Abstract: Post-foundational discourse analysis, also labeled as Essex School in Discourse Analysis, has been observed to suffer from a considerable methodological deficit that limits its applicability in empirical research. The principal aim of this article is to overcome this methodological deficit by constructing the research program of the post-foundational discourse analysis that facilitates its operationalization in empirical research. In accordance with Imre LAKATOS (1970) and David HOWARTH (2004a), a research program is referred to an internally consistent and openly scrutinizable system of theoretical, methodological and phenomenal concepts that opens up the possibility to distinguish between the "negative" and the "positive" heuristics of empirical research. The first three sections develop the positive heuristics of the post-foundational discourse analysis by elucidating its theoretical foundations, methodological position and phenomenal framework. The concluding fourth section draws on the presented positive heuristics to outline the analytical stages and strategies of the post-foundational discourse analysis and discusses suitable methods for sampling and interpreting empirical data.

Key words: David Howarth; discourse; Ernesto Laclau; Essex School in Discourse Analysis; Imre Lakatos; methodological holism; post-foundational discourse analysis (PDA); Rainer Diaz-Bone; research program; second-order hermeneutics

Social sciences have been the scene for a discursive turn that has given rise to various analytical approaches to discourse analysis such as "critical discourse analysis" (CDA) (e.g. WODAK & MEYER, 2009), discourse analysis after structuralism (e.g. ANGERMÜLLER, 2007), "Foucaultian discourse analysis" (e.g. DIAZ-BONE, 2006, 2007, 2010), "sociology of knowledge approach to discourse analysis" (e.g. KELLER, 2011a, 2011b) and governmentality studies (e.g. MARTTILA, 2013a). In contrast to these very distinctive and well-elaborated approaches to empirical discourse analysis, the so-called "post-foundational discourse analysis" (PDA)¹⁾—more commonly labeled as "Essex School in Discourse Analysis" (PURVIS & HUNT, 1993, pp.473f.)—has been attested a considerable "methodological deficit" (HOWARTH, 2006, p.23). As a result of this methodological deficit, there is hardly any information available about the analytical instructions and methodical guidelines of the PDA (cf. KELLER, 2011a, p.164). The aim of this account is to develop general methodical guidelines that can solve the PDA's methodological deficit. [1]

Methodological questions can hardly be reduced to issues related only to gathering and interpreting empirical data. Instead, as GLYNOS and HOWARTH (2007, p.6) underline, methodological questions touch inevitably on "the ontological and epistemological dimensions of any social inquiry." HOWARTH (2006, p.23) argues, therefore, that we are but left with the option to solve the PDA's methodological deficit by constructing its entire "research program." In accordance with LAKATOS (1968, 1970), a research program's characteristic grammar of concepts consists of a "hard core" of theoretical, methodological and phenomenal concepts (LAKATOS, 1970, p.135), which together "tell[s] us what paths of research to avoid (negative heuristic), and ... what paths to pursue (positive heuristic)" (1968, pp.167f.). In other words, the "grammar of concepts" that is characteristic of PDA's research program constitutes a heuristic framework that capacitates us "to elucidate a particular phenomenon in a specific way" (HOWARTH, 2004a, p.245). This account departs from the notion of the research program as an internally coherent "grammar" of theoretical, methodological, phenomenal and analytical concepts (ibid.). PDA's research program is developed by means of the elaboration of its "theoretical foundation," which embraces general propositions about the structural organization of social life (Section 2), "methodological premises" conceived to both facilitate and constrain empirical discourse analysis (Section 3), and a "phenomenal framework" of phenomenal features in whose form the world's assumed structural organization becomes accessible for empirical observations (Section 4). Drawing on this tripartite grammar of theoretical, methodological and phenomenal concepts, the aim of the concluding part (Section 5) is to solve the PDA's methodological deficit by explicating the general methodical guidelines for empirical research, shedding light on the analytical stages (Section 5.1), and identifying suitable methods for generation and interpretation of empirical data (Section 5.2). [2]

The PDA's theoretical foundations derive from SAUSSURE's (1959 [1916]) structuralist theory of language and its subsequent revision in the works of DERRIDA (e.g. 1981a [1972], 1981b [1967]). I follow the example of MARTTILA (2015/Forthcoming) and use the concept of "post-foundational discourse analysis" (PDA) instead of the more commonly used concept of "post-structural discourse analysis" to underline the fact that PDA does not abandon the "relational epistemology" of meaning that characterized SAUSSURE's structuralist theory of language. In general terms of speaking, the notion of relational epistemology of meaning implies that meaning is supposed to be the contingent outcome of the relations between elements conveying meanings. Whereas the essentialist epistemology of meaning presupposes partly the presence of a "one-to-one relation" between words and their referential objects and partly that objects' meanings reflect their objective phenomenal properties (LACLAU & MOUFFE, 1990, p.109), SAUSSURE (1959 [1916], p.120) argued that language does not consist of "positive," but rather of differential terms, whose meanings become determined by their mutual relations. This relational epistemology of meaning implies that "the identity [i.e. signified] of a sign ... is constituted on the basis of defining and asserting itself in terms of what it is not—that is, on the basis of difference" (BOWMAN, 2007, p.12). Relational epistemology of meaning implies that meaning conveying entities—i.e. signs—must be conceived of as connecting two mutually distinctive levels of the signifier—the phonetic or graphic structure of the sign—and the signified—that is the concept expressed by the sign (SAUSSURE, 1959 [1916], pp.9, 113f., 120). [3]

SAUSSURE underlined that the mutual relations of signifiers, and hence their signifieds, reflected the socio-historical requirements that pertained to practices of language-use (i.e. articulation). For SAUSSURE, the contextually contingent nature of practices of articulation was proven by the fact that different languages used different signifiers (e.g. pferd, cavallo, häst) to express the same concept of "horse" (cf. LACLAU & MOUFFE, 2001 [1985], p.106). This implies that the signifier horse is not confined to one and the same signified, but may, in the context of agriculture, refer to a distinctive species of animals and, in the context of a nightclub, to the intoxicant heroin. Notwithstanding, SAUSSURE assumes that the practice of articulation—i.e. parole—is restricted by the "preconceived possibilities" of the system of language—i.e. langue (JAKOBSON, 1990, p.118). In other words, the range of socially accepted practices of articulation is determined by the rules that embrace the entire community of language users (SAUSSURE, 1959 [1916], pp.71ff.). SAUSSURE assumes that only the adherence to collectively shared rules would "permit individuals to exercise the faculty of language" (JAKOBSON, 1990, p.88). [4]

However, it was above all DERRIDA, whose works on deconstruction redressed the contradictions and shortcomings in SAUSSURE's structuralist theory of language, and which later on built the theoretical fundament of the PDA (e.g. LACLAU & MOUFFE, 2001 [1985], pp.111f.). DERRIDA's critique of SAUSSURE addressed the very status of social rules. For DERRIDA (1981b [1967], pp.278ff.), a consistently relational epistemology of meaning had to imply that practices of articulation regulating social rules could be but the results of antedating practices of articulation. This meant also that social rules persist only as long as they are maintained by consecutive social practices. DERRIDA's deconstruction of the objective status of social rules revealed the absence of any kind of objective regulation of social practices of articulation. As such, there does not exist any socially objective "absolute anchoring," which in an objective and self-evidential manner could regulate practices of articulation (1977, p.12). DERRIDA's insight that social practices were not regulated by objective rules, but by contingent rules—i.e. by rules that originated from preceding practices of articulation—implied that there could not exist any objective and socially uncontestable distinctions between valid and invalid meanings. Hence, practices of articulation were likely to involve in continuous modification and replacement of already prevailing structures of meaning (DERRIDA, 1981a [1972], p.29, 1981b [1967], p.280). After all, the absence of objective rules—i.e. "transcendental signified"—could only mean that the range of valid practices of articulation was extended ad infinitum (DERRIDA, 1981b [1967], p.112). [5]

In PDA the concept of discourse refers to any differential arrangement of discursive elements—i.e. signifiers—in which these elements receive their distinctive meanings—i.e. signifieds. However, PDA makes use of the concept of discourse also to describe the structural logic, which in the absence of any objective rules, regulates the range of socially valid and acceptable practices of articulation (LACLAU, 1990, p.91). More precisely, discourse refers to a spatio-temporally distinctive structural arrangement of signifiers that functions as a "totalizing horizon" of intelligibility (LACLAU & MOUFFE, 2001 [1985], p.144). A discourse represents a presupposedly natural and self-evident way of "fixing of the relation between signifier and signified" and, by doing so, defines objects' (i.e. signifiers') social meaningfulness (LACLAU, 1993, p.435). The discourse specific horizon of intelligibility regulates social production of meaning partly by means defining the presupposedly self-evidential meanings of objects and partly by means of distinguishing the sets of articulations that social subjects must adhere to in order to confirm to the discourse specific conceptions of the truth. Discourses regulate the range of meaningful practices of articulation in both regards and subjugate articulations conducting subjects to "repeat" the already prevailing meanings that inform us about objects' very being and meaningfulness (LACLAU & MOUFFE, 2001 [1985], p.108). [6]

The structural impact exerted by discourses can be manifested with regard to the distinction between discursive "moments" and "elements" (LACLAU & MOUFFE, 2001 [1985], p.113). The concept of moment refers to a signifier with a distinctive and persisting position within a differential chain of signifiers. Due to their temporally persistent positions in such a chain of signifiers, moments become associated with distinctive meanings (i.e. signifieds) that continue from one practice of articulation to another. In contrast to a moment, an element refers to a signifier that falls short of any definite and persisting position within a chain of signifiers. As such, an element has the status of a fluctuating signifier, whose meaning varies from one articulation to another (cf. GLASZE, 2007, §18). The difference between moments and elements indicates that the "temporal" stability of the meaning associated with a signifier equals the "structural" stability of the chain of signifiers. Even more importantly, the assumed relation of homology between the stability of meaning and the stability of the structural order of discourse underlines that social subjects' adherence to the same discourse is manifested by the mutual coherence of their practices of articulation (LACLAU, 1990, p.207; see further in Section 4). [7]

