The Caribbean today still faces a number of old problems and is threatened by new and emerging ones. Such problems include: marketization and global capitalism, gender, health, sexuality, class, cultural, ethnic, colour and national difference, unjust social relations caused by these differences, poverty, inequality and unemployment (which may be a consequence of this difference) the imposition of knowledge, oppression, in-civility, corruption, organized violence hegemonic transnational institutions and their control of nation states and social actors, global surveillance, issues of privacy and the manipulation of identities through global information and communication technologies, and underdevelopment. Despite the many research projects to address these problems they still persist. For the most part, such projects have tended to focus on systems, forces of structure as units of analysis - limited access to resources, lack of political will, geo-politics, poor economic policies, weak social structures, cultural historicalness and so on. Fairclough et al, (2004) and in many ways Bourdieu, and Wacquant, (2001) as well as Jessop, (2000) and Held et al. (1999) have elsewhere argued that problems of such nature are as a result of how things are represented or framed in text/language and how power exists in and over text/language. Fairclough (2003) however suggests that these texts, operates conjointly with vocal and visual elements (depiction, gesture, graphics, typography), in the context of meaning-laden architects, with the semiotics of action itself, and with music or other extra-linguistic auditory signs (Fairclough et al. 2004: p. 5) which taken together represents discourse. In other words what Fairclough is saying here is that these problems are problems of discourse.

Such texts include, among others: magazines such as Time, Newsweek, Vogue; newspapers; transcripts of spoken conversations and interviews, regulations in books and guidelines/recommendations in speeches, contracts, rules, accords, memoranda of understandings, resolutions, agreements in text/documentations and visual images from interaction, conversations with specific vocabularies and terminologies about ordering and patterns as well as influential sounds and other images from televisions, radios, internet websites, broadcasts and text messages. Some of the discourses associated with these texts include specific styles, ways of acting, representing and organizing, as well as specific identities associated with their (the texts) production, distribution, transformation and consumption.

Following Fairclough (1992; 2003), Bourdieu, and Wacquant, (2001) as well as Jessop, (2000) and Held et al. (1999), I argue that such texts and their associated discourses (together discourses) play an important role in regulating social life. These 'discourses of regulation' are very often constructed by hegemonic monopolistic and/or quasi-monopolistic agents (Microsoft and UNDP is an example of this - See Microsoft, 2004). These bodies through their knowledge and power use projects of normalization and mystification to influence behaviour as well as enact their own agendas and interests which very often translated into projects of control. Indeed such projects have been popular over the years and represent a continuation of the spread of the 'spirit of capitalism' and the drive to preserve the status quo. Today however their power have multiplied manifold with the advent of globalization and particularly information and communication technologies (ICTs).

To understand social life today thus requires a proper deconstruction of such texts and their associated discourses, these mechanisms of control. Such a projects Derrida (1971) tells us will help researchers to deconstruct and better understand the cosmologies of social life, what they encourage, how the interests of a few are manifested in them, how some discourses are interpreted by many and how others evoke acts of resistance

Ostensibly speaking, based on a review of several leading social scientific related journal publications which were published in the Caribbean during this millennium (Specifically Social and Economic Studies and the Journal of Eastern Caribbean Studies) it is possible to argue that many social scientists in the region have not fully recognized such problems as moments of language and their associated discourse (together discourse). There are several different factors which, either taken together, or individually, may explain this oversight. Some of these may include: the view that language and discourse are elements of the discipline of Linguistic and the perception that such units of analysis are complex to deconstruct without a background in linguistics; a possible disconnect between the humanities and the social sciences; the fustian-type discussions of those championing the analysis of discourse and/or the multidimensional and complex process which they recommend for analyzing discourse. In addition to these, Fairclough et.al (2004) has also suggested that:

…there is also a widespread suspicion of discourse analysis amongst social scientists, a perception that it is often vague and ill-defined, supported by the manifold definitions of discourse in social theory (for example in Foucault as opposed to Habermas), in different national academic traditions (for example Germany as opposed to Britain and the USA), as well as in various areas of language study (for example pragmatics, text linguistics, as well as discourse analysis itself). Many social research papers identify discourses in whatever material they are analyzing without giving much indication of what particular features characterize a particular discourse and help us to recognize its presence, or the grounds for claiming that there are different discourses, or for distinguishing three rather than, say, five discourses in a given context. Another cause for suspicion is the assumption, correct in a few instances but incorrect for most critical discourse analysis, that discourse analysts reduce the whole of social life to discourse, leaving no space for analysis of the material world or social structures (p. 3).

There have however been a few social scientists that, despite these challenges, have focussed on discourse as a unit of analysis (See for example Dann, 2004). This select few have not demonstrated in any way a real appreciation for systematically analyzing the discourses of Caribbean social life. Certainly this leaves them open to criticisms from positivists, postpositivists and even interpretivist colleagues regarding rigour, trustworthiness and substantiveness. This has wide implications for the acceptance of findings as well as policy recommendations.

This theoretical article is pedagogical: its aim is to highlight and draw attention to: (1) the importance of text (language) and discourse as units of analysis for analyzing Caribbean social life and understanding Caribbean social problems; (2) the text(language).discourse linkages and differences and the relationship between text(language) and social practice (the oscillation between social relations, events, structure, action and agency and (3) one possible methodology and analytical tool for systematically analyzing Caribbean social life and understanding Caribbean social problems - Fairclough's Critical Discourse Analysis methodology.

