The Propaganda Model: Theoretical and Methodological Considerations

This article provides an overview of the Propaganda Model and rehearses central theoretical considerations concerning the model’s overall understanding of media behaviour. The article then advances a contemporary state-of-the-art discourse on the methodological techniques that may be utilized in applying the model, highlights potentially complementary approaches to the critical study of mass media behaviour and explores criticisms of the model. The ‘Propaganda Model’ (PM) of media operations advanced by Herman and Chomsky (1988) is analytically and conceptually concerned to engage with the question of how ideological and communicative power connect with economic, political and social power, and to explore the consequent effects upon media output. This article overviews the PM and highlights central arguments associated with the PM regarding overall patterns of media behaviour. The article then considers ways in which the PM may be applied, exploring the methodological techniques that may be utilized in applying the model. The article also highlights potentially complementary approaches to the critical study of mass media behaviour and explores criticisms of the model. The Herman–Chomsky Propaganda Model – An Overview The PM’s five filter elements draw attention to the main structural constraints that impact overall patterns of media performance. Herman and Chomsky correctly observe that most mainstream media are themselves typically large corporations, ‘controlled by very wealthy people or by managers who are subject to sharp constraints by owners and other market-profit-oriented forces’ (Herman and Chomsky, 1988, 14). Their model suggests that ownership, size and profit ______________________________ Westminster Papers in Communication and Culture © 2009 (University of Westminster, London), Vol. 6(2): 43-58. ISSN 1744-6708 (Print); 1744-6716 (Online) Westminster Papers in Communication and Culture 6(2) 44 orientation will influence media behaviour in a range of ways and will ultimately encourage a right-wing bias within mainstream media discourses. Their PM observes that advertising is the principal source of revenue for most mainstream, commercial media, thus media discourses tend to reflect the interests of advertisers and the market. Taken together, the first two filters suggest that political-economic dimensions play heavily into news production processes, highlighting the macro-level structural dimensions that in effect ‘shape’ mainstream news discourses. The third filter of Herman and Chomsky’s model draws attention to and highlights the ways in which news discourses are socially constructed vis-à-vis sources (see Herman and Chomsky, 1988, 19–23; Klaehn, 2005a, 4–7). Institutionally affiliated sources (the ‘primary definers’ of social reality) typically dominate news discourses. As a result, news comes to reflect institutional interests on a macro level. Within individual news stories, preferred meanings are typically encoded into media texts, influencing how news articles are constructed vis-à-vis their headlines and leads, as well as overall story presentation, particularly in relation to choices of emphasis and overall tone. Encoding/decoding is associated with the work of Stuart Hall (1980), whereas Herman and Chomsky’s PM is typically thought to be concerned to identify bias within media discourses (see Chomsky, 1989; Klaehn, 2002a; 2002b; 2003a; 2006a; 2006d; 2008). The PM does not embrace encoding/decoding but is concerned to delineate the extent to which particular features of media texts (such as the aforementioned basic constituent elements of particular news stories) structure news discourses so as to encourage intended, preferred readings. The PM makes professionalism as ideology relevant in relation to both journalists and the institutionally affiliated sources that typically define what comes to be understood as ‘news’ in the first instance. The model stresses the symbiotic relationship between with journalists and agents of power. Preferred meanings that are structured (encoded) into news discourses are overwhelmingly ‘those that are functional for elites’ (Herman and Chomsky, 1988, 23). Herman (2000) writes that: ‘Studies of news sources reveal that a significant proportion of news originates in public relations releases. There are, by one count, 20,000 more public relations agents working to doctor the news today than there are journalists writing it.’ The model’s fourth filter element also brings the concept of power into play, stressing that dominant institutional actors possess the requisite social-political power to exert subtle or not-so-subtle control over patterns of media performance (see Herman and Chomsky, 1988, 26; Klaehn, 2005a, 4; see also Dinan and Miller, 2007; Everton, 2005; Jensen, 2005, 2006; Lee and Solomon, 1990; McChesney, 1999, 2008; McMurtry, 1998; Miller, 2007; Parenti, 1986; Scatamburlo-D’Annibale, Klaehn, The Propaganda Model... 45 2005; Winter, 2002; 2007). Herman and Chomsky hypothesize that these first four filter elements dominate ‘real-world’ news production processes. While the filters operate on an individual basis, they also continuously interact with one another. How the various elements play out in reality will depend upon specific time/place situational contexts. The model’s fifth filter was originally ‘anti-communism’ but has since been modified and broadened to refer to dominant ideological elements. Analytically, the fifth filter is extremely useful and applicable to a range of case studies. It may play out in different ways at different times, contingent upon specific time/place contexts, and is extremely broad (as are many other concepts within the social sciences, such as hegemony and/or patriarchy, for instance). That the fifth filter is so generalized makes it relatable to a range of social phenomenon, and creates space for the PM to be utilized in a variety of social scientific research. For instance, it could provide a framework for assessing othering in the mainstream media (for additional context, see Said, 1978; 1997; 2007). Within the context of the PM’s overall critical approach, the ideological representations of race and ethnicity in media discourses may be explored and connected to historical and contemporary political-economic dimensions. Such analysis would broaden understanding of the ways in which media discourses intersect with broader power struggles. The fifth filter element may be related to any number of case studies involving power and powerlessness, and seems particularly well-suited for analysis concerned to investigate media and the legitimatization of power (see also Bourdieu, 1991; Chomsky, 1989; 1991; Herman, 1992; 1999; Herman and O’Sullivan, 1990; 1991; Jensen, 2006; Kellner, 1992; Miller, 2003; Pilger, 1998; Winter, 2007). Scholars working within the traditions of Communication and Cultural Studies may apply the PM to a wide range of topic and subject areas. Students wishing to explore how capitalism is sustained vis-à-vis legitimizations might take up research involving workplace safety and insurance, workers’ rights issues (see Chen et al., 2006) and/or the depowering of unions (see McMurtry, 1998; Winter, 2007), for instance, and would find the PM extremely useful in enabling empirical research within these areas. The fifth filter element, in addition to being particularly relatable, is oriented toward broadening understandings of how ideological power intersects with political-economy and dimensions of social class. This is a facet of the PM that has yet to be fully explored but will almost certainly provide a foundation for much empirical research in the future. The focus of the modified fifth filter remains consistent with its original incarnation; this filter may now be termed simply ‘the dominant ideology.’ Herman and Chomsky (1988, 29) originally wrote of the fifth filter element: ‘This ideology helps mobilize the populace against an enemy, and because the concept is Westminster Papers in Communication and Culture 6(2) 46 fuzzy it can be used against anybody advocating policies that threaten property or support accommodation with Communist states and radicalism.’ The filter may also refer to recurring thematic elements that tend to go unquestioned but are observable over extended periods of time, such as the merits of capitalism, the deregulation of private enterprise and media bias against unions (see Herman, cited in Wintonick and Achbar, 1994, 108). Concurrently, the fifth filter element is an optic with which to explore how fear is deployed within media discourses. Chomsky (Chomsky and Barsamian, 1998, 41) explains that media may generate fear and/or redirect existing fear, depending upon specific contexts, whenever it is ideologically serviceable to the interests of power. Fear may be deployed as an ideological control mechanism and used to legitimize policies, mobilize resources and push specific agendas. How the fifth filter plays out – be it in relation to dominant ideological principles, power and/or othering – is upon specific time/place situational context. The five filter elements outlined above constitute the foundations of the PM (see Herman, 2000; Klaehn, 2002a; 2003a; 2003b). Taken together, the five filters provide a framework that illuminates why and how structural dimensions encourage a systematic right-wing bias and limited range of debate within mainstream media discourses. The five filters provide a basis for the PM’s general argument that the news which is deemed ‘fit to print’ will overwhelmingly be that which is politically and ideologically advantageous to the interests of power. The PM makes first-, secondand third-order predictions (see Klaehn, 2003b; 2005a, 12–15). Its first-order predictions relate to overall patterns of media behaviour (see Chen et al., 2006; Cromwell, 2006; Eglin, 2005; Everton, 2005; Winter, 1992; Winter and Klaehn, 2005). Academic discussions of the PM typically begin and end with the five filter elements. The PM’s overall scope, however, is significantly more far-reaching than standard treatments suggest. The model highlights how ideology, communicative power and media texts link to social organization, cultural education and pervasive social, political and economic inequalities (see Chomsky, 1997a; 1997b; Winter, 2007). The model suggests that mainstream commercial media are structurally predisposed to endorse, legitimize and promote the interests of power (Chomsky, 1989; Klaehn, 2002a; 2002b; 2003a; also see Bagdikian, 1992; Barker, 2009; Carey, 1995; Edwards and Cromwell, 2005; Herman, 1999; Herman and McChesney, 1997). Herman (cited in Klaehn, 2008) explains that the PM’s theoretical origins reside in ‘the economic model of industrial organization that traces back to the great British economist Alfred Marshall’, which: ... assumed its more modern form at Harvard with Edward S. Mason and his student and later Berkeley academic Joe S. Bain. Our thinking was also Klaehn, The Propaganda Model... 47 influenced by pioneering media analysts whose ideas also flowed into our work: Warren Breed, Gaye Tuchman; Ben Bagdikian, Philip Elliott, Eric Barnouw, Peter Golding, Stuart Hall, Leon Sigal, and others. The PM argues that mainstream commercial mass media function as a ‘guided market system’ and explains why media discourses will reflect the interests of power within different time/place contexts. According to Herman and Chomsky, ideological control will be most effective when there is consensus among elites, whereas disagreement among elites will create space for dissent (see Mullen, 2008). As noted above and highlighted elsewhere (see, most notably, Klaehn, 2003a), the PM does not predict that media are closed to dissent. This is a rather basic point that critics of the PM often miss, which is (by this point in time) at best perplexing, especially considering that Herman and Chomsky had explicitly addressed this point in the opening pages of the original version of Manufacturing Consent, stating that: ‘The news media are not a solid monolith on all issues’ (Herman and Chomsky, 1988, xiii). Herman and Chomsky have both written at length about dissent culture (see Chomsky, 1997a; Klaehn, 2002a). Herman (2000) has pointed out that critics of the PM often ignore the fact that the PM is actually ‘about how the media work, not how effective they are’. Numerous studies provide support for the view that media establish priorities, perpetuate events and mobilize public opinion in line with elite interests (see, for example, Babe, 2005; Barker, 2009; Chomsky, 1989; Edwards and Cromwell, 2005; Herman, 1992; 1999; Kellner, 1992; Klaehn, 2002b; Lee and Solomon, 1990; McChesney, 1999; 2008; McMurtry, 1998; Miller, 2003; Parenti, 1986; SourceWatch, 2009; Winter, 1992; 2002; 2007; Winter and Klaehn, 2005; see also Millar and Dinan, 2008). While many assume that media are dedicated to the broader interests of democracy and public/cultural education, Herman (cited in Klaehn, 2008) states that: ... democracy and public education are not primary aims of the mainstream media; the former, if fully realized, might well be damaging to the ends of the powerful; the latter also, unless properly channeled and limited, could be injurious to the powerful. These incompatibilities are likely to increase if inequality grows and if a military ethos and culture become steadily more important (as they have). The mainstream media will respond with attacks on and marginalization of ‘populism’ with its equalitarian tendencies, and will normalize enormous military budgets and wars. Westminster Papers in Communication and Culture 6(2) 48 Central Methodological Techniques Associated with the PM While Herman and Chomsky have both written about the methodological techniques associated with the PM, they have not yet produced a single article or book chapter devoted exclusively to methodology. One would commonly be required to consult a wide range of articles, interviews and book chapters in order to delineate the full range of methodological techniques associated with the model. This section provides a comprehensive overview of the PM’s methodology, outlining ways in which the model may be applied. Rather than presuming that news discourses somehow exist within a vacuum – apart from particular time/place contexts and various relevant dimensions of social life – the PM explores media discourses in relation to historical and contemporary political-economic contexts. The PM makes historical and politicaleconomic elements directly relevant to the whole enterprise of media analysis, regardless of the specific approach one might take in relating and applying the model. One of the strengths of the PM is that it may be applied in a range of different ways, and it affords opportunities for multiplicity of focus. Fundamentally, the model suggests that how issues and topics are treated by media will be bound to the interests of power. The quantity and quality of coverage accorded certain news stories and events will differ accordingly. Herman and Chomsky maintain that observable disparities in media treatment of co-occurring historical events, or paired examples, can enable critical insight into patterns of media behavior (see Klaehn, 2005a, 10–12). This methodological technique may be modified so as to enable insight into how media prioritize and treat similar cases/incidents which, while perhaps not co-occurring, may share common and/or contrasting core contextual elements. A ‘case study approach’ such as this could usefully be applied and would enable insight into how media function and socially construct news across different time/place contexts. It would also enable exploration of relatively recent news events. Another methodological technique associated with the PM entails analysis of how victimized parties are represented within media discourses. The PM predicts that ‘worthy victims’ (victims of state terror enacted by official enemy states) will be accorded a significant volume of coverage and will be humanized within this coverage. News relating to victims of official enemy states will be constructed in ways that trigger an emotional response and mobilize public opinion and outrage. Conversely, the model predicts that ‘unworthy victims’ (victims of state terror undertaken by the United States, its allies and client states) will be accorded only slight detail and comparatively minimal humanization. The PM predicts that unworthy victims will be accorded treatment that does not prompt or elicit reader Klaehn, The Propaganda Model... 