The absence of any objective foundations of discourse means that discourses can only originate from practices of articulation installing them (cf. MARCHART, 2007, p.14). Discourses become installed by "hegemonic agents" that partake in "construction, defense, and naturalization of new frontiers" of discourse (GLYNOS & HOWARTH, 2007, p.141). Hegemonic agents give rise to hegemonic—i.e. presupposedly self-evident—discourses, which social subjects must adhere to in their practices of articulation (ÅKERSTRØM ANDERSEN, 2003, p.56). In more precise terms, the construction of hegemonic discourses consists of both "counter-hegemonic acts" (here: "radical acts"), which contest an already prevailing discourse (p.115), and "hegemonic acts," which install an unprecedented discourse or extend an already prevailing discourse to apply to new social contexts, issues, themes, etc. (ibid.). A discourse has attained a hegemonic status when its distinctive conception of the world has developed into an inter-subjectively shared and socially largely taken-for-granted horizon of intelligibility. The social acceptance of a discourse is supported by the phantasm of objective necessity: that it was "always already there in the first place" (TORFING, 1999, p.167). In other words, a discourse becomes more self-evident and temporally durable the more it is based on "the illusion that it [i.e. discourse] was already there, i.e. that it was not placed there by us" (ŽIŽEK, 1995, p.95). [8]

The relational epistemology of meaning does not only constitute the very possibility to conceive of social life as being structured in the form of discourse, but also identifying the general characteristics of discourses. The absence of any objective foundation of discourse means that discourses are necessarily defined ex negativo; that is by means of distinguishing them "from what they are not" (DYRBERG, 1997, p.119). This means that a discourse comes into existence as the result of the distinction drawn between itself and the "general field of discursivity"—the totality of other discourses (LACLAU & MOUFFE, 2001 [1985], p.106). Discursive limits open up the possibility to distinguish between "valid" and "invalid" practices of articulation and command social subjects to adhere to the discourse specific conceptions of the truth; they constitute the very condition of the possibility of meaning. While signifiers can obtain their meanings only "through their mutual differences in the discourse ... [t]he condition of possibility of mutual differences ... is that the elements are identical or equivalent in respect to belonging to the discourse and existing within the boundaries of the discourse" (ÅKERSTRØM ANDERSEN, 2003, p.54). However, discursive limits are anything but self-evident. LACLAU and MOUFFE (2001 [1985], pp.105f.) argue that a discourse can be distinguished from other discourses only if there is some "point of reference" that symbolizes the identity of a discourse and, by doing so, makes it possible to conceive of the logic of commonality that binds signifiers together. In PDA, so-called "nodal points"—also described as "empty signifiers"—constitute the "points of reference" that symbolize the identity of a discourse, making it possible to distinguish a discourse from other discourses, and conceive of the logic of commonality that attaches signifiers to relations of mutual differences (ŽIŽEK, 1989, p.95). TORFING (1999, p.225f.) describes how the concept of "welfare state" acts as a nodal point that makes it possible to symbolize the common identity of various "political strategies, institutional forms and power networks ..." and makes them distinguishable from other discourses. [9]

PDA does not reduce discourses to linguistic phenomena and regard them as resulting from written and spoken practices of articulation. LACLAU (1980, p.87) argues that discourse does not "refer to 'text' narrowly defined, but to the ensemble of the phenomena in and through which the social production of meaning takes place." The relational epistemology of meaning implies that discourses cannot be based upon any objective necessities. Instead, the conceived self-evidentiality of a discourse results from the process of discursive "sedimentation." There are two different types of discursive sedimentation. The first type consists of "forgetting" the contingent origins of a discourse (GLYNOS & HOWARTH, 2007, p.116). A partial decoupling of a discourse and its historical origins—that is hegemonic acts that installed and motivated its very presence—conceals the lacking objective necessity of a discourse and, at the same time, links it with the image of naturalness (LACLAU, 1990, p.35). In contrast, the so-called "reactivation" denotes the rediscovery of "the 'constitutive' activity" that gives rise to a discourse (GLYNOS & HOWARTH, 2007, p.116). The recovery of the historical origins increases social subjects' capacity to reflect upon the lacking objectivity of the socially presupposed objective necessity of a discourse (LACLAU & MOUFFE, 2001 [1985], p.viii). Reactivation takes place when social subjects become aware of "the contingent nature of the so-called 'objectivity'" and start questioning the necessity of the prevailing social order and the discourse that motivates it (LACLAU, 1990, p.35). Neither forgetting of historical origins of discourse nor their reactivation constitutes a self-evidential process. While the process of forgetting of historical origins results from hegemonic acts, which attempt to establish a new discourse with its distinctive limits and distinctions from other discourses, reactivation of historical originates from counter-hegemonic acts that "attempt to disarticulate it [i.e. the prevailing discourse] in order to install another form of hegemony" (MOUFFE, 2008, p.4). This first type of discursive sedimentation qua forgetting and concealment of a discourse's historical origins can to some extent explain its relative temporal stability. [10]

The second type of sedimentation—"discursive materialization"—has been given little attention in recent contributions to the PDA. This is rather surprising because discursive materialization is of crucial relevance for the temporal stability of a discourse. When a discourse becomes a socially taken-for-granted horizon of intelligibility, it is conceived to give objective meaning to and to motivate non-linguistic objects such as artifacts, pragmatic practices and institutions (TORFING, 1999, p.35). In other words, sedimentation qua discursive materialization transfers a discourse specific horizon of intelligibility into a corresponding "discursive materiality." Objects motivated and generated by means of discursive materialization achieve the status of mere "objective presence" (LACLAU, 1990, p.34). LACLAU's notion of "objective presence" implies that social subjects conduct routinely particular practices of articulation, identify themselves with particular subject roles and accept the social and symbolic authority of institutions without having to reflect upon their social meaningfulness. In other words, as HOWARTH (2000, p. 9) argues, discursive materialization consists of a partial decoupling of "historically specific systems of meaning" and "concrete systems of social relations and practices." The crucial point about discursive materialization is that the emergent discursive materiality restricts the range of socially meaningful practices of articulation and regulates social subjects' "enunciative possibilities" (STÄHELI, 2001, p.208; my translation). According to HOWARTH (2005, p.343) we can operationalize "enunciative possibilities" by conceiving of practice of articulation as consisting of two phenomenal dimensions of énoncé and énonciation. Whereas énoncé refers to the practice of articulation as an activity of speaking, writing and acting, énonciation refers to the sedimented "conditions and roles that must be satisfied" if a particular practice of articulation is "to qualify as meaningful" (ibid.). Unfortunately, recent contributions to PDA do not clarify further what such practices of articulation regulating contextual "conditions" consist of and how they can be made empirically visible. In contrast to the recently burgeoning literature on "new materialism" (LEMKE, 2015) and its deliberate attempts to conceptualize the contextual premises of social agency, PDA has remained without further operationalization of the discursive materiality. Basically, the question is what kinds of discursively constructed social structures can be assumed to regulate social subjects' practices of articulation and hence give stability to discourses. [11]

Obviously, and as the literature on "new materialism" also indicates, there is not just one way of thinking and conceptualizing the composition of the discursive materiality (LEMKE, 2015). One possible way of thinking about discursive materiality is to see discourses as giving rise to particular sets of "institutions" and "subject roles" that regulate social subjects' practices of articulation by making them accept and stick to discourse-specific conceptions of the world. For ÅKERSTRØM ANDERSEN subject roles refer to relatively elaborate and mutually distinctive social positions, which social subjects must possess in order to "speak and act meaningfully in a specific way" (2003, p.24). Also GLASZE (2007, §38) argues that subject roles are "institutionally stabilized positions ... which are related to specific conditions of access, and which connect specific possibilities, taboos and expectations with language-use" (my translation). GLASZE's conception of a subject role as an institutional(ized) position indicates that discourses compel social subjects "to engage in acts of identification" with discourse specific subjects roles (GLYNOS & HOWARTH, 2007, p.129). Therefore, subject roles are not freely accessible, but linked with particular criteria of access, which together with these criteria maintaining and enforcing bodies—i.e. institutions—determine "who is granted access and what is expressible" by a given subject in a given situation (STÄHELI, 2001, p.208; my translation). The access to subject roles can be a matter of more or less formalized processes of selection and require the possession of certain qualifications such as academic degrees or specific social skills (TORFING, 1999, p.70). Subject roles are also connected with rules relating to their subjective self-appropriation. Such rules of self-appropriation mean that a social subject is expected "to ... behave in a certain way, to pursue particular goals, and to have distinctive affections" in his/her capacity as holder of a subject role (BOURDIEU, 2005, p.19; my translation). [12]

The concept of institution constitutes something of a "black box" in discourse analysis in general (cf. CLEGG, COURPASSON & PHILLIPS, 2006) and the PDA in more particular (cf. BOUCHER, 2008, pp.97ff.; CARPENTIER & SPINOY, 2008, p.15). In the absence of any sufficient previous conceptualization of institutions, I am keen to argue (in an admittedly pragmatic manner) that institutions can be related to all kinds of social bodies with the socially accepted authority to sustain and implement the validity of particular discourses, regulate social access to subject roles, and sanction and control social subjects' adherence to presupposedly natural courses of both linguistic and pragmatic actions. Institutions originate from hegemonic and counter-hegemonic acts, which either motivate the installment of particular institutions or overdetermine the meaningfulness of already prevailing institutions and redefine their appropriate functions and purposes. Institutions implement and sustain discourse-specific "unevenness of the social ..." (LACLAU, 1996, p.43), because they support particular conceptions of the world, sanction and punish observed deviations from these conceptions, and regulate social subjects' access to socially influential and valuable subject roles. This means that institutions are of crucial relevance for the implementation and maintenance of discursively defined and motivated "subordinate-superior relationships" between different (groups of) social subjects (CLEGG et al., 2006, p.330). Institutions constitute crucial sources of stabilization of discourses, not least of all because they are assigned the responsibility to oversee social subjects' compliance with the rules of self-appropriation connected with subject roles (cf. GLASZE, 2007, §38; STÄHELI, 2001, p.208; TORFING, 1999, p.153). Altogether, widely sedimented institutions possess the authority to oblige social subjects to conduct particular practices of articulation, which again stabilize particular discourses. [13]