The Text(Language) and Discourse Link and Discourse as a Unit of Analysis

Texts, whether written or spoken or representation of language in images, movements or sounds, are the "key raw materials out of which specific discourses, understood as bounded (sometimes strictly so) ways of representing the world, get shaped" (Fairclough, 2003: p. 2) and can be understood. Texts thus provide us with insights into language. According to Fairclough, (2003), "language is an irreducible part of social life, dialectically interconnected with other elements of social life, so that social analysis and research always has to take into account language" (p. 2) or how language is represented in text. Language constructs and is constructed by social relations, events, structure, action and agency. Language also provides a good description of structures, events, social practices, social networks, and relations between and among people, between and among institutions, between institutions and people. Thus language provides us insights into discourse.

There are many different definitions of discourse. For example Habermas (1984) sees it as a form of communicative action which is an expression of rationality. Macdonell (1986) defines it as a process of social exchange of language which is organized around rules and regulations involving social intercourse. Harvey (1996) explains that it is a combination of languages bounded together to represent the world. Gee (1999) notes that it is the use of language to socially identify and position oneself. And Parker (1992) represents it as "a system of statements which constructs an object" (p.5). But it is Fairclough's definition of discourse, which extends the work of Foucault (1972, 1977, 1980) that I wish to use in this article and promote. According to Fairclough (1989, 1992, 1995a, 1995b, 1995c, 2003) and Chouliaraki and Fairclough, (1999) discourse is a way of constituting knowledge about a particular topic at a historical moment through language in speech and text or images and sounds, or representations in postures, movements and structures which shape or are shaped by institutions, situations and structures. Language in speech and text or images and sounds, or representations in postures, movements and structures are constituted by discourse, and these discourses help to identify a subject's characteristics and possibility. Discourse provides an image about the reality of the subject (individual, group, institution), and how meanings are constructed in certain situations. Thus what language (body language, spoken language, written language and environmental language) really does is that it gives us insights into discourse.

For Fairclough, discourse also includes 'other elements' sometimes referred to as 'social practices', or 'moments' of social life or extra-discursive elements (Fairclough, 1995a, 1995b, 1995c, 2003; Harvey, 1996). Moments, social practices or extra-discursive elements (and I use these terms interchangeably) are relatively stabilized forms of social actions. The literature identifies two different types of moments. These are (a) discursive practices and (b) socio-cultural practices. Discursive practices includes action and interaction, social relations, the material world, material practices, as well as the rituals, beliefs, attitudes, values, desires of people and institutions (Chouliaraki and Fairclough, 1999: p. 28). It also include power and discourse (Harvey, 1996: p. 78), as well as forms of consciousness, time and space, objects, instruments, subjects and their social relations and activities (Fairclough, 2001: p. 1; Fairclough, 2003: p. 205). It can also include abstract social structures, concrete social events as well as relatively stabilized forms of social activity (Fairclough, 2004:p. 5).

Socio-cultural practices are the wider socio-cultural, political, ideological and institutional structures and processes in a historical context (Fairclough, 2003; See also Weiss and Wodak, 2002). The concept of social practice allows one to capture the changeability and interactive flow between social structure, as well as social action and agency and the role of discourse in this context. Moments, social practices or extra-discursive elements are the oscillation between the perspectives of social structure, action and agency.

Discourse internalizes these other moments without them being reducible to each other (Fairclough, 2003). In other words, these elements are dialectically related in that they are the active entanglement of relations, interactive discussions, contradictions, permutations, difference combinations, interactions, argumentation, reasoning and reactions which is likened to the process of exchanging propositions (thesis) and counter-propositions (antithesis) to produce synthesis - a socially constructed truth (either someone's truth, an agreed-upon truth, an abstraction of the truth or an imposed truth). Therefore analyzing discourse requires a form of 'dialectical thinking' (Harvey, 1996: p. 49). One needs to understand the processes, flows, fluxes, circulatory framework, and relations over the analysis of elements, things, structures and organized systems (p. 49) and how, why and in what way they were constructed. One needs to also have an understanding that "elements, things, structures, and systems do not exist outside of or prior to the processes, flows, and relations that create, sustain, or undermine them" (p. 49).

According to Fairclough (2001), discourse "features in broadly three ways of social practices" (p. 2): as genres (ways of acting and organizing or action or ways of relating) - interacting through speaking and writing; as discourses (ways of representing or representation) - as particular ways of representing the world, and as styles (ways of being, ones identity) - "particular social or personal identities" (Fairclough, 2003: p. 26). Discourses are considered as a combination of text, event, the wider physical and social world, and to "the persons involved in the event" (p. 27). This relationship between ways of representing, ways of acting and ways of being, within the context of social practices, is a dialectical one. Fairclough (2001) attempts to illustrate the link between these three elements in his 2001 paper The Dialectics of Discourse (and more comprehensively in his 2003 publication Analyzing Discourse: Textual analysis for social research). His arguments are worth quoting in length:

Discourses include representations of how things are and have been, as well as imaginaries - representations of how things might or could or should be….In terms of the concept of social practice, they imagine possible social practices and networks of social practices - possible syntheses of activities, subjects, social relations, instruments, objects, space-time,…values, forms of consciousness. These imaginaries may be enacted as actual (networks of) practices - imagined activities, subjects, social relations etc can become real activities, subjects, social relations etc. Such enactments include materializations of discourses - economic discourses become materialized for instance in the instruments of economic production, including the 'hardware' (plant, machinery, etc) and the 'software' (management systems, etc). Such enactments are also in part themselves discoursal/semiotic: discourses become enacted as genres….Discourses as imaginaries may also come to be inculcated as new ways of being, new identities….Inculcation is a matter of, in the current jargon, people coming to 'own' discourses, to position themselves inside them, to act and think and talk and see themselves in terms of new discourses…. Inculcation also has its material aspects: discourses are dialectically inculcated not only in styles, ways of using language, they are also materialized in bodies, postures, gestures, ways of moving, and so forth.