49 emotion, interest or sympathy (Herman and Chomsky, 1988, 35). The model’s general argument, then, is that how victims and events are portrayed within mainstream media discourses will largely be dependent upon the interests of power within specific time/place contexts. This area of focus could also be significantly broadened and expanded to include events, issues and other groups of actors (beyond victims) (see, for example, Winter and Klaehn, 2005). The model’s focus upon how specific actors are represented within media discourses is relatable to both domestic and international events. Although the PM originated in the United States, a range of recent studies demonstrate that the model may be usefully applied to media discourses within other countries, where the media systems and political structures in play may be very different (see Edwards and Cromwell, 2005; Everton, 2005; Gunn, 1994; SourceWatch, 2009; Winter, 2007; Winter and Klaehn, 2005; see also Corner, 2003, 367; Klaehn, 2003b; 2006a; 2006b; 2006c; 2006d). This research indicates that the PM is international in overall scope and applicability. One may go further still and suggest that the model’s analysis of victim representation should be explicitly linked to power. The methodological technique most favoured by Chomsky in his political writings entails exploration of the ‘boundaries of the expressible’ or range of permitted opinion on crucial topics within media discourses. This involves exploring and assessing what particular facts, details and/or arguments are present within and absent from mainstream media discourses (see Chomsky, 1989, 59). The general argument here is consistent with the PM’s overall conceptualization regarding patterns of media behaviour: that which conforms to the interests of power is permissible and that which may be threatening to the interests of power will typically falls outside the range of permitted opinion. Applying the PM in the first instance, then, entails the quantitative and qualitative exploration of media discourses. Assessing the quantity of news coverage accorded particular cases and/or issues is relatively straightforward but can lend significant insight into how media prioritize and cover specific issues/events as opposed to others. The PM emphasizes qualitative exploration of the boundaries of debate. Sources, emphasis, placement, fullness of treatment, context, tone and evident range of debate on central issues and topics are observable dimensions of media discourses that may be qualitatively assessed when utilizing the PM to undertake detailed media analysis. The PM assumes that news discourses are bound to power and predicts that the primary sources of news will be agents of power. The model may be utilized to explore how power pervades news discourses vis-à-vis source selection. Applying the PM entails consideration of the extent to which news narratives and debates Westminster Papers in Communication and Culture 6(2) 50 are set within parameters that conform to the interests of power. Utilizing the PM to explore the extent to which official sources dominate mainstream media discourses can enable more nuanced and sophisticated understanding than mere quantitative content analysis. Testing the PM should entail qualitative exploration of the degree to which official sources are favoured within news discourses and the extent to which news narratives and discernible boundaries of debate within them are in effect defined by official sources. Evaluating data for content, omissions and style of presentation should be undertaken in order to delineate the extent to which news discourses are ideologically inflected. Such analysis will involve assessing media choices regarding how news stories are framed and ways in which particular events, issues, topics and/or actors are represented within particular media texts. Assessing what is present within and absent from media discourses will connect in turn to source selection and boundaries of debate. Another dimension of news discourses regarded as centrally important to the PM is the inclusion or absence of photographs within media texts. Since visual imagery can powerfully impact awareness and interpretation, the PM attaches particular importance to photographs. Analysis of these elements can enable understanding on the part of the critical analyst relating to overall patterns of media behaviour. In its original incarnation, the PM stressed media choices relating to overall story selection and story treatment. Herman (cited in Klaehn, 2008) comments that: The overlaps with critical discourse frames are numerous, but this is because the subject is immense and many tracks can be followed that are often not inconsistent with one another but stress different things. We don’t stress subtle language variations and/or the nuances in effects when the elite are split and a certain amount of dissent becomes permissible. Our emphasis is on the broader routes through which power affects media choices, how this feeds into media campaigns, and how it results in dichotomization and systematic double standards. The propaganda model focuses heavily on the institutional structure that lies behind news-making in ‘a world of concentrated wealth and major conflicts of class interest.’ This leaves lots of room for other tracks and subtracks in areas we deal with. The PM and critical discourse analysis both suggest that analysis of textual prominence within media discourses can impart insight into how power and meaning intersect (see, most notably, Fairclough, 1989; 1995a; 1995b; 2002; see also Giroux, 2001; Jhally, 2006; Lee and Solomon, 1990; Van Dijk, 2008). Such analysis involves careful consideration of a number of particular features of media texts and analysis of the ways in which media texts are structured and framed (see Klaehn, The Propaganda Model... 