Section 2 manifested that, while practices of articulation are structured by sedimented and materialized discourses, they are at the same time the very origin of any discourse. This reciprocal relation of social practice (of articulation) and structure (discourse) is akin to the methodological position of the neostructural "second-order hermeneutics" (FRANK, 1984, pp.279ff.; my translation). I would like to argue that second-order hermeneutics constitutes the first methodological premise of the PDA (see Section 3.1). Basically, second-order hermeneutics implies that the research is focused on the supra-subjective—existing beyond social subjects' conscious conceptions of reality—and social structures recurring from one context to another (see Section 3.1). However, it is hardly possible to define and describe social structures without some kind of a priori conception of their general characteristics. While the theoretical foundations of the PDA render visibility to both the origins and general characteristics of discourses, they also introduce an epistemic bias into empirical research. After all, we are unlikely to be capable of carrying out PDA without accepting the epistemic horizon of it. In order to analyze social reality in consistence with the theoretical foundations of the PDA, conceptions of discourse and discursive materiality characteristic of PDA must not only inform us "about the nature of the reality being studied ...", but also instruct us about "the ways by which one can study that reality" (ALASUUTARI, 1996, p.373). As BOURDIEU's (2008 [2002]) and DIAZ-BONE's (2006, 2010) works on a "reflexive methodology" manifest, the consistence between a priori assumed theoretical foundations and practical research can be best ensured by making "holistic" use of theory in empirical research. I therefore suggest that "methodological holism" should constitute the second methodological premise of the PDA (see Section 3.2). [14]

Second-order hermeneutics is a methodological stance to empirical research that was above all influenced by HEIDEGGER's (e.g. 2008 [1923]) critique of the phenomenological methodological position of Edmund HUSSERL. The crucial issue about the distinction between HEIDEGGERian post-phenomenological methodology and the HUSSERLian phenomenological methodology is that they motivate two distinctive logics of social inquiry: first- and second-order hermeneutics (ALVESSON & SKÖLDBERG, 2000, pp.52ff.). [15]

The phenomenological methodological position is centered around the assumption of a unilinear relationship between the meanings that social subjects assign to objects and their overall social meaningfulness. Social subjects are not only assumed to be fully aware of the meanings related to objects, but their self-conscious conceptions constitute the very "'condition of possibility' of [their] significance, meaning and reference" (FRANK, 1984, p.282; my translation). The other way round, the meaning of an object cannot but be the result of the "process carried out by a subject upon [this] object" (COSTACHE, 2011, p.498). In other words, the social meaningfulness of objects equals the "perfect self-knowledge of the subject" (TILES, 1984, p.36). Subjects possess the ontological status of "transcendental subjects" because they are not passive beholders of prevailing meanings attached to or contained in things, but originators whose conscious experiences and intentions give rise to collectively shared conceptions of the world (CROWELL, 1990, p.511). It is not the prevailing bodies of knowledge—such as cultural, ideological and religious edifices—but subjects' "transcendental experience" that constitutes the origin of any social meaningfulness (p.512). The presence of the absolute and "pure subjectivity" (ibid.) postulated by HUSSERL gave no reason to distinguish between the "objective"—supra-subjective and subjectively inaccessible—and the "subjective"—subjectively conceivable and reflexible—meaningfulness of the world (cf. DIAZ-BONE, 2010, p.188). As a result, it is of little value to try to disclose some kind of objective, or factical level of the social world that could explain why social subjects ascribe specific meanings to certain objects. When studying the social meaningfulness of a social order, we can only take recourse to the intersubjectively shared and subjectively reflexible structures of meaning (ALVESSON & SKÖLDBERG, 2000, p.87). In the absence of any clear distinction between the objective and subjective foundations of socially meaningful objects, a phenomenological methodological position cannot but motivate analysis of "the meanings that the originators of texts and acts—authors and agents—associate with" particular objects (p.52). The absence of any objective and supra-subjective constitution of the world means that we have little reason to pay attention to anything other than "the contents of human consciousness" (GUMBRECHT, 2004, p.60). The principal analytical challenge associated with the phenomenological methodological position is to render visibility to social subjects' experiences of the world by means of feeling and putting "oneself into the situation of the acting (writing, speaking) person" (ALVESSON & SKÖLDBERG, 2000, p.54). [16]

Post-phenomenological methodology is based upon the assumptions about "the ontological groundlessness of transcendental subjectivity" (GADAMER, 1975, p.227) and the presence of context-specific and subjectively inaccessible "limits of cognition" (p.231). These assumptions open up the possibility to assume that social subjects perceive their objects of observation within context-specific epistemic horizons, which, in their turn, remain beyond the scope of their conscious self-conceptions (MARTTILA, 2010, p.98). As such, the relationship connecting the knowing subject—the knower—with the subjectively perceived object—the known—cannot be based upon the "transcendental" cognitive faculties of the subject. GADAMER (1975, p.232) argues that "neither the knower nor the known are present-at-hand" because both the cognitive faculties of the subjects as well as the meanings they have cognized derive from a supra-subjective and socio-historically distinctive "mode of being." In other words, social subjects are, without their own knowing, "thrown into" some kind of socio-historically distinctive mode of being (HEIDEGGER, 2008 [1923], p.67). It is this ontological status of the subject as one "being-thrown-into-the-world" (ibid.) that opens up the epistemological possibility to start analyzing what such "thrownness" (ibid.) consists of in a particular social context and where it originates from. The reciprocal relationship between the objective—discursive throwing of social subjects into the world located beyond their self-conceptions—and the subjective—throwing themselves into the discursively constituted beingness of the world accomplished by social subjects—which is characteristic of post-phenomenological methodology, constitutes the first methodological premise of the PDA. This distinction between the objective—i.e. "discursive"—constitution of the social meaningfulness of objects, and the subjective—i.e. "articulatory"—self-appropriation of the discursively defined meanings of objects—opens up the methodological condition of possibility to start searching for context-specific forms and processes of the world's discursive structuration (p.58). [17]

Discourses and processes of discursive structuration can be referred to an objective level of reality because they constitute the "deeper structure of factical life that underlies [any] ... intentional correlation between us and the world" (KISIEL, 2010, p.21). However, it is worth noticing that such a factical level is not separated from social subjects' "subjective" self-conceptions, but constitutes instead their "objective" conditions of possibility. The distinction between the ontological "beingness of the world" and the subjectively perceivable "being of the world," which is characteristic of post-phenomenological methodology, motivates empirical research aiming at rendering visibility to subjectively unperceived discourses and discursive materialities. At the same time, a reciprocal relation between discourses and practices of articulation and general structural characteristics of discourse as it is described by the PDA opens up the possibility to achieve an "epistemological break" with social subjects' self-conceptions of the world (FOUCAULT, 2009 [1969], p.206). Epistemological break implies that we abandon social subjects' self-conscious common sense conceptions of the world in order to interpret these conceptions as contingent outcomes of their conditions of conventions, norms, ideologies, social structures, discourses, and the like that constitute possibility and are largely unconscious to social subjects (MARTTILA, 2010, p.105). In other words, epistemological break cannot be achieved independently of theoretically derived conceptions of the "objective" structures that regulate social subjects' faculties to make meaningful conceptions of the world. The inevitable epistemological consequence of the epistemological break is that neither empirical social research in general, nor PDA in more particular, can help falling prey to an epistemically biased conception of the world. After all, the condition of possibility to conduct PDA depends on our acceptance of its key assumptions of the general forms and processes involved in the world's discursive constitution. The question arises as to how we as researchers should deal with our own epistemic bias when carrying out empirical research. [18]

The theoretical framework of the PDA opens up the possibility to achieve "epistemological break" with social subjects' conscious self-conceptions and identify sedimented discourses that regulate their practices of articulation. Many postmodern critics (e.g. LAW, 2004; LYOTARD, 1993; RORTY, 1979) have criticized scientific aspirations to achieve "epistemological break" with social subjects' conscious self-conceptions by means of regressing from their common conceptions of the world (MARTTILA, 2013b, §18). Instead, as among others LAW (2004, p.102) argues, the epistemically biased character of scientific observation makes it meaningless to distinguish between theoretically founded second-order interpretations of social subjects' self-interpretations and social subjects' conscious first-order self-interpretations. For LAW, the scientific knowledge about "[w]hat there is and how it is divided up should not be assumed beforehand. Instead, it arises in the course of interactions between different actors" (ibid.). However, postmodern critique of second-order hermeneutics disregards that the observed lacking objectivity of scientific practice is itself based upon an equally biased conception of the (relativist) nature of knowledge. In other words, the refusal of the second-order interpretation of social subjects' first-order interpretations cannot get rid of the epistemic bias as such. Therefore, the abandonment of any a priori theoretical conceptualization of the world, as it has been suggested by LAW (2004), does not constitute any viable means to deal with the epistemic bias of scientific practice. [19]