The dialectical process does not end with enactment and inculcation. Social life is reflexive. That is, people not only act and interact within networks of social practices, they also interpret and represent to themselves and each other what they do, and these interpretations and representations shape and reshape what they do. (p. 2-3).[1]

What Fairclough (2003) suggests here is it is possible to argue that particular 'ways of representing' social life (discourses) may in many ways be enacted in specific 'ways of acting' (genres or social practices), and inculcated in certain 'ways of being' (certain styles). They can be seen together in texts as what brings subject, objects and action - the cosmologies of a phenomenon - to the fore. Thus taken together discourse, genre, styles and social practices model the structure/agency/action link and a systematic analysis of text can capture this constellation. According to Fairclough, focusing on discourse can provide inroads into these other moments of social life and how they are arranged around a social phenomenon - their cause and consequences.

The notion of social practices suggests that the entire world is thus not reduced to discourse as some scholars critical of discourse analysis techniques have argued (see for example Widdowson, 1995). Rather discourse must be understood dialectically in relations to the other moments (Fairclough, 2003). This dialectical process can provide an understanding of these other moments as they (the other moments) all have some form of discursive property: and thus provides a researcher with insights into the discursive processes and the dynamics of a phenomenon.

Fairclough (2003) has also argued that discourse can have various meanings and represent a multiplicity of subjects and objects based on the audience, environment, history, position of the producer, and the recipient. Through a communicative process, discourse can position and label people in different ways (Foucault 1972, 1977, 1980; Weiss and Wodak, 2002; Harvey, 1996; Macdonell, 1986; Purvis and Hunt, 1993; Fraser, 1997; Fairclough, 1989, 1992, 1995a, 1995b, 1995c, 2003; Hall, 2000). For example discourse can define the roles of social actors - as someone who can (or cannot) bring about a change in a condition or as someone who is (or is not) knowledgeable. It can also define their status, for example, as someone who needs (or does not need) help, as someone who cannot (or can) understand complex things or as someone who is (or is not) in the lower-class, middle-class, or upper-class. Discourse can also be used to define the significance or existence of social objects. Discourse can thus be viewed as socially constitutive of, as well as socially conditioned from constituted objects, subjects, processes, events and phenomena - it is constitutive both in the sense that it helps to sustain and reproduce the social status quo, and in the sense that it contributes to transforming it (Fairclough and Wodak, 1997: p. 258).

In many ways then, it may be possible to argue that discourse is normalized by, and normalizes social practice. Many people act and organize through and around discourses in specific ways. Such action helps to shape and create discourses while at the same being shaped by discourse. They (the action) represent ways of acting and organizing, and produce and are produced by imaginary projections of new or alternative ways of acting and organizing (Fairclough et al, 2004). For example, terms such as 'the knowledge society', 'bridging the digital divide', and 'ICT for development' may be constructed as discourses that specify ways of (inter)acting which become modes of operation. At the same time ways of inter(acting) help to shape and redefine the notion of 'the knowledge society'.

The ability of a discourse to influence social practice or vica versa is dependent on the dynamics of power. Thus having an understanding of the power relations of a discourse is indeed important. Foucault, who does not produce a dialectical version of discourse, argues that the power to define terms determines the outcome of a discourse. For example, the power of an institution, a group or an individual can influence social actions and relations. According to Titscher et al, (2000) (extending Foucault's arguments to a dialectical moment) "[p]ower relations have to do with discourse….Society and culture are dialectically related to discourse: society and culture are shaped by discourse, and at the same time constitute discourse" (p. 148).

Analyzing Discourse

In the last decade or so, many different tools have been developed to analyze discourse (See in Weiss and Wodak, 2002; Titscher, et. al, 2000). Because of this diversity Weiss and Wodak, (2002) have correctly asserted that studies in analyzing discourse are multifarious, derived from quite different theoretical backgrounds and oriented towards very different data and methodologies as well as epistemological influences. A taxonomy of these different tools suggests that there are two different poles. For example, on the one hand there is a focus on a detailed analysis of texts - the linguistic features of texts. Such an approach may be considered to be narrow. On the other hand there is a focus on the social aspects of text production, transformation, distribution, consumption and redistribution - focussing only on discourse. Such an approach may be regarded as wide. Within these two poles also exists other poles - a normative and a critical approach. The former is an attempt to understand the configurations of a discourse while the later is specific to deconstructing hegemonic relations of power in and over discourse and how this undermines social justice.

