in Curtis, 2003: 55). As Curtis (2003: 55-56) points out, '[t]his was in March 2002, well after estimates of thousands of civilian deaths - but these deaths, being of mere Afghans, are not worthy of being called a "price"'.

This moral apathy towards the deaths of Muslims has persisted despite the barefaced illegality of this retaliation and of the subsequent US doctrine of 'preventative war', pioneered by the Bush-Blair alliance and tested in Iraq.

Conclusion

The boundaries between racist anti-Muslim argumentation and legitimate critique of both Islamic beliefs and practices and the wrong doing of Muslims, are rather fuzzy and often difficult to differentiate. Here I assume that critical discussion, or even arguments, about values and value systems are valid, providing, first, the arguers do not overstep the rules of reasonableness listed above; and second, this discussion takes place in a social context in which the standpoint of one side or the other is not used to derogate or discriminate. Further, while it is not always possible to consider the full consequences of one's actions, writers or commentators on Islam and the actions of Muslims should also bear particular attention to the possibility that their arguments may be used by third parties to support their own nefarious causes or conclusions. This should be of particular interest and concern to powerful genres of communication, such as journalism and academic discourse, given their manifest abilities to maintain and transform social practices.

In summary, an argument or text may be considered prejudiced against Islam if it is founded upon an essentialised representation of Islam and a rigid stratification of religions and other belief systems in which Islam is viewed as inherently inferior. Further, the insights of CDA demonstrate that communication should be viewed as a social practice, and that every communicative act reproduces or transforms society and culture, including its racialised power relations. Hence, argumentation on or about Islam and Muslims may additionally be said to be racist if underlying racial or ethnicist inequalities 'are actualised and reinforced' by the text (Essed, 1991: 52). In other words, we may consider a discourse to be racist if analysts can (plausibly) demonstrate that it encourages the social stratification of religious communities, perpetuates discrimination against Muslims or, in the case of 'the war against terror', it degrades the value of Muslim life to the point where innocent Muslims can be killed readily and with impunity.

Bibliography

Al-Azmeh, A. (1993). Islams and Modernities (2nd edition), London: Verso.

Amossy, R (2003) 'The Argumentation Dimension of Discourse' in van Eemeren, Blair, Willard & Snoeck Henkemans (eds.) Proceedings of the Fifth Conference of the International Society for the Study of Argumentation, pp. 13-16. Amsterdam: ISSA.

Aristotle (1981) Rhetoric and On Poetics, Pennsylvania, PA: The Franklin Library.

Beinin, J. & Stork, J. (1997) 'On the Modernity, Historical Specificity and International Context of Political Islam' in Beinin & Stork (eds.) Political Islam: Essays from Middle East Report, pp. 3-28, London: I.B. Tauris.

Bitzer, L. F. (1959) Aristotle's enthymeme revisited, Quarterly Journal of Speech, 45: 399-408.

Bourdieu, P. (1991) Language and Symbolic Power, edited by J. B. Thompson, translated by G. Raymond. London: Polity Press.

British National Party (2003) News, www.bnp.org.uk/news/2003_march/news_mar15
created March 15th 2003, accessed April 4th 2003.

Brown, R. (2003) 'Spinning the War: Political Communications, Information Operations and Public Diplomacy in the War of Terrorism' in Thussu & Freedman (eds.) War and the Media: Reporting Conflict 24/7, pp.87-100, London: Sage.

Cameron, D. (2001) Working With Spoken Discourse, London: Sage

Campion, J. (2003) 'Responses to a Muslim's conversion' (letter), The Times, January 10th.

Conley, T. M. (1984) The enthymeme in perspective, Quarterly Journal of Speech, 70: 168-187.

Cottle, S. (2000) 'Media Research and Ethnic Minorities: Mapping the Field' in Cottle (ed.) Ethnic Minorities and the Media: Changing Cultural Boundaries, pp. 1-30, Buckingham: Open University Press.

Curtis, M. (2003) Web of Deceit: Britain's Real Role in the World, London: Vintage

Daniels, J. (1997) White Lies: Race, Class, Gender and Sexuality in White Supremacist Discourse, New York: Routledge.

Deacon, D., Pickering, M., Golding P. & Murdock, G. (1999) Researching Communications: A Practical Guide to Methods in Media and Cultural Analysis, London: Arnold.

Dijk, T. A. van (1988) News as Discourse, Hillsdale, NJ: Lawrence Erlbaum.

Eemeren, van F.H., Grootendorst, R. (1987) 'Fallacies in pragma-dialectical perspective' in Argumentation, 1: 283-301.

Eemeren, F. H. van & Grootendorst, R. (1992). Argumentation, Communication and Fallacies: A Pragma-Dialectical Perspective, Hillsdale, NJ: Lawrence Erlbaum.

Eemeren, van F.H., Grootendorst, R. (1994a). 'Rationale for a pragma-dialectical perspective' in van Eemeren & Grootendorst (eds.) Studies in Pragma-Dialectics, pp. 11-28, Amsterdam: Sic Sat.