GLYNOS and HOWARTH (2008, p.15), LACLAU (2004, p.323) and MARTTILA (2013a, pp.83ff.) make the case for a reflexive methodological position that underlines the necessity to not try to abandon, but rather explicate the epistemic bias involved in empirical research. The holistic use of theory in empirical research opens up the possibility to choose, define and interpret empirical objects "in terms of the distinctions brought about by [our] ontology" (LACLAU, 2004, p.323). Holistic use of theory necessitates mutual consistence between the empirical "facts"—i.e. the observations of the empirical being of the analyzed objects—to reflect their a priori assumed "factual propositions" about objects' ontological beingness (LAKATOS, 1999, p.97). Following NADEL's (1962, p.1) suggestion, the consistence between the theoretical frameworks and empirical observations is ensured by the operationalization of theoretical propositions into corresponding analytical concepts and categories. Such a heuristic use of theory ensues when our theoretical framework is operationalized into an empirically applicable "body of propositions," which makes it easier for us "to map out the problem area and thus prepare the ground for its empirical investigation by appropriate methods" (ibid.). Following the lead of MARTTILA (2013b, §22), such a theoretically derived and analytically usable body of propositions can be defined to consist of "theoretical codes." [20]

I consider it expedient to define theoretical codes by contrasting them to empirical codes. Empirical codes are developed ad hoc during the empirical analysis and mirror empirically observed phenomenal characteristics of objects (cf. KELLE, 2005, §49f.). In contrast, theoretical codes are derived from the theoretical framework that serves the function of an "interpretive frame[s] from which [we] ... view realities" (CHARMAZ, 2006, p.128). Moreover, while the analytical validity of empirical codes is determined by their correspondence with the observed empirical properties of objects, the analytical usefulness of theoretical codes depends both on their consistency with the a priori assumed phenomenal properties of the objects and on the possibility to relate them to concrete empirical phenomena. Even though theoretical codes are only available "independently of data collection and data analysis" (KELLE, 2005, §49), they are of little analytical use unless they can, similar to middle-range concepts, interlink "abstract statements and complex empirical evidence" (RAPPERT, 2007, p.695). Although middle-range concepts reflect our ontological and theoretical commitments, they must still be "specific enough so as to allow for the observation of their correspondence with the empirical features of the studied phenomena" (MARTTILA, 2013b, §10). In other words, middle-range concepts provide an excellent means to translate our theoretical frameworks into a set of theoretical codes that allow us to think about and interpret empirical data in consistence with the PDA's theoretical framework. The following section aims to operationalize the theoretical framework of the PDA into a theoretically consistent and analytical applicable heuristic framework of theoretical codes. [21]

As a reminder, discourse refers to a structural arrangement of discursive elements that originates from practices of articulation, and which, in its turn, constrains social subjects' potential practices of articulation. GLYNOS and HOWARTH (2007, p.140) point out that social subjects' adherence to the same discourse is manifested by the mutual coherence of their practices of articulation. GLYNOS and HOWARTH (ibid.) suggest that the mutual coherence of practices of articulation is manifested by "patterns" of discursive elements. They argue further that mutually coherent "patterns of discursive articulation" (ibid.) become empirically visible in form of recurrent patterns of discursive "relations"—which connect discursive elements with each other—and "identities"—which discursive elements have in connecting discursive relations. There are two good reasons to argue that concepts of discursive relations and identities can constitute the above (in Section 3.2) described theoretically consistent and analytically applicable categories of "theoretical codes." Firstly, the concepts of discursive relations and identities support the realization of the holistic use of theory in empirical research because they can be derived from the PDA's theoretical framework. Hence, discursive relations and identities serve for the methodological holism's essential idea of "theory-driven construction of phenomena" (DIAZ-BONE, 2007, §35; my translation). Secondly, discursive relations and identities constitute empirically applicable middle-range concepts. The following two sections (4.1 & 4.2) will elaborate discursive "relations" and "identities" in greater detail. [22]

We must bear in mind that the relative regularity of articulations is the contingent outcome of discursively defined "enunciative possibilities" sedimented by means of discursive materialization. In accordance with the earlier elaboration of the concept of discursive materiality (Section 2.2), the mutual coherence of articulations bears witness to the influence exerted by sedimented subject roles and institutions. In other words, we cannot settle for the analysis of the meaning contents of practices of articulation, but interpret discourses observed in social subjects' articulations against the background of the subject roles, which social subjects adapt in their practices of articulation, and institutions, which restrict the social access to subject roles. In order to take into account the material conditions of possibility of a discourse, I argue that the analytical focus must be extended beyond the analysis of discursive structures, and also involve analysis of the reciprocal relations between discourse and discursive materiality. I will argue in Section 4.3 that discourse analysts must render visibility to discursively motivated discursive materialities sustaining discursive stability. [23]

In their recent works, MARTTILA (2013a) and NONHOFF (2006, 2007) have argued that discursive elements can be connected by means of four different kinds of discursive relations: contrariety, representation, difference and equivalence. The initial discussion (Section 2.1) indicated that a discourse exists as a relatively coherent and temporally stable structure only when it is separated from other concomitant discourses. LACLAU and MOUFFE (2001 [1985], p.143) argue that a discourse exists as a distinctive structural arrangement of discursive elements only insofar as it can be "cut out as a totality with regard to something else beyond them." In other words, the "relation of contrariety"—the distinction between the interior realm of discursive elements and its exterior, the disregarded and excluded discourses—constitutes the very condition of possibility of any spatially distinctive and temporally durable discourse. Recent contributions to the PDA have equated the relation of contrariety with the "relation of antagonism" that consists of a distinction between two (or more) mutually opposed elements, such as just/unjust, friend/enemy, democratic/totalitarian, etc. (e.g. GLYNOS & HOWARTH, 2007, p.106). A sample of discursive elements can be conceived of as belonging together only if they are installed into a relation of equivalence with regard "to the other, to the antagonistic adversary," which opposes their identity (ŽIŽEK, 1990, p.253). The distinction of a discourse (A) from what opposes it (non-A) opens up the possibility to constitute discursive limits that separate a discourse from what "it is not" (LACLAU & MOUFFE, 2001 [1985], p.143). MARTTILA (2013a, p.44f.) and STÄHELI (2004, p.235) have questioned the necessity to reduce the relation of contrariety to a relation of antagonism. If we follow LUHMANN (1997, pp.45ff.), discursive boundaries could be likewise well established by way of distinction between "marked" and "non-marked" discursive elements. In other words, practices of articulation sustain a prevailing discourse by reproducing constitutive distinctions of "actual"—i.e. cognizable—and "potential"—i.e. non-cognizable—discursive elements. The distinction between the actual and the potential arrangements of discursive elements does not separate two marked realms of discursive elements, but consists of a distinction made between subjectively perceived and non-perceived realms of discursive elements. I therefore argue that the distinction between the marked and unmarked discursive elements consists of a "relation of dissociation." [24]

Besides the relations of antagonism and dissociation, the relation of contrariety may also appear in form of the "relation of incommensurability" (MARTTILA, 2013a, pp.60ff., 2015, Ch.6.2). Basically, relation of incommensurability means that the commensurability of discursive elements is negated with regard to a discursive element, which represents the overall principle of their incommensurability. In other words, the relation of incommensurability is but one distinctive sub-type of a relation of contrariety, which separates two (or more) discursive elements (e.g. A & B) with regard to a third element (e.g. C) that symbolizes the common point of reference—i.e. the nodal point—against the background of which A and B can be conceived of as being mutually incommensurable. It would be possible, for instance, to establish the limits of the welfare state discourse by separating welfare-related issues from non-welfare-related issues. While mass-employment (A) is considered to be of crucial relevance for social welfare (C), gender inequality (B) is deemed as a social problem of little relevance for social welfare. Hence, mass-unemployment and gender inequality are separated by means of their mutual incommensurability with regard to the nodal point of social welfare. To sum up, relation of antagonism, relation of dissociation and relation of incommensurability constitute the discursive relations that make it possible to distinguish a discourse from other discourses. [25]

While the different types of relations of contrariety make it possible to draw a distinction between a discourse and the totality of other discourses, the "relation of representation" makes it possible to symbolize the common identity of the elements that belong to the same discourse. In other words, the "relation of representation" makes it possible to symbolize the logic of commonality that binds discursive elements into a discourse (DYRBERG, 1997, p.125). "Relation of representation" is exerted by the so called nodal points, which represent the common identity of discursive elements and make it possible to conceive of them as "a single unity" (TORFING, 1999, p.98). More precisely, a nodal point (C) constitutes a generic concept that manifests different discursive elements (A, B, C, D) to be mutually equivalent. For instance, the nodal point of social welfare (C) can be used to represent the logic of commensurability that connects discursive elements of mass-unemployment, social security, public schools and state pension insurance and makes it possible to grasp their shared commonality. This example shows that the relation of representation does not only connect mutually distinctive discursive elements, but also consists of a distinctive kind of a relation between the "representative"—the nodal point—and the "represented"—the sample of discursive elements whose common identity is symbolized by the nodal point. While a nodal point acts as a representative that symbolizes and "embodies" the presupposedly intrinsic essence of the "represented" discursive elements, the represented discursive elements substantiate the nodal point representing their common identity (LACLAU, 1996, p.97). [26]

LACLAU and MOUFFE (2001 [1985], p.129) argue that discursive limits and nodal points open up the possibility to observe the shared commonality—i.e. equivalence—of the otherwise mutually distinctive discursive elements. To recall: discursive elements can obtain particular identities—i.e. signifieds—in differential relations to other elements. However, these differential relations are also partially "cancelled out by the equivalent relation provided by the elements' attachment to the discursive structure" (ÅKERSTRØM ANDERSEN, 2003, p.54). After all, discursive elements can be established in "relations of difference" only if they are also mutually equivalent—partly with regard to the nodal points, which represent their common identity, and partly with regard to their distinction from other discourses. In other words, to a discourse belonging elements are not only related by means of the relation of difference, but also any such relation of difference motivating “relation of equivalence." This means that discursive elements are always "equivalent in one respect and different in another" (JAKOBSON, 1990, p.119). Therefore, discursive elements become connected both by means of their mutual distinctiveness sustaining a relation of difference and a relation of equivalence that partly supersedes such distinctiveness. [27]