51 Van Dijk, 1988; Winter, 1992; 2007). Analysis of various dimensions of media texts is often required and may be utilized in applying the PM. Headlines are crucially important given the significant roles that they play in impacting how readers understand and subsequently interpret news stories. Readers are conditioned to presume that the most important or ‘newsworthy’ information is conveyed at the outset of media texts. Because headlines and leads structure interpretation, they are of importance with regard to analyzing textual prominence. Of course the most central aspect of textual prominence is the quantity of news coverage accorded particular stories. Beyond this, fore-grounding and backgrounding are important considerations, as are questions of what is made explicit within media discourses and what is presented merely as incidental background fact or omitted altogether (see Fairclough, 1995a, 103–8). For the PM, omissions are of central importance and connect to the boundaries of debate. Media choices regarding sources, textual prominence, presuppositions, fore-grounding and backgrounding influence how new articles are understood and interpreted. Such choices create meanings and structure interpretative frameworks through presentation of content. Media discourses often include presuppositions about previously occurring events that also may be ideologically inflected. The media analysis techniques set out by Herman and Chomsky in Manufacturing Consent and elsewhere enable sophisticated analysis of media discourses. The PM connects media and the ideological realm with social, political, economic dimensions of social life. The fact that the PM’s explanatory power is not limited by geographic borders is a great strength of the model, in my view. The model is relatable to both international and domestic news events. Critiques Associated with the Propaganda Model Criticisms of the PM within the context of scholarly discourse have by this point in time been quite exhaustively rehearsed (Herman, 2000; Klaehn, 2002a; 2003a; 2005b; 2008; Mullen, 2008), yet misconceptions about the model continue to circulate widely: in textbooks, university departments, classrooms, and on the worldwide web. Herman (2000) comments that many of the initial critiques advanced against the PM ‘displayed a barely-concealed anger, and in most of them the Propaganda Model was dismissed with a few superficial clichés (conspiratorial, simplistic, etc.), without fair presentation or subjecting it to the test of evidence’. On why the anger and hostility existed, Herman (cited in Klaehn, 2008) explains: The resistance and hostility to the propaganda model had several sources. One is that it is a radical critique, whose implication is that modest reforms that don’t alter the structure very much aren’t going to affect media performance very much. This is hard for non-radicals to swallow. Another source of resistance has Westminster Papers in Communication and Culture 6(2) 52 been based on our relatively broad brush strokes with which we model a complex area. This makes it allegedly too mechanistic and at the same time lacking in a weighting of the elements in the model! But we don’t claim that it explains everything and we are clear that elite differences and local factors (including features of individual media institutions) can influence media outcomes. We argue that the model works well in many important cases, and we await the offering of one that is superior. But we also acknowledge that there remains lots of room for media studies that do not rest on the propaganda model. This same room opens the way to criticizing the model for its failure to pursue those tracks and fill those spaces. Critique that deliberately avoids the whole issue of evidence and factual support for the PM continues to persist (see Corner, 2003; Danesi, 2008, 44; see also Klaehn, 2003b; 2005b) and (at least in Canada) it is extremely rare to find introductory textbooks within the fields of Sociology, Communication Studies or Cultural Studies that even mention the PM. When the PM is mentioned, often only the five filter elements are discussed. Sometimes critique of the model is also presented. Generally, the model is simply ignored. As noted, recent research indicates that the PM has international resonance. Scholars from Canada, the US and Europe have demonstrated the model’s applicability to a diverse range of issues and topics. The PM has been usefully employed to undertake analysis involving a broad range of international issues (such as the ‘War on Terror’ and the near-genocide in East Timor) and numerous domestic social and political issues (for a comprehensive listing of studies applying the PM, see SourceWatch, 2009). It is certainly true that the PM argues, in the first instance, that media discourses are shaped by market forces. This is for the simple reason that ownership, size and profit orientation of dominant media impact significantly upon the contexts in which discourses are conceived and produced (see Bagdikian, 1992; Golding and Murdock, 1991; Lee and Solomon, 1990; McChesney, 1999; 2008; Murdock and Golding, 1977). The PM does not assume that media are monolithic, nor does the model ignore dissent. Chomsky has in fact described the media system as inherently unstable (see Chomsky, 1997a; Klaehn, 2003b; 2005b, 231). While the PM describes a system in action, it makes no claims regarding how effective media may or may not be. Scholars have demonstrated, however, that media do have significant impacts in relation to public and cultural education (Ginsberg, 1986; Klaehn, 2003b; Lee and Solomon, 1990; Miller, 2003) and are also politically influential (see Chomsky, 1989; 1991; Cromwell, 2006; Herman, 1985; 1999; Klaehn, 2002b; 2006b; 2006d; Winter, 2002; 2007). Generally speaking, criticisms of the PM fall into one or more of the following three categories: (1) critique motivated by political, ideological and sometimes Klaehn, The Propaganda Model... 53 even personal opposition to the model; (2) critique of the PM’s assumptions, arguments and/or methodological approaches; and (3) critique centered around the fact that the PM is not all-encompassing and does not explain everything, in every context. Critique motivated by political, ideological and/or personal opposition to the PM may be implicit within (2) and (3). Such bias against the model may play out in different ways depending upon specific time/place contexts. Some suggest that the PM casts media audiences as passive and easily manipulated, but (as the author has noted elsewhere) the PM actually makes no claims regarding audience effects, nor does it take media audiences or ruling-class interests for granted (for a detailed discussion, see Klaehn, 2003a; 2005b, 227–33). It is concerned merely to delineate the extent to which media discourses are ideologically inflected in relation to structural dimensions of power and it affords ways of easily ‘testing’ the substantive predictions it advances. The PM does not assume that media audiences read texts in ways that are simple or passive (Chomsky, 1989, vii; Klaehn, 2002a; 2003b; 2005a; 2005b; Rai, 1995, 53). Herman and Chomsky actually suggest that so much effort and expense are deployed on a daily basis by elites (vis-à-vis spin and PR) precisely because influencing public opinion is an ongoing process. Consent is typically negotiated and must be actively won and re-won (hegemony) over time. Regarding claims that the Internet weakens and/or disproves the PM, Herman (2000) comments that: Some argue that the Internet and the new communication technologies are breaking the corporate stranglehold on journalism and opening an unprecedented era of interactive democratic media. There is no evidence to support this view as regards journalism and mass communication. In fact, one could argue that the new technologies are exacerbating the problem. They permit media firms to shrink staff even as they achieve greater outputs, and they make possible global distribution systems that reduce the number of media entities. Although the new technologies have great potential for democratic communication, there is little reason to expect the Internet to serve democratic ends if it is left to the market. The PM is a democratic and critical model that engages directly with the question of how economic, political, social and communicative power intersect with discourse phenomena. The model affords clear pathways for empirical investigation. Its emphasis on political-economy has also made it a target for critique. Herman (cited in Klaehn, 2008) comments that the PM: ... is a radical model, a class-based and class-bias model, and that in itself will explain much of the hostility. That will make it ‘political,’ whereas analyses that Westminster Papers in Communication and Culture 6(2) 54 take the status quo as a given and that confine themselves to modest reforms are ‘non-political.’ This kind of critique is implicitly political.