In this article I wish to focus on an approach by Norman Fairclough - Critical Discourse Analysis (Fairclough, (1989, 1992, 1995a, 1995b, 1995c, 2003), Chouliaraki and Fairclough, (1999) or CDA. Such an approach attempts to transcend these vertical and horizontal divisions. In other words such an approach "bridges texts, the micro-level of face-to-fact interaction, and the more macro-level social, political and economic process" (Bloome, and Talwalkar, 1997: p.112). It is also an attempt to merge the normative and critical approach to text and discourse analysis. Fairclough's CDA also provides "a theory-method linkage that is absent in many sociological discussion of everyday life and language use and in many linguistic discussions of social dynamics (Bloome and Talwalkar, 1997: p. 105). Fairclough describes his approach:

On the one hand, any analysis of texts which aims to be significant in social scientific terms has to connect with the theoretical questions about discourse (e.g. the socially 'constructive' effects of discourse). On the other hand, no real understanding of the social effects of discourse is possible without looking closely at what happens when people talk or write (Fairclough, 2003: p. 3).

Fairclough (2003) further states that "text analysis is an essential part of discourse analysis…discourse analysis is not merely the linguistic analysis of text" (p. 3) rather discourse analysis oscillates between the focus on specific texts and the focus on the relatively durable social structuring of language which in many ways is itself merely one of many elements of a relatively durable structuring and networking of extra-discursive practices - social practices and sociocultural practices. In other words it is the slippage between content/description of text their creation/production, distribution, projected (how texts are represented in terms of the different discourses, genres, and styles they draw upon and articulate together), transformation and their receipt - discursive processes. Thus for Fairclough, texts are connected to broader social structures, events, actions, processes and practices.

This approach to analyzing discourse and text first made its début in 1992 with the publication of Discourse and Social Change. In this text Fairclough had introduced various critical approaches to discourse analysis which he consolidated in 1995 with the publication of Critical Discourse Analysis. His approach was made more robust in a joint project with Lilie Chouliaraki in a 1999 publication Discourse in Late Modernity - Rethinking Critical Discourse Analysis. Here both himself and Chouliaraki outlined a more focused and substantive approach to the critical analysis of discourse which was further refined over the years through approximately 20 publications influenced by the application of the methodology within the discipline of government and politics as well as media. His most recent publication Analyzing discourse: textual analysis for social research is a culmination of more than a decade of theorizing about, experimenting with, and developing methodologies for critically analyzing discourse.

Today, Fairclough's CDA has been defined as fundamentally interested in analyzing opaque as well as transparent structural relationships of dominance, discrimination, power and control manifested in language (Weiss and Wodak, 2002). It is an approach to deconstructing society which aims to investigate critically social inequality as it is expressed, constituted, legitimized, and so on, by language use (or in discourse). (p. 15). According to Chouliaraki and Fairclough, (1999) the approach emerged from Critical Theory as a method for accomplishing the critical agenda, an aspect of Fairclough's CDA which sets it apart from other approaches to the analysis of discourse.

Critical theory is considered as any approach which seeks to "upset institutions and threatens to overturn sovereign regimes of truth" (Kincheloe and McLaren, 2005: p. 433). Critical Theory, Agger (1991) explains, was developed by the Frankfurt School to "explain why the socialist revolution prophesised by Marx…did not occur as expected" (p. 107) and "the changing nature of capitalism" (Kincheloe and McLaren, 2005: p. 434).According to Kincheloe and McLaren, (2005), Critical Theory:

…analyses competing power interests between groups and individuals within a society - identifying who gains and who loses in specific situations. Privileged groups, criticalists argue, often have an interest in supporting the status quo to protect their advantages; the dynamics of such efforts often become a central focus of critical research…critical research attempts to expose the forces that prevent individuals and groups from shaping the decisions that crucially affect their lives (Kincheloe and McLaren, 2005: p. 437).

For Alvesson, and Deetz, (2000) the agenda of critical theory therefore is:

1. Identifying and challenging assumptions behind ordinary ways of perceiving, conceiving and acting;
2. recognizing the influence of history, culture, and social positioning on beliefs and actions;
3. imaging and exploring extraordinary alternatives, ones that may disrupt routines and established orders;
4. being appropriately sceptical about any knowledge or solution that claims to be the only truth or alternative (Alvesson, and Deetz, 2000: p. 8).

Critical theory is thus an emancipatory process which is committed to engaging oppressed groups and undertake democratic theorizing about what is common or different about "their experiences of oppression and privilege….A constant focus is given to the material and cultural practices that create structures of oppression" (Denzin, 1998: p. 332). It is an approach which gives the oppressed a space to speak, to tell their story. This process is dialogic and collaborative as the researcher assists the oppressed through data analysis and interpretation.

Fairclough's CDA extends Critical Theory by connecting critical theory to an understanding of the ways in which people are unequally positioned through an analysis of discourse, how people socially construct the meanings of objects and subject (and the influences behind these constructions) in the production and consumption of language in spoken and written form. In other words, a critical analysis of discourse explores the connections between the use of language and the social, historical and political contexts in which it occurs, how language is used in social interactions and how language influences social relations and practices. It also extends Foucault's (1972, 1977, 1980) project in unravelling power relations through the analysis of "competing power interests between groups and individuals" (Kincheloe and McLaren, 2003: p. 437) by analyzing the dialectical relationships between "discourse …and other elements of social practice" (Fairclough, 2003: p. 205) (locating power in discourse and power over discourse in a historical, socio-cultural and political context). In addition to the influences of Foucault, the Fairclough's CDA model also builds on the works of other social theorists such as Antonio Gramsci, Louis Althusser, and Jürgen Habermas as well as literary theorists such as Michael Halliday, Michel Pecheux, Gunther Kress and Mikhail Bakhtin, (Fairclough, 1992; Chouliaraki and Fairclough, 1999). This interdiscliplinary/transdiscliplinary (Chouliaraki and Fairclough, 1999: p. 16) aspect of CDA positions the methodology as a theory formulation instrument which synthesizes both linguistic and sociological conceptual tools so as to better analyze and explain the complex interactions, and interrelations between and among the communicators of social life (Weiss and Wodak, 2002: p. 7). It positions the researcher to:

…systematically explore often opaque relationships of causality and determination between (a) discursive practices, events and texts, and (b) wider social and cultural structures, relations and processes; to investigate how such practices, events and texts arise out of and are ideologically shaped by relations of power and struggles over power; and to explore how the opacity of these relationships between discourse and society is itself a factor securing power and hegemony (Fairclough, 2003: p. 135)

According to Bloome, and Talwalkar (1997), this is based on Fairclough's assumption that:

…people use language to accomplish a variety of social goals; he also assumes that any analysis of language must be wedded to a social theory that encompasses both every social practices… and social institutions in which they occur…, as well as the broader ideological context…issues of power and control are central to the social theories that Fairclough builds upon (Bloome, and Talwalkar, 1997: p. 107).

CDA thus attempts to locate, describe, understand, interpret and explain these nuances of unequal relations of power as well as those of dominance, hegemony, marginalization, exploitation and/or inequality (and other forms of social injustices) in or over discourse practice through an analysis of language. It seeks to understand and explain how these social injustices are initiated, hidden, transformed, reproduced, and legitimized, the agencies that generate, normalise, mystify, alter or change them, and their links to wider socio-cultural, political, historical, ideological and institutional contexts in an attempt to introduce to, and make humanity aware of, the influences of discourse and thus encourage participation in processes leading to positive change. This is accomplished by detecting struggles between the strategies of change of different groups of social agents.

This processes of detecting is done through three different types of critique tied to the agenda of Critical Theory- strategic, ideological and rhetorical critique. He has defined these:

Ideological critique: critique of how a system of social relations is sustained through representations of a social order which are in contradiction with its realities.
Rhetorical critique: critique of the subordination of considerations of truth and sound argumentation to the will to persuade.
Strategic critique: critique of how discourse figures in the development, promotion and dissemination of the strategies for social change of particular groups of social agents, and in hegemonic struggle between strategies, and in the implementation of successful strategies (Fairclough, 2004:p. 7).

In critiquing, CDA often targets dominant groups or "power elites that sustain social inequality and injustice" (Weiss and Wodak, 2002: p. 38) through social relations. CDA studies their potential for reproducing social structures and the consequence for social inequality. Such social power relations are based on preferential access to or control over scarce social resources by the dominant group. Certainly this has always been the concern of Caribbean social scientists. These resources are not only material, but also symbolic and knowledge related. They include access to public discourse which Weiss and Wodak, (2002) argue is "among the major symbolic power resources of contemporary society" (p. 87). However such forces have largely been downplayed in the Caribbean social sciences. CDA puts the voice of the power elite into question to reveal hidden needs, interests and agendas, demonstrating how they use knowledge in public discourse to control and dominate social relations that serve self interests, maintaining social inequality and injustice. CDA makes the marginalized aware of the dynamics of their social situation, position and circumstances and helps to make their voice legitimate and heard. This is done by uncovering ideological assumptions and clarifying the connections between the use of language, ideologies and the exercise of power/knowledge through a systematic exploration of the relationships between discursive practices, texts, social, socio-cultural structures, institutions, ideologies, and associated processes. According to Chouliaraki and Fairclough (1999) CDA is inherently beneficial in theorizing transformations by creating an awareness of "what is, how it has come to be, and what it might become, on the basis of which people may be able to make and remake their lives" (p. 4). CDA then contributes to making a positive change to the lives of the less privileged and is thus well suited for this research project.

Using Fairclough's CDA to systematically analyze discourse

Fairclough offers his approach to the critical analysis of discourse as an ideal tool for the analysis of language and discourse systematically. This has been based on a decade of work on refining his approach. As mentioned earlier, for Fairclough, analyzing discourse is a synergy of the aforementioned three levels of analysis, that is (1) the actual text (which he describes as discourse-as-text) (2) producing, distributing, transforming and consuming the texts (which he describes as discourse-as-discursive practice) and (3) the larger social context which may have influenced the creation of the texts (which he describes as discourse-as-social-practice) - which takes into consideration the underlining socio-cultural and power structures in society. Figure 1 presents a graphical representation of this process, a process which Fairclough states is a continual movement back and forth among the various levels of analysis - descriptive, interpretive and explanatory:

To fully understand this model, a detailed discussion is necessary. At the text level, there is a focus on describing the texts (language in text) - their discursive content - as well as how these texts are linked to other discourses, genres and styles. Fairclough refers to this process as intertextuality. Essentially at this level, the researcher looks at what is represented in the text. Here the analysis is descriptive and in many ways, can be described as a form of linguistic analysis of texts in that texts are analyzed by looking at vocabularies (wording and metaphors), semantics, utterances, grammar (transitivity, modality) to identify "representations, categories of participant, constructions of participant identity or participant relations" (Fairclough, 1995: p. 58) of subjects, objects, social positions, how subjects and objects were positioned, and instances of relations of power in the use of language. It is also important to look at "collocations, patterns of co-occurrences of words in text, simply looking at which other words most frequently precede and follow any word which is in focus (Fairclough, 2003:p. 131). Also of specific interest at this level are the genres to which specific discourses belonged, whether the texts conform to that particular genre, the intertextual linkages, semantic relationships, how elements of social events (processes, people, objects, means, times, places) are represented (Fairclough, 2003: p. 133), the absence or inclusion of specific characteristics in the genre, generalizations, how events were ordered, the angle which is being taken, what is emphasized (foregrounding) what is not (backgrounding) and, among other things, what were the main assumptions and presuppositions of the generic and discursive configurations of text.