Distinction between different types of relations of contrariety is of crucial relevance for the empirical discourse analysis because it allows us to distinguish between different general structural organizations of discourse. As argued before, discursive boundaries can be based upon different types of relations of contrariety. A popular discourse is characterized by the relation of antagonism that separates a discourse and its antagonistic other. In popular discourse, discursive elements "are [not] equivalent ... insofar as they share a positive property ... but, crucially, insofar as they have a common enemy" (GLYNOS & HOWARTH, 2007, p.144). This means that a discourse exists in form of a popular discourse when its limits are constructed by means of representations of "common negation or threat" (ibid.). Elements belonging to a popular discourse are not exclusively but primarily connected by means of the relation of equivalence because they are mutually equivalent with regard to their opposition to the antagonistic other. In contrast to the popular discourse, the limits of a pluralist discourse are constituted without any explicit representation of the antagonistic other (MARTTILA, 2013a, pp.72ff.). The absence of an antagonistic other means that the elements of a pluralist discourse are not subdued to elements of a chain of equivalence, within which they tend to lose their distinctive meanings. Hence, elements belonging to a pluralist discourse are primarily connected by means of the relation of difference because they possess mutually distinctive meanings. Notwithstanding, elements of a pluralist discourse are still mutually equivalent in their relation to their common identity representing nodal points. It is obvious that popular and pluralist discourses do not refer to two absolutely contrary discursive logics, but refer to two ideal types of discourse. [28]

MARTTILA (p.69) argues that while the Thatcherite political discourse constitutes the archetype of the popular discourse, the Blairite political discourse is a case in point of a pluralist discourse. The following Table 1 provides an overview of the outlined discursive relations and describes their general phenomenal features.

The observation of mutually coherent practices of articulation is not only facilitated by the identification of discursive relations, but also by the conceptualization of the identities that practices of articulation attribute to discursive elements (GLYNOS & HOWARTH, 2007, p.140). Phenomenal conceptualization of discursive elements is a heuristic means to identify relatively regular patterns of social practice, highlight their similarities and differences, and detect their diachronic changes. Following SILVERMAN (1997, pp.72ff.), phenomenal conceptualization of discursive elements opens up the possibility to detect both the paradigmatic coherence of practices of articulation—i.e. their attribution of the same identity to the same element—and their syntagmatic coherence—i.e. the regular reoccurrence of particular discursive elements with their distinctive identities. However, to detect paradigmatically and syntagmatically coherent practices of articulation, we need to ponder first of all upon the epistemological possibilities and constraints related to the empirical conceptualization of the identities of discursive elements. [30]

The initially (in Section 2.1) discussed relational epistemology of meaning underlined that discursive elements cannot possess any inherent identities—i.e. signifieds (GLASZE, 2007, §30). Instead, the identities possessed by discursive elements vary in accordance with their actualized relations to other elements. Relational epistemology of meaning indicates that the identities of discursive elements vary in accordance with the "part[s] that [they are] playing in the particular structure under consideration" (HALLIDAY, 1985, p.29). In other words, relational epistemology of meaning underlines that the identity of a discursive element depends on its semantic function in the context of its appearance. Whether a stone has the identity of "a projectile or an object of aesthetic contemplation" depends on the composition of the "discursive configuration" of other elements, in which it appears (LACLAU & MOUFFE, 1990, p.101). We may expect a stone to be observed as an "aesthetic" object by an artist, be depicted on canvas, etc. We may describe the stone in terms of a "projectile" when it is conveyed through the air, aimed at hitting an object, etc. [31]

The methodological position of second-order hermeneutics makes it inevitable to achieve an epistemological break with social subjects' self-conceptions of the meanings of discursive elements. Consider, for instance, the statement "boys throw stones" (HALLIDAY, 1985, p.31). An analysis based on the methodological position of first-order hermeneutics would describe the meanings that social subjects associate with these elements (i.e. boys/throw/stones). In other words, the analytical focus would be directed at social subjects' conscious self-conceptions. The relational epistemology of meaning and the methodological position of second-order hermeneutics make it necessary to achieve an epistemological break with social subjects' conscious self-conceptions and interpret the identities of "boys," "throw" and "stones" in accordance with the parts that they play in the context of their appearance. HALLIDAY (p.27) underlines that such relational conceptualization of identities cannot but consist of intentional labeling and "putting names on things." In terms of VAN LEEUWEN (2005, p.4), the identity of a discursive element can be conceptualized by looking at its context-specific "semiotic potential"—i.e. "the kinds of meanings it affords" due to its actualized relations to other elements. Conceptualization of semiotic potential is about "studying how that resource [i.e. element] has been, is, and can be used for purposes of communication" (p.5). Reflecting the relational epistemology of meaning, the "meaning potentials" must be disclosed with regard to the roles that different discursive elements play in their relations to other elements. Let us return to the statement "boys throw stones." A conceptualization of discursive identities that is consistent with the relational epistemology of meaning would imply that the discursive element "boys" can be conceived of as possessing the "meaning potential" of a "subject" because its co-occurrence with "throw" and "stones" indicates that it performs a particular social practice: throwing stones. The discursive element of "throw" can be understood as playing the role of an "action" because it refers to a social practice performed by a "subject." "Stones" can be related to the role of a "resource" because its presence constitutes the (physical) condition of possibility of the performed "activity." [32]

The previously introduced methodological premise of methodological holism makes it necessary to conceptualize the identities of discursive elements in accordance with the phenomenal dimensions contained in the PDA's theoretical framework. In practical terms, such theoretically holistic conceptualization of discursive elements can be related to "theoretical coding," which in terms of the grounded theory methodology (GTM) means that researchers make use of the phenomenal categories of their specific theoretical framework "to give meaning to data" (STRAUSS & CORBIN, 1990, p.42). Unfortunately, however, the PDA's theoretical framework offers only few indications of phenomenal categories, which would allow us to conceptualize the identities of discursive elements. GLASZE (2007, §43-50) has argued that we are only left with the possibility to combine theoretical coding with empirical coding. In terms of KELLE (2005, §22), empirical coding means that the identities of discursive elements are conceptualized in accordance with phenomenal categories that we "find suitable for the data under scrutiny." In other words, phenomenal categories conceptualizing discursive identities must be developed pragmatically or emerge "ad hoc during 'open coding'" (§12; cf. CHARMAZ, 2006, p.128). However, empirical coding of discursive elements must take into consideration the nodal points and discursive limits that are essential for the very constitution of discourse. In order to take into account that discourses are constituted by means of binary distinctions separating a discourse from other discourses, the applied phenomenal categories must be divided into positively and negatively charged identities. As regards the phenomenal categories of discursive identities, nodal points—embodying the idealized states of social being—and antagonistic others—symbolizing the threat to the ideal state of social being—constitute the first phenomenal category of ethical "values." As outlined above, the very presence and temporal stability of discourses depends on their capacity to install social subjects in discourse-specific subject roles. It is therefore reasonable to argue that "subjectivity" should constitute the second phenomenal category of discursive elements. Reflecting the binary division of discourse into positively and negatively charged elements, subject roles can be divided into "protagonists" conceived of as facilitating the attainment or sustainment of ethical values and "opponents," who endanger and impede the attainment of ethical values (MARTTILA, 2013a, pp.67ff.). Moreover, both categories of subject roles can be differentiated further with regard to "helpers" that are believed to support their respective aspirations and courses of action. It is also expedient to follow the example of GREIMAS (e.g. 1990 [1976], p.126f.) and distinguish between the subject roles of "destinators," which refer to the supposed instigators of ideas and courses of action, and "receivers" on whose behalf social subjects partake in particular activities (TITSCHER, WODAK, MEYER & VETTER, 1998, p.168). [33]

To recall: subject roles are associated with particular courses of both linguistic and non-linguistic actions and it is worthwhile noticing that institutions are sources of discursively motivated and authorized activities. It is therefore reasonable to assume that "activity" constitutes the third phenomenal category of discursive elements. Very pragmatically regarded, the phenomenal category of activity can be subdivided into "actions" that refer to activities undertaken by subjects or institutions, "interactions" that interlink subjects and their actions, "objects" that social subjects or institutions act upon or manipulate for the sake of achieving particular outcomes, "resources" that capacitate and equip social subjects to conduct particular actions, and finally the "strategies" that refer to courses of action that social subjects and institutions utilize for the sake of achieving the aspired results of action. The following Table 2 summarizes the phenomenal categories of discursive identities and describes their general phenomenal features.

The presented typologies of discursive relations and identities constitute a heuristic means to identify the paradigmatic and syntagmatic coherence of practices of articulation. On a more aggregated level, the discursive relations and identities open up the possibility to render visibility to the structural arrangement of discourse, which social subjects adhere to in their practices of articulation. The following Figure 1 offers a view of the kind of structural topography of a discourse that can be detected by making analytical use of the typologies of discursive relations and identities.