The 'Propaganda Model' (PM) of media operations advanced by Herman and Chomsky (1988) is analytically and conceptually concerned to engage with the question of how ideological and communicative power connect with economic, political and social power, and to explore the consequent effects upon media output. This article overviews the PM and highlights central arguments associated with the PM regarding overall patterns of media behaviour. The article then considers ways in which the PM may be applied, exploring the methodological techniques that may be utilized in applying the model. The article also highlights potentially complementary approaches to the critical study of mass media behaviour and explores criticisms of the model.

The Herman-Chomsky Propaganda Model -An Overview
The PM's five filter elements draw attention to the main structural constraints that impact overall patterns of media performance. Herman and Chomsky correctly observe that most mainstream media are themselves typically large corporations, 'controlled by very wealthy people or by managers who are subject to sharp constraints by owners and other market-profit-oriented forces' (Herman and Chomsky, 1988, 14). Their model suggests that ownership, size and profit ______________________________ Westminster Papers in Communication and Culture © 2009 (University of Westminster, London), Vol. 6(2): 43-58. ISSN 1744-6708 (Print); 1744-6716 (Online) orientation will influence media behaviour in a range of ways and will ultimately encourage a right-wing bias within mainstream media discourses. Their PM observes that advertising is the principal source of revenue for most mainstream, commercial media, thus media discourses tend to reflect the interests of advertisers and the market. Taken together, the first two filters suggest that political-economic dimensions play heavily into news production processes, highlighting the macro-level structural dimensions that in effect 'shape' mainstream news discourses.
The third filter of Herman and Chomsky's model draws attention to and highlights the ways in which news discourses are socially constructed vis-à-vis sources (see Herman and Chomsky, 1988, 19-23;Klaehn, 2005a, 4-7). Institutionally affiliated sources (the 'primary definers' of social reality) typically dominate news discourses. As a result, news comes to reflect institutional interests on a macro level. Within individual news stories, preferred meanings are typically encoded into media texts, influencing how news articles are constructed vis-à-vis their headlines and leads, as well as overall story presentation, particularly in relation to choices of emphasis and overall tone. Encoding/decoding is associated with the work of Stuart Hall (1980), whereas Herman and Chomsky's PM is typically thought to be concerned to identify bias within media discourses (see Chomsky, 1989;Klaehn, 2002a;2002b;2003a;2006a;2006d;. The PM does not embrace encoding/decoding but is concerned to delineate the extent to which particular features of media texts (such as the aforementioned basic constituent elements of particular news stories) structure news discourses so as to encourage intended, preferred readings.
The PM makes professionalism as ideology relevant in relation to both journalists and the institutionally affiliated sources that typically define what comes to be understood as 'news' in the first instance. The model stresses the symbiotic relationship between with journalists and agents of power. Preferred meanings that are structured (encoded) into news discourses are overwhelmingly 'those that are functional for elites' (Herman and Chomsky, 1988, 23). Herman (2000) writes that: 'Studies of news sources reveal that a significant proportion of news originates in public relations releases. There are, by one count, 20,000 more public relations agents working to doctor the news today than there are journalists writing it.' The model's fourth filter element also brings the concept of power into play, stressing that dominant institutional actors possess the requisite social-political power to exert subtle or not-so-subtle control over patterns of media performance (see Herman and Chomsky, 1988, 26;Klaehn, 2005a, 4; see also Dinan and Miller, 2007;Everton, 2005;Jensen, 2005Jensen, , 2006Lee and Solomon, 1990;McChesney, 1999McChesney, , 2008McMurtry, 1998;Miller, 2007;Parenti, 1986;Scatamburlo-D'Annibale, 2005;Winter, 2002;. Herman and Chomsky hypothesize that these first four filter elements dominate 'real-world' news production processes. While the filters operate on an individual basis, they also continuously interact with one another. How the various elements play out in reality will depend upon specific time/place situational contexts.
The model's fifth filter was originally 'anti-communism' but has since been modified and broadened to refer to dominant ideological elements. Analytically, the fifth filter is extremely useful and applicable to a range of case studies. It may play out in different ways at different times, contingent upon specific time/place contexts, and is extremely broad (as are many other concepts within the social sciences, such as hegemony and/or patriarchy, for instance). That the fifth filter is so generalized makes it relatable to a range of social phenomenon, and creates space for the PM to be utilized in a variety of social scientific research. For instance, it could provide a framework for assessing othering in the mainstream media (for additional context, see Said, 1978;1997;. Within the context of the PM's overall critical approach, the ideological representations of race and ethnicity in media discourses may be explored and connected to historical and contemporary political-economic dimensions. Such analysis would broaden understanding of the ways in which media discourses intersect with broader power struggles. The fifth filter element may be related to any number of case studies involving power and powerlessness, and seems particularly well-suited for analysis concerned to investigate media and the legitimatization of power (see also Bourdieu, 1991;Chomsky, 1989;Herman, 1992;Herman and O'Sullivan, 1990;Jensen, 2006;Kellner, 1992;Miller, 2003;Pilger, 1998;Winter, 2007). Scholars working within the traditions of Communication and Cultural Studies may apply the PM to a wide range of topic and subject areas. Students wishing to explore how capitalism is sustained vis-à-vis legitimizations might take up research involving workplace safety and insurance, workers' rights issues (see Chen et al., 2006) and/or the depowering of unions (see McMurtry, 1998;Winter, 2007), for instance, and would find the PM extremely useful in enabling empirical research within these areas. The fifth filter element, in addition to being particularly relatable, is oriented toward broadening understandings of how ideological power intersects with political-economy and dimensions of social class. This is a facet of the PM that has yet to be fully explored but will almost certainly provide a foundation for much empirical research in the future.
The focus of the modified fifth filter remains consistent with its original incarnation; this filter may now be termed simply 'the dominant ideology. ' Herman and Chomsky (1988, 29) originally wrote of the fifth filter element: 'This ideology helps mobilize the populace against an enemy, and because the concept is fuzzy it can be used against anybody advocating policies that threaten property or support accommodation with Communist states and radicalism.' The filter may also refer to recurring thematic elements that tend to go unquestioned but are observable over extended periods of time, such as the merits of capitalism, the deregulation of private enterprise and media bias against unions (see Herman, cited in Wintonick and Achbar, 1994, 108).