The analysis then expands to the discursive practice level - discourse(s)-as-discursive practice the researcher analyses what are the factors which influences how social actors interpret an event and how this process influences the production, distribution, transformation and consumption of texts. At this level, an interpretation of discursive practices in relation to events, inter-discursivity (how different discourses are related), orders of discourse (a network of a socially ordered set of genres, social practices and discourses associated with a particular social field) and the power relations between people in an event is undertaken. It is at this level that moments such as action, interaction, social relations, the material world, material practices, the rituals, beliefs, attitudes, values, desires of people and institutions, power and discourse, forms of consciousness, time and space, objects, instruments, subjects and their social relations, activities as well as abstract social structures and concrete social events is taken into consideration in relation to text production, distribution, transformation and consumption.

Finally, understanding the wider socio-cultural, political, ideological, institutional and historical context and structures surrounding the text and their associated discourses - sociocultural practice - is an important activity in explicating the dynamics of a text and how the text is produced, distributed, transformed and consumed. Here the analysis is at the explanatory level. At this level one also takes into consideration the underlining power relations which might be reproduced, how they facilitate the exploitation and marginalization of groups as well as possibilities of change and resistance. The process also helps a researcher to identify, understand and explain the causal and circular logic at work - how and why powerful discourses and powerful agentic forces "shape beliefs, fantasies and desires so as to regulate practices of institution building that set the stage for material production and reproduction activities that in turn construct social relations that finally return to ensure the perpetuation of power" (Harvey, 1996: p. 82).

Fairclough's CDA - Problems and Solutions: A Bricoleurian Approach to building Trustworthiness and Rigor

In concluding this article it is important to note that the CDA methodology has been heavily criticized for being too interpretive and subjective with little room for objectivity (see Widdowson, 1998). In other words it is too political and bias-prone. For contemporary qualitative researchers objectivity is questionable and the multiplicity in truths which politics brings to a research is welcomed. What is really of importance for qualitative methodologies are issues of trustworthiness and rigour. Ostensibly speaking Fairclough's CDA in its existing form is indeed questionable in this regard. I have thus attempted to minimize this problem by juxtaposing Fairclough's CDA with other methodological and data analytical tools which I believe makes his approach a more trustworthy and rigorous approach.

This juxtaposition is made possible by what Fairclough describes as the fluidity of his CDA approach. According to him like many other qualitative methodologies one of the strengths of his approach to analyzing discourse is that it may be used in conjunction with other methodologies such as ethnography. In my use of this tool (Fairclough's CDA approach) to analyze the discourses surrounding an ICT for development initiative in Jamaica for microenterprise entrepreneurs operating in the tourism industry I have found this to be true. For example I was able to juxtapose Fairclough's CDA with various elements of Strauss, and Corbin's, (1990, 1998) Constant Comparative Analysis of Grounded Theory, Yin's (1994) Case Study method approach, and Miles and Huberman's (1994) Matrix Technique. Ostensibly speaking, this was a quintessential 'bricoleurian' approach (Denzin and Lincoln, 2005) in that it was an "emergent construction that changes and takes new forms as the bricoleur adds different tools, methods, techniques of representation and interpretation to the puzzle" (p. 4) - the mixing of methodologies and methods of data analysis (triangulation) to suit a particular research agenda.

The Matrix approach of Miles and Huberman for example can be used in helping to simplify the analysis of data collected for a particular purpose. With matrices a researcher is able to sort data into increasingly manageable display and also helps in avoiding aggregation - they keep the data distinct. Matrices can be grouped by cases and Yin's Case Study approach is excellent in this regard. This approach can be used to unpack the matrices to construct cases tied to the unit of analysis at various levels of abstractions - cases as individuals, as events, actions, structures, relations and so on.

The nature of the case study method allows a researcher to focus on specific events to reconstruct the details of these proceedings and activities as they occur over time and in space. In a sense then, the case-study approach is like an examination of halted reality in putting together events and activities. It allows snapshots of moments in an ongoing process which, thereafter, the researcher interprets with the aim of explaining the phenomenon under investigation. In less abstract terms, with the case study approach, once the facts of an event or an activity is collected, they were then explored, examined and compared with other similar situational occurrences to draw out these and other conclusions. The analysis of the dynamics of each case contributes to an understanding of the comparative (similar/dissimilar) dynamics of the cases. This process provides a means of clustering, sorting, identifying and isolating cases, and reducing them into themes.