Figure 1: Structural topology of discourse. Please click here for an increased version of Figure 1. [35]

The earlier explication of the mutually constitutive relation of discourses and discursive materialities (Section 2.2) underlined that social subjects' sticking to a discourse is supported by subject roles and institutions, which restrict social subjects' enunciative possibilities and, by doing so, stabilize discourses. Following DEAN (1999, p.27), we can conceptualize this kind of system of reciprocal relations between discourse and discursive materiality in terms of a "discursive regime" (cf. GLYNOS & HOWARTH, 2007, pp.104ff.). Discursive regimes come into being as subsequent outcomes of counter-hegemonic acts that contested preceding regimes, and hegemonic acts that installed the present regime thus motivating discourse (see Figure 2). Hence, discursive regimes are altered by restructuring the discourses that motivate their very being. Bearing in mind the reciprocal relationship between discourse and practices of articulation, a discourse regime is maintained only as long as social subjects' practices of articulation sustain the discourse that constitutes the very rationality and meaningfulness of a discursive regime. GLYNOS and HOWARTH (2008, p.11) argue therefore that a discursive regime cannot prevail independent from the presence of corresponding "regime[s] of practice." In the following passages I will explicate how the reciprocal relations between social subjects' practices (of articulation), discourse, subject roles and institutions form spatially distinctive and temporally stable discursive regimes, and also elucidate how these reciprocal relations can be made accessible for empirical observations. [36]

Discursive regimes embrace discursively constructed subject roles that capacitate social subjects to conduct particular practices of articulation. Subject roles are "discursive positional[ities]" that are rationalized by discourses constituting and motivating their very being (TORFING, 1999, p.89). Discursive construction of subject roles means that subject roles are defined by means of distinctive structural arrangements of discursive elements. MARTTILA (2013a, pp.186ff.) has observed how the conception of the subject role of the "entrepreneur" has varied in accordance with the arrangement of discursive elements that define and motivate it. While the "neoclassical" economic discourse relates the entrepreneur to a strictly economic role and defines it by relating it to economic activities such as "risk-taking," "investment" and "competition," the "neoliberal" economic discourse extends the social function of the entrepreneur by referring to society-wide activities like "creativity," "innovation" and "will to change" (ibid.). In other words, we can detect subject roles by identifying discursive elements that symbolize the subject roles that different discourses assign to social subjects. [37]

However, discourses do not only name subject roles, but also define them by associating them with corresponding "actions, responsibilities, rights and duties, and equip [their holders] ... with capital, interests, intentions, goals and preferences" (TEUBNER, 2006, p.519). This means that discourses subjectivate social subjects by defining their subject roles and linguistic and non-linguistic practices associated with them. Following KELLER (2011a, p.256) subject roles become empirically visible in the form of linguistic and non-linguistic "discourse generated model practices" (my translation). Just like the mutual coherence of practices of articulation bears witness to them belonging to the same discourse, the mutual coherence of pragmatic actions serves as an indication of social subjects' belonging to the same subject roles. In other words, subject roles become empirically observable in the form of the relative homogeneity of pragmatic actions. Social subjects' adherence to a subject role is, at the same time, a source of continuous validation and sustainment of the discourse that defined this subject role. As the earlier elaboration of the process of sedimentation indicated, it is reasonable to assume that widely sedimented subject roles become subjectively performed without social subjects being fully aware of the underlying discursive origins and motivations of their subject roles. [38]

Similar to the relationship between discourses and subject roles, the relationship between discourses and institutions is likewise mediated by social subjects' practices. Generally speaking, institutions are rationalized by discourses "that bring them into being ... maintain them, and that may, at times lead them to disappear" (PHILLIPS, 2003, p.221). This means that the rationality of institutions and the meaningfulness of their specific sets of purposes, functions and operations derive from distinctive "discursive field[s] of representation ..." (KELLER, 2011a, p.141; my translation). Discourses do not only rationalize the presence of institutions, but also make the reality "amenable" for particular institutional interventions into the world (ROSE & MILLER, 1992, p.179). On their part, institutions validate discourses by retaining their function as the epistemic foundation for institutional operations. Moreover, institutions materialize discourses by giving them material effects. Such institutional materialization of discourse may, amongst other things, mean that institutions implement discursively motivated social hierarchies, enforce distribution of material resources such as financial assets, and symbolic resources such as access to socially influential subject roles. However, the relationship between discourses and institutions is also (to a quite considerable extent) mediated by subject roles. The functionality—i.e. practicability—of institutional operations depends on the presence of corresponding institutional populace, which accommodates itself to the functional needs of institutions and adapts social practices that support institutional operations (TEUBNER, 2006, p.519; TORFING, 1999, p.70). Social subjects' acceptance of particular interests, preferences and conducts brings them into accordance with institutional "actor fictions"—i.e. assumptions about social subjects' assumedly natural conceptions of the world, capacities, inclinations and likely future conducts (MARTTILA, 2013a, p.18). In other words, subject roles do not only constitute a link between social subjects' practices and institutions, but even more importantly a crucial source of corroboration of institutional operations. [39]

The reciprocal relationship between subject roles and institutions can be operationalized further by assuming it as consisting of an interplay of different types of "technologies of power" (LEMKE, 2007, p.50). Technologies of power embrace various kinds of means and strategies that subjugate social subjects "to effect by their own means or with help of others a certain number of operations on their own bodies and souls, thoughts, conduct, and way of being" (FOUCAULT, 1988, p.18). Moreover, technologies of power can be distinguished from institutionally exerted "technologies of the other," which subjectivize social subjects to particular subject roles and model practices linked with these, as well as "technologies of the self" such as chastity or life-long learning, which social subjects adapt and appropriate in order to concur with their subject roles (ibid.; cf. LEMKE 2007, pp.50ff.). Technologies of the other embrace all kinds of "programs, calculations, techniques, apparatuses, documents, and procedures" with "which authorities seek to embody and give effect to [their] ... ambitions" (ROSE & MILLER, 1992, p.175). Amongst other things, technologies of the other restrict social subjects' enunciative and pragmatic possibilities by regulating access to subject roles, enforcing social subjects' adherence to discourse-specific conceptions of the world and assessing their compliance with model practices. The notion of the technologies of the self refers to means and strategies that social subjects adapt in order to bring themselves into line with discursively motivated and institutionally enforced subject roles. In other words, technologies of the self refer to ways of subjective "self-appropriation" of their designated subject roles (ÅKESTRØM ANDERSEN, 2003, p.24). [40]

To conclude, discursive regimes consist of spatially distinctive and temporally relatively durable systems of reciprocal relations between discourses, subject roles and institutions. Even though discursive regimes are installed and maintained by means of discourses motivating their very presence, discursive regimes cannot be reduced to systems of meaning. The process of sedimentation in general and the materialization of discourse effectuated with subject roles and institutions in more particular, effectuate discourses in the form of material effects and increase their structural impact upon social life. The following Figure 2 provides a schematic overview of the elaborated phenomenal characteristics of discursive regimes.

Figure 2: Discursive regime. Please click here for an increased version of Figure 2. [41]

The aim of this concluding section is to make use of the above introduced framework of theoretical, methodological and phenomenal concepts to explicate the general guidelines for the conduct of empirical discourse analysis. While the first part (Section 5.1) is devoted to explaining the PDA's analytical stages, the second part (Section 5.2) elucidates the general criteria for the methods for generation and interpretation of empirical data. [42]

The above (in Section 3.1) described general epistemic bias of cognitions means that researchers can never encounter "brute facts ... only theoretically informed and culturally shaped descriptions of a discursively constructed reality" (TORFING, 2005, p.27). Hence, our objects of research are not "ready to be identified and mapped": they must be elaborately conceptualized and constructed (JØRGENSEN & PHILLIPS, 2002, p.144). All this being considered it is reasonable to suggest that the first stage of empirical discourse analysis consists of conscious co-construction of the research object. The methodological premise of methodological holism underlines the necessity to start empirical discourse analysis with the "theory-driven" co-construction of the research object. BOURDIEU (2008 [2002], p.92) argues that we are the more capable of observing the world in accordance with our consciously chosen epistemic horizon "the more completely [we] ... have objectivated [our] ... (social; academic, etc.) position[s] and interests." Theory-driven co-construction of research objects can be conceived of as consisting of two steps. Firstly, the lacking objective foundation of meaning identified by the PDA's theoretical framework underlines "the contingent formation of social phenomena" (TORFING, 2005, p.22). The a priori known discursively contingent constitution of socially meaningful objects opens up the possibility to deconstruct the research object's socially accepted implicitness and understand its presupposedly self-evidential social meaningfulness as originating from discursive attempts to define and fix its meaning (ÅKERSTRØM ANDERSEN, 2003, pp.57ff.). Secondly, deconstruction of socially sedimented common sense conceptions of the world opens up the possibility to think about the general phenomenal characteristics of discourses that constitute the conditions of possibility of any socially meaningful objects. The PDA's theoretical framework allows us to conceive of the general phenomenal characteristics of discourses involved in the creation of any socially meaningful object and think about research questions, analytical strategies, and research designs, which make it possible to lend visibility to the discourse that constitutes the condition of possibility of our research object. [43]

Theory-driven co-construction of the research object paves the way for the second stage of reconstruction, whose aim it is to lend visibility to the discourse, which we conceive to be responsible for the constitution of our research object in the focused spatio-temporal context. In accordance with FOUCAULT's (1977) distinction between "archaeological" and "genealogical" discourse analysis, analytical reconstruction can assume either a "synchronic" (i.e. archaeological) or a "diachronic" (i.e. genealogical) focus (cf. ÅKERSTRØM ANDERSEN, 2003, pp.20ff.). While the synchronic analysis focuses on the structural organization and material prerequisites of a relatively stable discourse, diachronic analysis is aimed at studying historical processes of sedimentation and reactivation, which together determine the formation and transformation of discourses. However, diachronic analysis cannot be dissociated with synchronic analysis. ÅKERSTRØM ANDERSEN (p.20) and FOUCAULT (1977, p.151) point out that we can identify and conceptualize the extent and direction of discursive change only if we know the general structural characteristics of the contested and displaced social structure. In other words, diachronic analysis necessitates analytical comparison of the points of disjunction that distinguish the contested and contesting discourses. After all, only the empirically manifested structural differences between different diachronically distinctive discourses may serve as empirical indications of anything similar to a discursive change. [44]