Concurrently, the fifth filter element is an optic with which to explore how fear is deployed within media discourses. Chomsky (Chomsky and Barsamian, 1998, 41) explains that media may generate fear and/or redirect existing fear, depending upon specific contexts, whenever it is ideologically serviceable to the interests of power. Fear may be deployed as an ideological control mechanism and used to legitimize policies, mobilize resources and push specific agendas. How the fifth filter plays out -be it in relation to dominant ideological principles, power and/or othering -is upon specific time/place situational context.
The five filter elements outlined above constitute the foundations of the PM (see Herman, 2000;Klaehn, 2002a;2003a;2003b). Taken together, the five filters provide a framework that illuminates why and how structural dimensions encourage a systematic right-wing bias and limited range of debate within mainstream media discourses. The five filters provide a basis for the PM's general argument that the news which is deemed 'fit to print' will overwhelmingly be that which is politically and ideologically advantageous to the interests of power.
The PM makes first-, second-and third-order predictions (see Klaehn, 2003b;2005a, 12-15). Its first-order predictions relate to overall patterns of media behaviour (see Chen et al., 2006;Cromwell, 2006;Eglin, 2005;Everton, 2005;Winter, 1992;Winter and Klaehn, 2005). Academic discussions of the PM typically begin and end with the five filter elements. The PM's overall scope, however, is significantly more far-reaching than standard treatments suggest. The model highlights how ideology, communicative power and media texts link to social organization, cultural education and pervasive social, political and economic inequalities (see Chomsky, 1997a;1997b;Winter, 2007). The model suggests that mainstream commercial media are structurally predisposed to endorse, legitimize and promote the interests of power (Chomsky, 1989;Klaehn, 2002a;2002b;2003a; also see Bagdikian, 1992;Barker, 2009;Carey, 1995;Edwards and Cromwell, 2005;Herman, 1999;Herman and McChesney, 1997). Herman (cited in Klaehn, 2008) explains that the PM's theoretical origins reside in 'the economic model of industrial organization that traces back to the great British economist Alfred Marshall', which: … assumed its more modern form at Harvard with Edward S. Mason and his student and later Berkeley academic Joe S. Bain. Our thinking was also influenced by pioneering media analysts whose ideas also flowed into our work: Warren Breed, Gaye Tuchman; Ben Bagdikian, Philip Elliott, Eric Barnouw, Peter Golding, Stuart Hall, Leon Sigal, and others. The PM argues that mainstream commercial mass media function as a 'guided market system' and explains why media discourses will reflect the interests of power within different time/place contexts.
According to Herman and Chomsky, ideological control will be most effective when there is consensus among elites, whereas disagreement among elites will create space for dissent (see Mullen, 2008). As noted above and highlighted elsewhere (see, most notably, Klaehn, 2003a), the PM does not predict that media are closed to dissent. This is a rather basic point that critics of the PM often miss, which is (by this point in time) at best perplexing, especially considering that Herman and Chomsky had explicitly addressed this point in the opening pages of the original version of Manufacturing Consent, stating that: 'The news media are not a solid monolith on all issues' (Herman and Chomsky, 1988, xiii). Herman and Chomsky have both written at length about dissent culture (see Chomsky, 1997a;Klaehn, 2002a). Herman (2000) has pointed out that critics of the PM often ignore the fact that the PM is actually 'about how the media work, not how effective they are'.
Numerous studies provide support for the view that media establish priorities, perpetuate events and mobilize public opinion in line with elite interests (see, for example, Babe, 2005;Barker, 2009;Chomsky, 1989;Edwards and Cromwell, 2005;Herman, 1992;Kellner, 1992;Klaehn, 2002b;Lee and Solomon, 1990;McChesney, 1999;McMurtry, 1998;Miller, 2003;Parenti, 1986;SourceWatch, 2009;Winter, 1992;Winter and Klaehn, 2005; see also Millar and Dinan, 2008). While many assume that media are dedicated to the broader interests of democracy and public/cultural education, Herman (cited in Klaehn, 2008) states that: … democracy and public education are not primary aims of the mainstream media; the former, if fully realized, might well be damaging to the ends of the powerful; the latter also, unless properly channeled and limited, could be injurious to the powerful. These incompatibilities are likely to increase if inequality grows and if a military ethos and culture become steadily more important (as they have). The mainstream media will respond with attacks on and marginalization of 'populism' with its equalitarian tendencies, and will normalize enormous military budgets and wars.
Central Methodological Techniques Associated with the PM While Herman and Chomsky have both written about the methodological techniques associated with the PM, they have not yet produced a single article or book chapter devoted exclusively to methodology. One would commonly be required to consult a wide range of articles, interviews and book chapters in order to delineate the full range of methodological techniques associated with the model. This section provides a comprehensive overview of the PM's methodology, outlining ways in which the model may be applied.
Rather than presuming that news discourses somehow exist within a vacuumapart from particular time/place contexts and various relevant dimensions of social life -the PM explores media discourses in relation to historical and contemporary political-economic contexts. The PM makes historical and politicaleconomic elements directly relevant to the whole enterprise of media analysis, regardless of the specific approach one might take in relating and applying the model.
One of the strengths of the PM is that it may be applied in a range of different ways, and it affords opportunities for multiplicity of focus. Fundamentally, the model suggests that how issues and topics are treated by media will be bound to the interests of power. The quantity and quality of coverage accorded certain news stories and events will differ accordingly.
Herman and Chomsky maintain that observable disparities in media treatment of co-occurring historical events, or paired examples, can enable critical insight into patterns of media behavior (see Klaehn, 2005a, 10-12). This methodological technique may be modified so as to enable insight into how media prioritize and treat similar cases/incidents which, while perhaps not co-occurring, may share common and/or contrasting core contextual elements. A 'case study approach' such as this could usefully be applied and would enable insight into how media function and socially construct news across different time/place contexts. It would also enable exploration of relatively recent news events.