Once these themes emerge, they can then be further grouped/reduced into higher order themes - higher level of abstractions of discursive themes. At this level the researcher can begin to map the characteristics and configurations of these discursive themes (broader discourses) with the aim of understanding their relationships with, linkages to, dynamics of, and effects on individual themes and cases specifically in relation to the orders of discourse, social practices, and genres. At this level both discourse-as-discourse practice and discourse-as-sociocultural practice are taken into consideration in the construction of more definitive discursive themes, and explore case configurations to be able to see possible linkages (network of relations) across themes and discursive themes.

The researcher now can attempt to identify shared meanings in discourse, as well as common ways of socially constructing objects, subjects, processes and phenomena as well as possible differences (isolated for more deconstructive analysis) which would be outlined in texts thereby capturing analysis and interpretation from which discursive themes could be further refined.

In this data analysis/interpretive process researchers should be mindful of their own discourses. This is indeed important given the qualitative nature of CDA and the criticisms that the tool is biased as it focusing on the interpretation (influenced by presuppositions) and not on the data. But Miles and Huberman (1994) remind us that bias is inescapable for the social research: According to them:

We all have our preconceptions, our pet theories about what is happening. The risk is taking them for granted, imposing these willy-nilly, missing the inductive grounding that is needed. We do not hold as strongly as Glaser (1992) that all predefined concepts should be abjured in favour of solely 'grounded theory' emergence" (p. 208

Certainly using CDA requires some form of reflection. The researcher needs to keep an open mind as regards to what might emerge from the data. The researcher will need to ground his or her analyses of the data while generating meaning - interpreting, describing and explaining. The researcher will need to attempt to make conceptual/theoretical coherence, through the process of constant analytical comparisons (seeing and understanding themes, patterns, relationships, associations and causal networks with causal connections drawn from multiple analysis) while undertaking data reduction and pattern coding, and making as well as understanding contrasts and comparisons of degrees of effect among the cases and themes - as discussed above.

This can further lead to which I refer to as purifying discursive themes. And this is indeed possible if one uses Strauss and Corbin's constant comparative analysis of the grounded theory. With this tool the research further groups, orders, amalgamates/clusters cases, themes and discursive themes and expands others. Secondary data can be consulted to compare and contrast causal connections, disconnections, and constructs which may conflict with, as well as support the data. This process can lead to the shifting from concepts to attributes to theories which account for "the how and why of the phenomena under study" (Miles and Huberman, 1994: p. 261). In other words at this juncture tentative conclusions, explanations about a phenomenon (theories), and/or hypotheses can emerge.

As mentioned earlier, this approach was used in an interdiscliplinary qualitative study I had undertaken to analyze an Information and Communication Technology (ICT) for livelihood development initiative in Jamaica, introduced by the United Nations Development Programme - the Jamaica Sustainable Development Networking Programme (JSDNP). The primary objective of this initiative is to provide the poor in Jamaican communities with access to, and training in ICTs with the aim of improving their livelihoods. In this research, I specifically focus on the implications of the discourses surrounding the JSDNP Cybercentre Project for a group of microenterprise entrepreneurs in the Jamaican tourism industry to access the epistemological assumptions of this project.

Through the use this approach it was found that (at one level) the JSDNP Cybercentre initiative discourage the indigenization of non-indigenous technologies, represent the achievement of livelihood development through the use of specific commercial technologies while marginalizing and/or excluding other more democratic (fluid/flexible) ones. For example it was found that the inability or ability of microenterprise entrepreneurs operating in the Jamaican tourism sector to benefit from ICTs is not only a result of their lack of access, infrastructure, training and resources but also how the objects, subjects and processes which were required to benefit from these ICTs were represented - the particular discourses which were encouraged.

Subsequently, on one hand, these discourses were more favourable to the operational and structural dynamics of 'non-poor' microenterprise entrepreneurs. On the other hand however the discourses were incompatible with the operational and structural configurations of those microenterprise entrepreneurs working in the Jamaican tourism industry who had represented themselves as poor. What this had suggested is that the discourses play a fundamental role in perpetuating already existing entrenched inequalities through the preservation of social practices, along with their associated systems and structures.

At another level it was also found that these modalities the discourses promoted limited the operational processes of all microenterprise entrepreneurs who were exposed to the Cybercentre Project - both poor and non-poor. These entrepreneurs have limited control over the configuration of non-indigenous technologies; their technological and creative capabilities are restricted; their ability to indigenize non-indigenous technologies impaired; and they are highly dependent on non-indigenous technologies (which themselves have a number of limitations). All this significantly undermined their ability to achieve livelihood expansion.

A Final Note

It is hoped that this article has demonstrated the importance of deconstructing discourse, the discourse/text links, and the utility of Fairclough's CDA model as well its fluidic characteristics. Given the power dynamics within and among Caribbean societies I believe that Fairclough's CDA model would be an appropriate methodology for scholars in the region trying to understand and explain Caribbean social life. Thus it is also hoped that this article has peaked the interest of Caribbean scholars. It is important for readers to also note that this article is not only relevant to Caribbean social scientists but may be of importance to other social scientists in other regions of the developing world facing similar social problems and who are in a similar intellectual position - they have overlooked discourse as a unit of analysis. Agger, B. (1991). Critical Theory, Poststructuralism, Postmodernism: Their Sociological Relevance. Annual Review of Sociology, 17, 105-131.

Alvesson, M., & Deetz, S. (2000). Doing Critical Management Research. London: Sage Publication.