The first step of the discourse analysis begins with the compilation of the "corpus of data" embracing those practices of articulation, which our contextual background knowledge classified as essential for the constitution and sustainment of our research object's social meaningfulness.²⁾ The relevance and analytical sufficiency of the initially compiled corpus of data is proven subsequently by the observed empirical coherence of the practices of articulation. To avoid our analysis confirming merely conventional wisdoms, the compiled data must remain open for continuous revision in the light of newly emerging empirical findings. Amongst other things, our research can relativize the extent of assumed mutual coherence of practices of articulation and give reason to extend the corpus of data to embrace initially disregarded subjects. The second step of synchronic analysis consists of the contents-related analysis of the compiled data and aims at reconstructing the discourse the included social subjects adhered to and generated in their practices of articulation. The discursive relations and discursive identities constitute theoretical codes that allow us to investigate the extent of paradigmatic and syntagmatic coherence of articulations. Reflecting the general phenomenal characteristics of discourse, our empirical analysis must identify nodal points that constituted the overarching identity of the discourse under scrutiny and observe the logics and locations of discursive limits that separate our discourse from other discourses. While nodal points can be detected by means of "identify[ing] key words constantly referred to and used as supreme justification" of a particular relational arrangement of discursive elements (CEDERSTRÖM & SPICER, 2014, p.195), discursive limits become visible in the form of articulations made to explain and justify the reasons for the separation of the discursive elements linked by nodal points from other discursive elements (cf. GLASZE, 2007, §47-50). Having reconstructed the structural order of discourse, the aim of the third step is to identify the material prerequisites of the observed discourse. Reflecting the earlier described phenomenal characteristics of discursive materiality (see Sections 2.2 & 4.3), the analytical focus is set on the sedimented subject roles and institutions, which regulate social subjects' enunciative possibilities. Subject roles and institutions can be identified by means of two analytical strategies. We can make use of the overall phenomenal characteristics of discursive regimes (see Figure 2) to search for empirical indications of context-specific reciprocal relations between discourses, subject roles and institutions. The synchronic analysis has been completed once we have succeeded in reconstructing both the discourse, which at a given point in time constitutes the social meaningfulness of the research object, as well as the subject roles and institutions, which sustain this discourse by subjecting social subjects to conducting mutually coherent practices of articulation. [45]

Also the diachronic analysis can be conceived as consisting of three analytical stages. The first step makes use of the already accessible contextual knowledge to conceptualize the social change under review and its spatio-temporal coordinates and identify social subjects whose practices of articulation can be supposed to have instigated and enforced the observed change. This initial conceptualization of social change opens up the possibility to compile a corpus of data that embraces both the practices of articulation that have sustained the previous social status quo, and the practices of articulation that have induced and were involved in the generation of social change. The second step of diachronic analysis consists of a contents-related analysis of the compiled data and aims at the reconstruction of the discursive structures of contested and contesting discourses and the identification of their structural differences. Moreover, identification of the counter-hegemonic acts, which initiated the contestation of the prevailing discourse, and hegemonic acts, which installed a new hegemonic discourse, opens up the possibility to locate the historical and social origins of discursive change. The third step of diachronic analysis focuses on the reciprocal relations between contested and contesting discourses and their respective material conditions of possibility. The analytical aim is to identify the subject roles and institutions that sustained the stability of the contested discourse, the extent to which these subject roles and institutions were contested, modified and replaced by the counter-hegemonic acts, and the unprecedented subject roles and institutions, whose installment was motivated by the new hegemonic discourse. TORFING (1999, p.152) states that an analysis of discursive change must also take into account the subject roles and institutional affiliations of hegemonic agents because these may to some degree explain the credibility and subsequent social acceptance of the extent and direction of social change (cf. LACLAU, 1990, p.77). [46]

The absence of objective foundations of discourse, which is underlined by the PDA's theoretical framework, also means that analytically reconstructed discourses reflect context-specific historical origins and conditions of possibility. This contextually contingent character of discourse constrains the possibility to generalize the empirical findings to be applied to other social contexts. However, I would still like to argue that empirical findings should be collocated to enhance their subsequent reception and facilitate their comparison with the observations made in other research projects. As such, the third stage of empirical discourse analysis consists of collocation of empirical findings (cf. GLYNOS & HOWARTH, 2007, p.189). Collocation of research results consists of an accentuation of those phenomenal aspects, which we assume to reflect the differentia specifica of the discursive constitution of our research object, and which we also assume to be worthwhile comparing with the discursive construction of similar phenomena in other contexts. Following WEBER's (1949 [1904], p.90) instructions regarding the construction of "ideal types," collocation can be facilitated by the "one-sided accentuation" of some overarching phenomenal dimension or aspect that makes it possible to compare our research findings with the research conducted on similar social phenomena. Due to their function as middle-range concepts, the overall characteristics of popular and pluralist discourses and the general phenomenal characteristics of discursive regimes may serve as heuristic means to identify trans-contextually occurring discourses as well as forms and directions of discursive change, etc. Also the different types of counter-hegemonic and hegemonic acts elaborated by NONHOFF (2006, 2007) can serve as empirically generalizable analytical categories to identify processes of historical change recurring from one context to another. Following the approach of MARTTILA (2013a, 2013b), we can also detract different subject roles from empirical data featuring their typical and potentially cross-contextually generalizable phenomenal characteristics. GLYNOS and HOWARTH (2007, pp.104ff.) point out the possibility to condensate empirical findings to identify "logics" such as "economization," "bureaucratization" and "scientification," which we understand as characterizing what goes on in our context of analysis. The following Table 3 resumes the PDA's analytical stages along with their respective aims and strategies.

The above outlined methodological premise of methodological holism implies that we must ensure that "theory, methodology and methods" form an internally coherent "aesthetic context" in empirical research (DIAZ-BONE, 2006, §6; my translation). In order to adhere to methodological holism we must accept that our theoretical framework instructs us how "reality manifests itself and how it can be investigated, and how it cannot" (ibid.; my translation). Discourse analysts cannot freely choose among the prevailing methods for gathering and interpreting data because each of these methods has been developed against the background of distinctive assumptions concerning the phenomenal characteristics of the objects under scrutiny. GAUKROGER (1976, p.222) underlines that scientific methods do not "only describe," but actually "produce phenomena." In order to realize the methodological premise of methodological holism in empirical research, we need to consciously reflect upon the extent to which different methods for gathering and interpreting data actually concur with the a priori conceived phenomenal characteristics of discourse and discursive structuration of social life. With the exception of hints given in the works of GLASZE (2007) and MARTTILA (2013a), there is hardly any information about the scientific methods that could facilitate the attainment of methodologically holistic empirical PDA. To serve the purpose of methodologically holistic empirical research, the following two sections will make a recourse to the theoretical, methodological and phenomenal foundations of the PDA to identify consistent methods for "generating" (Section 5.2.1) and "interpreting" (Section 5.2.2) data. [48]

Post-foundational discourse analysis consists of a theoretically informed approach to empirical research, whose primary aim is to lend empirical visibility to all parts of discourses constituting and structuring social life. Bearing in mind that practices (of articulation) are conducted on behalf of some discourse, which also regulates social subjects' meaningful practices, discourse analysts need to locate the set of social practices that install and effectuate a particular discourse. Similar to the FOUCAULTian (archaeological) discourse analysis (e.g. FOUCAULT, 2009 [1969]), PDA also needs to begin with a preliminary outlining of the "archive"—i.e. the body of practices—, which is revisable during the course of concrete empirical research and which we observe to adhere and give effect to the discourse responsible for the formation of our research object (ÅKERSTRØM ANDERSEN, 2003, p.13). [49]

In accordance with the conventional wisdom of qualitative research methodology, discourse analysts can actively "generate" their empirical data by means of research interviews and field observations or "collect" data by relying on existing textual material (cf. DIEKMANN, 2014 [1995], pp.434ff.; FLICK, 2009, pp.128ff.). Here is not the right place to delve further into the general methodical questions and problems related to methods of research interviews, field observations and collecting texts. Instead, it is crucial to notice that any of these methods for gathering data is compatible with PDA as long as they are not intended to provide "immediate access" to social subjects' own conscious conceptions and understandings of the motives for their "actions and routines," but are in line with the second-order hermeneutics and render visibility to the structural organization of the discourse effectuated and sustained by these practices (FLICK, 2009, p.128; my translation). Moreover, considering that discourses do not only refer to a structural organization of language, but also embrace any relational structure of socially meaningful objects established, retained and changed by means of social practices, there is hardly any reason to constrain empirical discourse analysis to the study of textual material only. Considering that any social practice must be regarded as constituting a "fragment" of a more encompassing discourse (JÄGER & MAIER, 2009, p.47) that restricts social subjects' enunciative possibilities, we can then make use of methods of field observations to detect non-linguistic patterns of practice. As GUEST, NAMEY and MITCHELL (2013, p.78) point out, methods of field observation can render visibility to patterns of practice because they let us identify patterns of practice by empirical "counting [of] the frequency and/or intensity of specific behaviours or events or mapping the social composition and action of a particular scene." Also, methods of research interviews can be used to render visibility to routinely conducted patterns of practices. However, bearing in mind the previous critique of the methodological position of first-order hermeneutics, research interviews cannot be utilized to gain access to social subjects' consciously reflectible "life-worlds" (KVALE, 1983, p.174). In accordance with second-order hermeneutics, a research interview can constitute a means to make the interviewees provide the interviewer access to the "otherwise closed areas of practice ..." (DEPPERMANN, 2013, §3; my translation). [50]

The earlier elucidated methodological premise of methodological holism indicated that we need to gather empirical data in consistency with the already known phenomenal properties of the research object. Generally speaking, the analytical validity of the empirical data depends on their capacity to lend visibility to the discourse, which we assume to have been responsible for the formation of our research object. Even though social subjects' practices of articulation are always conducted on behalf of some discourse (see Sections 3.1, 4.1), we cannot identify the discourse that a social subject adheres to from his/her singular articulation. To reconstruct the discourse, which constitutes the horizon of possibility and intelligibility of a particular articulation, we need to detect the points of (syntagmatic and paradigmatic) coherence between articulations conducted by social subjects, whom we assume to possess subject roles in the studied discourse. Our sample must therefore cover a considerable part of the population, which we have reason to assume to adhere to the same discourse, or a representative sample of the totality of articulations conducted by this population. This means that gathering data is necessarily preceded by an initial conceptualization of spatio-temporal coordinates of the discourse under review and its (re-) producing population of subjects. [51]