Another methodological technique associated with the PM entails analysis of how victimized parties are represented within media discourses. The PM predicts that 'worthy victims' (victims of state terror enacted by official enemy states) will be accorded a significant volume of coverage and will be humanized within this coverage. News relating to victims of official enemy states will be constructed in ways that trigger an emotional response and mobilize public opinion and outrage. Conversely, the model predicts that 'unworthy victims' (victims of state terror undertaken by the United States, its allies and client states) will be accorded only slight detail and comparatively minimal humanization. The PM predicts that unworthy victims will be accorded treatment that does not prompt or elicit reader emotion, interest or sympathy (Herman and Chomsky, 1988, 35). The model's general argument, then, is that how victims and events are portrayed within mainstream media discourses will largely be dependent upon the interests of power within specific time/place contexts. This area of focus could also be significantly broadened and expanded to include events, issues and other groups of actors (beyond victims) (see, for example, Winter and Klaehn, 2005).
The model's focus upon how specific actors are represented within media discourses is relatable to both domestic and international events. Although the PM originated in the United States, a range of recent studies demonstrate that the model may be usefully applied to media discourses within other countries, where the media systems and political structures in play may be very different (see Edwards and Cromwell, 2005;Everton, 2005;Gunn, 1994;SourceWatch, 2009;Winter, 2007;Winter and Klaehn, 2005; see also Corner, 2003, 367;Klaehn, 2003b;2006a;2006b;2006c;2006d). This research indicates that the PM is international in overall scope and applicability. One may go further still and suggest that the model's analysis of victim representation should be explicitly linked to power.
The methodological technique most favoured by Chomsky in his political writings entails exploration of the 'boundaries of the expressible' or range of permitted opinion on crucial topics within media discourses. This involves exploring and assessing what particular facts, details and/or arguments are present within and absent from mainstream media discourses (see Chomsky, 1989, 59). The general argument here is consistent with the PM's overall conceptualization regarding patterns of media behaviour: that which conforms to the interests of power is permissible and that which may be threatening to the interests of power will typically falls outside the range of permitted opinion.
Applying the PM in the first instance, then, entails the quantitative and qualitative exploration of media discourses. Assessing the quantity of news coverage accorded particular cases and/or issues is relatively straightforward but can lend significant insight into how media prioritize and cover specific issues/events as opposed to others. The PM emphasizes qualitative exploration of the boundaries of debate. Sources, emphasis, placement, fullness of treatment, context, tone and evident range of debate on central issues and topics are observable dimensions of media discourses that may be qualitatively assessed when utilizing the PM to undertake detailed media analysis.
The PM assumes that news discourses are bound to power and predicts that the primary sources of news will be agents of power. The model may be utilized to explore how power pervades news discourses vis-à-vis source selection. Applying the PM entails consideration of the extent to which news narratives and debates are set within parameters that conform to the interests of power. Utilizing the PM to explore the extent to which official sources dominate mainstream media discourses can enable more nuanced and sophisticated understanding than mere quantitative content analysis. Testing the PM should entail qualitative exploration of the degree to which official sources are favoured within news discourses and the extent to which news narratives and discernible boundaries of debate within them are in effect defined by official sources.
Evaluating data for content, omissions and style of presentation should be undertaken in order to delineate the extent to which news discourses are ideologically inflected. Such analysis will involve assessing media choices regarding how news stories are framed and ways in which particular events, issues, topics and/or actors are represented within particular media texts. Assessing what is present within and absent from media discourses will connect in turn to source selection and boundaries of debate.
Another dimension of news discourses regarded as centrally important to the PM is the inclusion or absence of photographs within media texts. Since visual imagery can powerfully impact awareness and interpretation, the PM attaches particular importance to photographs.
Analysis of these elements can enable understanding on the part of the critical analyst relating to overall patterns of media behaviour. In its original incarnation, the PM stressed media choices relating to overall story selection and story treatment. Herman (cited in Klaehn, 2008) comments that: The overlaps with critical discourse frames are numerous, but this is because the subject is immense and many tracks can be followed that are often not inconsistent with one another but stress different things. We don't stress subtle language variations and/or the nuances in effects when the elite are split and a certain amount of dissent becomes permissible. Our emphasis is on the broader routes through which power affects media choices, how this feeds into media campaigns, and how it results in dichotomization and systematic double standards. The propaganda model focuses heavily on the institutional structure that lies behind news-making in 'a world of concentrated wealth and major conflicts of class interest.' This leaves lots of room for other tracks and subtracks in areas we deal with.
The PM and critical discourse analysis both suggest that analysis of textual prominence within media discourses can impart insight into how power and meaning intersect (see, most notably, Fairclough, 1989;1995a;1995b;; see also Giroux, 2001;Jhally, 2006;Lee and Solomon, 1990;Van Dijk, 2008). Such analysis involves careful consideration of a number of particular features of media texts and analysis of the ways in which media texts are structured and framed (see Van Dijk, 1988;Winter, 1992;. Analysis of various dimensions of media texts is often required and may be utilized in applying the PM. Headlines are crucially important given the significant roles that they play in impacting how readers understand and subsequently interpret news stories. Readers are conditioned to presume that the most important or 'newsworthy' information is conveyed at the outset of media texts. Because headlines and leads structure interpretation, they are of importance with regard to analyzing textual prominence. Of course the most central aspect of textual prominence is the quantity of news coverage accorded particular stories. Beyond this, fore-grounding and backgrounding are important considerations, as are questions of what is made explicit within media discourses and what is presented merely as incidental background fact or omitted altogether (see Fairclough, 1995a, 103-8). For the PM, omissions are of central importance and connect to the boundaries of debate. Media choices regarding sources, textual prominence, presuppositions, fore-grounding and backgrounding influence how new articles are understood and interpreted. Such choices create meanings and structure interpretative frameworks through presentation of content. Media discourses often include presuppositions about previously occurring events that also may be ideologically inflected.
The media analysis techniques set out by Herman and Chomsky in Manufacturing Consent and elsewhere enable sophisticated analysis of media discourses. The PM connects media and the ideological realm with social, political, economic dimensions of social life. The fact that the PM's explanatory power is not limited by geographic borders is a great strength of the model, in my view. The model is relatable to both international and domestic news events.