Eemeren, van F.H., Grootendorst, R. (1994b) 'Relevance reviewed: the case of "argumentum ad hominem"' in van Eemeren & Grootendorst (eds.) Studies in Pragma-Dialectics, pp. 51-68, Amsterdam: Sic Sat.

Eemeren, van F.H., Garssen, B. & Meuffels, B. (2003) 'The conventional validity of the pragma-dialectical freedom rule' in van Eemeren, Blair, Willard & Snoeck Henkemans (eds.) Proceedings of the Fifth Conference of the International Society for the Study of Argumentation, pp. 275-280, Amsterdam: ISSA.

Eemeren, F. H. van, Grootendorst, R., Snoeck Henkemans, F., Blair, J. A., Johnson, R. H., Krabbe, E. C.W., Plantin, Ch., Walton, D. N., Willard, C. A., Woods, J. and Zarefsky, D. (1996) Fundamentals of Argumentation Theory: A Handbook of Historical Backgrounds and Contemporary Developments, Mahwah, NJ: Lawrence Erlbaum Associates.

Essed, P. J. M. (1991) Understanding Everyday Racism: An Interdisciplinary Approach. Newbury Park, CA: Sage.

Fairclough, N. (1995) Media Discourse, London: Arnold.

Gee, J. P. (1999) An introduction to discourse analysis: theory and method, London: Routledge.

Grice, H. P. (1975) Logic and Conversation, in Syntax and Semantics, Cole, P. & Morgan, J. L. (eds) 41-58, London: Academic Press.

Hage, G. (1998) White Nation: fantasies of white supremacy in a multicultural society, Annandale, Australia: Pluto Press.

Halliday, F. (1996) Islam and the Myth of Confrontation, London: I. B. Tauris.

Halliday, F. (1999) 'Islamophobia reconsidered' in Ethnic and Racial Studies, 22(5): 892-902.

Honeyford, R. (1997) 'Muslims should change radically' (letter), Daily Telegraph, October 25th.

Horsnell, G. (1998) 'Islamic Schools' (letter), The Times January 19th.

Jackson, I. (1971) The Provincial Press and the Community, Manchester: Manchester University Press.

Karim, K. H. (1997) 'The Historical Resilience of Primary Stereotypes: Core Images of the Muslim Other' in Riggins (ed) The Language and Politics of Exclusion: Others in Discourse, pp.153-182, Thousand Oaks, California: Sage.

Law, I. (2002) Race in the News, Houndmills: Palgrave.

Mayes, I. (2001) 'Sincerely yours The readers' editor on... the value, and the hazards, of the letters page', Guardian June 23rd.

Mayes, I. (2003) 'Readers between the lines', Guardian, March 22nd.

Morris, S. (1997) 'Islamophobia' (letter) Independent, October 27th.

Perelman, Ch. (1979) The New Rhetoric and the Humanities, Dordrecht: Reidel.

Perelman, Ch. (1982) The Realm of Rhetoric, Notre Dame: University of Notre Dame Press.

Phillips, L. & Jørgensen, M. W. (2002) Discourse Analysis as Theory and Method, London: Sage.

Reisigl, M. & Wodak, R. (2001) Discourse and Discrimination: Rhetorics of Racism and Anti-Semitism, London: Routledge.

Richardson J.E. & Franklin, B. (2003) '"Dear Editor": race, readers' letters and the local press.' Political Quarterly, 74(2): 184-192.

Richardson, J. E. (2001) '"Now is the time to put an end to all this" Argumentative discourse theory and letters to the editor', Discourse and Society, 12(2): 143-168

Richardson, J. E. (2004) (Mis)Representing Islam: the racism and rhetoric of British Broadsheet Newspapers, Amsterdam: John Benjamins.

Ridley, Y. (2003) 'In the fog of war…' in Thussu & Freedman (eds.) War and the Media: Reporting Conflict 24/7, pp.248-252, London: Sage.

Rodinson, M (1979) Marxism and the Muslim World, London: Zed Press.

Runnymede Trust (1997) Islamophobia: A Challenge for Us All, London: Runnymede Trust.

Said, Edward W. (1997) Covering Islam: How the Media and the Experts Determine How We See the Rest of the World, London: Vintage.

Sifry M. L & Cerf, C. (eds.) (2003) The Iraq War Reader: History, Documents, Opinions, New York: Touchstone.

Titscher, S., Meyer, M., Wodak, R. & Vetter, E. (2000) Methods of Text and Discourse Analysis, London: Sage.

Toynbee, P (1997) 'In defence of Islamophobia', Independent, October 23rd.

Wober, J. M. (2004) 'Top people write to The Times' in British Journalism Review, 15(2): 49-54.

Wodak, R. & Meyer, M. (eds.) (2003) Methods of Critical Discourse Analysis. London: Sage.

Woods, L. A. & Kroger, R. O. (2000) Doing Discourse Analysis: Methods for Studying Action in Talk and Text, Thousand Oaks, Ca.: Sage.

Worsthorne, P. (1997) 'I believe in Islamophobia', Daily Telegraph, March 1st.

An earlier version of the argument articulated in this chapter appears in Richardson (2004).