Bourdieu, P. and Wacquant, L. (2001). New-Liberal Speak: notes on the new planetary vulgate Radical Philosophy 105, 2-5

Bloome, D., and Talwalkar, S. (1997). Critical discourse analysis and the study of reading and writing. Reading Research Quarterly, 32(1), 104-112.

Chouliaraki L & Fairclough N. (1999). Discourse in Late Modernity Edinburgh: Edinburgh University Press 1999

Chouliaraki, L., and Fairclough, N. (1999). Discourse in late modernity: rethinking critical discourse analysis. Edinburgh: Edinburgh University Press.

Dann, G. (2004). Representing the Other in the language of Tourism. Journal of Eastern Caribbean Studies 29(2), 76-97

Eagleton, T. (1991). Ideology. London: Verso

Fairclough, N. (1989). Language and power. London: Longman.

Fairclough, N. (1992). Discourse and Social Change. Cambridge: Polity Press.

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

Fairclough, N. (1995b). The Discourse Reader. London: Routeledge.

Fairclough, N. (1995c). Critical discourse analysis: the critical study of language. New York: Longman.

Fairclough, N. (2000). Language and power (2nd ed.). New York: Longman.

Fairclough, N. (2001). The dialectics of discourse. Retrieved December 23, 2003, from http://www.geogr.ku.dk/courses/phd/glob-loc/papers/phdfairclough2.pdf

Fairclough, N. (2003). Analyzing discourse: textual analysis for social research. London: Routledge.

Fairclough, N. (2004). Critical discourse analysis and change in management discourse and ideology: a transdisciplinary approach to strategic critique. Retrieved January 11, 2006 from http://www.ling.lancs.ac.uk/staff/norman/paper5.doc

Fairclough, N., Graham, P., Lemke, J., and Wodak, R. (2004). Introduction. Critical Discourse Studies, 1(1), 1-7.

Fairclough, N., and Wodak, R. (1997). Critical discourse analysis. In T. van Dijik (Ed.), Discourse Studies: A Multidisciplinary Introduction. London: Sage Publications.

Foucault, M. (1972). The Archaeology of Knowledge. London: Tavistock.

Foucault, M. (1977). Discipline and Punish. London: Allen Lane.

Foucault, M. (1980). Power/Knowledge: Selected Interviews and Other Writings 1972-1977. , Hassocks, Sussex: Harvester.

Foucault, M. (1981). The History of Sexuality Volt I: An Introduction. Harmondsworth: Penguin.

Foucault, M. (1984). The order of Discourse. In Shapiro (Ed.), The Language of Politics. Oxford: Blackwell.

Gee, J. (1999). An introduction to discourse analysis: Theory and method. New York: Routledge.

Habermas.J. (1984). The Theory of Communicative translated by Thomas McCarthy. Boston: Beacon Press.

Hall, S. (2002). The problem of ideology: Marxism without guarantees. In D. Morley and C. Kuan-Hsing (Eds.), Cultural dialogues in cultural studies. North Yorkshire: J & L Composition Ltd.

Harvey, D. (1996). Justice, Nature and the Geography of Difference. London: Blackwell.

Held, D. and McGrew, A., Goldblatt, D. and Perraton, J. (1999), Global Transformations: Politics, Economics and Culture, Polity Press, Cambridge

Jessop, B. (2000). The Crisis of the National Spatio-Temporal Fix and the Ecological Dominance of Globalizing Capitalism. International Journal of Urban and Regional Research, 24(2), 323-360.

Kincheloe, J., & McLaren, P. (2000). Rethinking critical theory and qualitative research. In N. K. Denzin, & Lincoln, Y.S (Ed.), Handbook of qualitative research. Thousand Oaks: Sage Publications.

Kress, G. (1989). Linguistic processes in sociocultural practice. Oxford: Oxford University Press

Lincoln, Y. S., & Denzin, N. K. (2003). Turning points in qualitative research: tying knots in a Handkerchief. Walnut Creek: AltaMira Press.

Macdonell. D. (1986). Theories of Discourse: An Introduction. New York: Basil Blackwell.

Microsoft. (2004). UNDP and Microsoft Announce Technology Partnership To Combat Poverty in Developing Nations Joint Initiatives Will Help Achieve Millennium Development Goals Through Technology Resources and IT Expertise. Retrieved June 10, 2005, from http://www.microsoft.com/presspass/press/2004/Jan04/01-23WorldEconomicForumPR.mspx

Miles, M. M., and Huberman, A.M. (1994). Qualitative Data Analysis: An Expanded Sourcebook. Newbury Park: Sage Publications.

Poster, M. (1989). Critical Theory and Poststructuralism. Cornell: Cornell University Press.

Purvis, T., and Hunt, A. (1993). Discourse, Ideology, Discourse, Ideology, Discourse, Ideology…. The British Journal of Sociology, 44(3), 473-499.

Strauss, A., and Corbin, J. (1990). Basics of Qualitative Research - Grounded Theory Procedures and Techniques. Newbury Park: Sage Publications.

Strauss, A., and Corbin,. (1998). Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory. Thousand Oaks: Sage Publications.

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

Weiss, G., and Wodak, W. (Ed.). (2002). Critical Discourse Analysis Theory and Interdiscliplinary. New York: Palgrave Macmillan.

Widdowson, H. (1995). Discourse analysis: a critical view. Language and Literature, 4(3), 157-72.

Yin, R. (1994). Case study research: Design and methods (2nd ed.). Thousand Oaks: Sage Publications.