Critiques Associated with the Propaganda Model
Criticisms of the PM within the context of scholarly discourse have by this point in time been quite exhaustively rehearsed (Herman, 2000;Klaehn, 2002a;2003a;2005b;Mullen, 2008), yet misconceptions about the model continue to circulate widely: in textbooks, university departments, classrooms, and on the worldwide web. Herman (2000) comments that many of the initial critiques advanced against the PM 'displayed a barely-concealed anger, and in most of them the Propaganda Model was dismissed with a few superficial clichés (conspiratorial, simplistic, etc.), without fair presentation or subjecting it to the test of evidence'. On why the anger and hostility existed, Herman (cited in Klaehn, 2008) explains: The resistance and hostility to the propaganda model had several sources. One is that it is a radical critique, whose implication is that modest reforms that don't alter the structure very much aren't going to affect media performance very much. This is hard for non-radicals to swallow. Another source of resistance has been based on our relatively broad brush strokes with which we model a complex area. This makes it allegedly too mechanistic and at the same time lacking in a weighting of the elements in the model! But we don't claim that it explains everything and we are clear that elite differences and local factors (including features of individual media institutions) can influence media outcomes. We argue that the model works well in many important cases, and we await the offering of one that is superior. But we also acknowledge that there remains lots of room for media studies that do not rest on the propaganda model. This same room opens the way to criticizing the model for its failure to pursue those tracks and fill those spaces.
Critique that deliberately avoids the whole issue of evidence and factual support for the PM continues to persist (see Corner, 2003;Danesi, 2008, 44; see also Klaehn, 2003b;2005b) and (at least in Canada) it is extremely rare to find introductory textbooks within the fields of Sociology, Communication Studies or Cultural Studies that even mention the PM. When the PM is mentioned, often only the five filter elements are discussed. Sometimes critique of the model is also presented. Generally, the model is simply ignored.
As noted, recent research indicates that the PM has international resonance. Scholars from Canada, the US and Europe have demonstrated the model's applicability to a diverse range of issues and topics. The PM has been usefully employed to undertake analysis involving a broad range of international issues (such as the 'War on Terror' and the near-genocide in East Timor) and numerous domestic social and political issues (for a comprehensive listing of studies applying the PM, see SourceWatch, 2009).
It is certainly true that the PM argues, in the first instance, that media discourses are shaped by market forces. This is for the simple reason that ownership, size and profit orientation of dominant media impact significantly upon the contexts in which discourses are conceived and produced (see Bagdikian, 1992;Golding and Murdock, 1991;Lee and Solomon, 1990;McChesney, 1999;Murdock and Golding, 1977). The PM does not assume that media are monolithic, nor does the model ignore dissent. Chomsky has in fact described the media system as inherently unstable (see Chomsky, 1997a;Klaehn, 2003b;2005b, 231). While the PM describes a system in action, it makes no claims regarding how effective media may or may not be. Scholars have demonstrated, however, that media do have significant impacts in relation to public and cultural education (Ginsberg, 1986;Klaehn, 2003b;Lee and Solomon, 1990;Miller, 2003) and are also politically influential (see Chomsky, 1989;Cromwell, 2006;Herman, 1985;Klaehn, 2002b;2006b;2006d;Winter, 2002;. Generally speaking, criticisms of the PM fall into one or more of the following three categories: (1) critique motivated by political, ideological and sometimes even personal opposition to the model; (2) critique of the PM's assumptions, arguments and/or methodological approaches; and (3) critique centered around the fact that the PM is not all-encompassing and does not explain everything, in every context. Critique motivated by political, ideological and/or personal opposition to the PM may be implicit within (2) and (3). Such bias against the model may play out in different ways depending upon specific time/place contexts.
Some suggest that the PM casts media audiences as passive and easily manipulated, but (as the author has noted elsewhere) the PM actually makes no claims regarding audience effects, nor does it take media audiences or ruling-class interests for granted (for a detailed discussion, see Klaehn, 2003a;2005b, 227-33). It is concerned merely to delineate the extent to which media discourses are ideologically inflected in relation to structural dimensions of power and it affords ways of easily 'testing' the substantive predictions it advances. The PM does not assume that media audiences read texts in ways that are simple or passive (Chomsky, 1989, vii;Klaehn, 2002a;2003b;2005a;2005b;Rai, 1995, 53). Herman and Chomsky actually suggest that so much effort and expense are deployed on a daily basis by elites (vis-à-vis spin and PR) precisely because influencing public opinion is an ongoing process. Consent is typically negotiated and must be actively won and re-won (hegemony) over time.
Regarding claims that the Internet weakens and/or disproves the PM, Herman (2000) comments that: Some argue that the Internet and the new communication technologies are breaking the corporate stranglehold on journalism and opening an unprecedented era of interactive democratic media. There is no evidence to support this view as regards journalism and mass communication. In fact, one could argue that the new technologies are exacerbating the problem. They permit media firms to shrink staff even as they achieve greater outputs, and they make possible global distribution systems that reduce the number of media entities. Although the new technologies have great potential for democratic communication, there is little reason to expect the Internet to serve democratic ends if it is left to the market.
The PM is a democratic and critical model that engages directly with the question of how economic, political, social and communicative power intersect with discourse phenomena.
The model affords clear pathways for empirical investigation. Its emphasis on political-economy has also made it a target for critique. Herman (cited in Klaehn, 2008) comments that the PM: take the status quo as a given and that confine themselves to modest reforms are 'non-political.' This kind of critique is implicitly political.

Conclusion
This article has provided a succinct overview of the PM, rehearsed central theoretical and methodological considerations associated with the model and explored various criticisms that have been leveled against the model in the past. In conclusion it bears noting that the PM readily enables critical, empirical research and is well-suited to a multiplicity of domestic and international issues: the internment of Japanese Canadians during the Second World War, globalization and the political-economy of the sex trade, workers and unions within the broader context of social class and human rights issues, public relations (PR) and corporate spin, environmental issues, terrorism, war, domestic state/police violence, instances of institutional victimization, domestic poverty, political inequalities and much else. The range of topics the PM can theoretically be applied to is limited only by the creativity and imagination of the researcher. The existent scholarly work on the PM has, in my view, barely scratched the surface of the potential the model affords in enabling empirical research, which will in turn further understanding of how ideological power and meanings intersect with politicaleconomy and social class. The PM complements critical scholarship that is concerned to examine other significant dimensions of the communicative process that impact patterns of media behaviour, such as media production processes and spin and PR strategies (see Dinan and Miller, 2007;Miller, 2007;Miller and Dinan, 2008;Nelson, 1989;Sussman, 2007;Tuchman, 1978;Van Dijk, 2008;Winter, 1998;.