However, the earlier described absence of any objective foundations of discourse means that discourses are not naturally limited to a given population, social context or institution (cf. ÅKERSTRØM ANDERSEN, 2003, p.13). Hence, the initial sample remains necessarily tentative and requires further sampling as our empirical knowledge about the spatio-temporal coordinates and the subjects belonging to it increases. The initial sample is necessarily based on some kind of discourse analysis predating secondary data such as previous research conducted on our research object. We can draw on secondary data to identify the common sense conceptions, which inform social subjects about the social meaningfulness associated with our research object, and thereafter retrace these common sense conceptions to social subjects installing and reproducing it. Having located the initial sample of data and reconstructed its characteristic structure of discourse, we can then, step by step, extend this sample of data to embrace further articulations. There are several possible methods to extend the data in order to obtain a fuller picture of the studied discourse. Firstly, we can search the compiled data for references to other subjects conceived of as influential and inspirational and, in this manner, extend the sample of data to embrace the articulations conducted by these subjects (cf. SOULLIERE, BRITT & MAINES, 2001, p.254). Secondly, we can make use of qualitative interviews with the already included social subjects in combination with network analysis to increase our information about the entire "network" of social subjects involved in the generation of a particular conception of our research object (BERNHARD, 2014, §11f.; CLARKE, 2003, p.559). Thirdly, we can also follow the example of GLASZE (2007, §25ff.) and assess the mutual coherence of different samples of data by methods of quantitative lexicometric analysis, which allows us to search the selected corpus of data for relatively frequent co-occurrences of lexical elements. The frequent co-occurrence of particular lexical elements (e.g. "entrepreneur," "innovation," "knowledge-based economy") serves as an empirical indication of the mutual structural coherence of different data and gives us reason to suggest that these data articulating subjects adhere to the same discourse (ibid.). Having collected sufficient empirical indications of the structural coherence of texts, we can continue with a more precise qualitative interpretation of the discursive relations and identities. [52]

As concerns the analytical validity of the gathered data, the absence of any objective foundations and natural locations of discourse together with the epistemically biased nature of empirical analysis impede/restrict the possibility to assess the sufficiency of data in terms of "theoretical saturation." The concept of theoretical saturation originates from the context of the GTM and means that gathering new data is completed when "fresh data no longer sparks new theoretical insights" (CHARMAZ, 2006, p.113). The absence of objective foundations and limitations of discourse and the epistemic relativism of (scientific) interpretations give reason to suggest that the sufficiency of the gathered data must be assessed pragmatically with regard to its empirical plausibility. The plausibility of our data can be ensured by means of two consecutive strategies. Firstly, our sample is plausible if we can manifest that the subjects, whose practices of articulation we analyze, can be reasonably assumed to possess subject roles within the studied discourse. Secondly, however, it does not only suffice to identify the relevant group of social subjects, but we also need to ensure that their practices of articulations are representative enough to lend visibility to the entire discourse under review. [53]

According to methodological holism, the a priori known phenomenal characteristics of the studied objects must determine "how ... reality manifests itself and how it can be investigated, and how not" (DIAZ-BONE, 2006, §5; my translation). Hence, our knowledge of the phenomenal characteristics of discourse and their material conditions of possibility should provide heuristic "frames from which to view realities" (CHARMAZ, 2006, p.128). The theoretical and methodological premises of the PDA underlined that discourse refers to a relational arrangement of elements that constitutes the social meaningfulness of an object. Even though a discourse has a meaning-generating function, discourse cannot be equated with either the subjectively possessed and inter-subjectively shared body of knowledge, or with the FOUCAULTian "episteme"—i.e. historically specific "horizons of perception" (DIAZ-BONE, 2007, §21; my translation). To recall: meaning (signified) is contingent on the relations that social subjects establish between meaning-containing elements (signifiers). There are two reasons why PDA can only study how meanings are produced, but not what the produced meanings actually are. Firstly, discursive sedimentation and the resulting subjection of social subjects and their roles lead to a situation, in which the social meaningfulness of an object is separated from social subjects' self-conscious conceptions of this object. Secondly, the epistemically biased nature of perception implies that the analyst is unlikely to achieve an access to the sedimented meanings that rationalize different social practices. Both these epistemological obstacles clearly show that empirical analysis of data is not focused on meaning-contents but on meaning-generating relations. Moreover, these epistemological obstacles reject the analytical usefulness and appropriateness of different types of "ethnomethodological" methods because these assume social subjects' actions to be motivated and based upon an intersubjectively shared "conjunctive space of experience" (BOHNSACK, 2004 [2000], p.218). Ethnomethodological methods such as participatory observation, narrative interviews, documentary analysis and conversation analysis are motivated by the epistemological assumption that they can lend visibility to subjectively conscious and intersubjectively shared "experiential spaces" (pp.214ff.; MAYNARD & CLAYMAN, 1991, pp.396ff.). Moreover, ethnomethodological methods are incompatible with the PDA also because they prefer the methodological position of first-order hermeneutics that necessitates "neither prior description, nor empirical generalization, nor formal specification of variable elements and their analytic relations" (p.387). Instead of using a priori defined theoretical codes to achieve an epistemological break with social subjects' conscious self-conceptions in ethnomethodological research the use of analytically sensitizing concepts should be "delayed until the situated meanings of concepts are discovered" (DENZIN, 1969, p.926). [54]

To recall, the primary analytical aim of the PDA is to give visibility to discourses and discursive materialities. While discourses become empirically observable in the form of "temporal fixations of elements" in social subjects' practices of articulation (GLASZE, 2007, §37), discursive materialities are observable in the form of subject roles and institutions, which together regulate social subjects' enunciative possibilities. The distinction between discourse and discursive materiality implicates that empirical discourse analysis involves "multimodal" research because it embraces interpretation of both linguistic and non-linguistic materials (VAN LEEUWEN, 2005, pp.5f.). The multimodal character of the PDA's research object requires corresponding recourse to a multimethod approach to empirical research. Even though texts contain indexical indications of subject roles and institutions constituting discursive materiality, their analysis is likely to require the use of non-text-centered methods such as "situative mapping" offered in "situational analysis" (CLARKE, 2003, 2005). In accordance with ERZBERGER and KELLE (2003, p.461), the multi-dimensionality (and -modality) of research objects requires the use of several methods, which together "provide different pictures of [the] ... object [and] ... yield a fuller and more complete picture of the phenomenon concerned." [55]

Demonstrating that practices of articulation are not constrained to linguistic practices, HOWARTH (2004b, p.265) has made the case for "creat[ive] misappl[ication of] the concept of discourse to encompass all dimensions of social reality and not just the usual practices of speaking, writing and communicating." Notwithstanding, GLASZE (2007, §27) has a point when he argues that discourses are most readily available in texts. In other words, we can use text-centered methods to discover discourses in textual material. Considering that discourses consist of relational arrangements of discursive relations and identities, we need to use methods of interpretation that allow us in the first step to discover the relational structure of singular articulations, and in the second step coalesce these partial structures into an all-embracing common structure of discourse. Methods developed in the context of structuralist approaches to narrative analysis (e.g. SOMERS, 1994, 1995) are not only consistent with the relational epistemology of meaning, but they also provide various strategies to identify mutually consistent "narrative patterns" (SOMERS, 1994, p.606). We can make use of such "narrative patterns" to identify for a social group characteristic narrative structures and "meta-narratives" (SOMERS, 1995, p.135), which again reveal this group to adhere to the same discourse (cf. BERNHARD, 2014; TITSCHER et al., 1998, p.206). Indeed, a number of empirical analyses aligned with the PDA (CARPENTIER & SPINOY, 2008; GLASZE, 2007; MARTTILA, 2013a) have already manifested the applicability of the methods of structural narrative analysis in empirical research. In order to adhere to the methodologically holistic interpretation of articulations we must have recourse to theoretical codes derived from the theoretical (Section 2) and phenomenal (Section 4.4.3) framework of the PDA, which serve the function of a "coding paradigm" in empirical research (KELLE, 2005, §22). Analytical use of a priori defined theoretical codes means that the interpretation of articulations does not follow ad hoc in the form of generation of "theoretical codes and coding families" that we "find suitable for the data under scrutiny," but means using the codes that are "rooted in [our] ... own theoretical tradition" (ibid.). [56]

There are at least three methods available to analyze the discursive materiality. Firstly, we can retrace social subjects' practices of articulation regulating subject roles and institutions by means of locating practices of articulation defining and motivating them. This analytical strategy was described above in terms of "diachronic" analysis (Section 5.1.2). Secondly, we may also make use of the analytical methods developed in "enunciative pragmatics" (e.g. ANGERMÜLLER, 2011; ZIENKOWSKI, 2012). ANGERMÜLLER (2011) and ZIENKOWSKI (2012) have argued that the so called "enunciative markers" can be used to locate indexical indications to the situational context of the subject of enunciation in practices of articulation. Amongst other things, enunciative markers can be used to lend visibility to "interpretive schemes which represent the relevant subject positions of discourses" (ANGERMÜLLER, 2011, p.2993). Thirdly, we can also detect discursive materialities by using the methods developed in "situational analysis" (e.g. CLARKE, 2005): E.g., the method of situative mapping provides an analytical means to map the "human and nonhuman, material, and symbolic/discursive elements," which the analyst conceives of as "making a difference" with regard to social subjects' capacities to act and pursue practices of articulation (p.87). DIAZ-BONE (2013, §17) underlines that situational analysis offers several methods such as the different types of "situational," "social worlds/arenas" and "positional maps" to identify and "grasp the actual resources of power," which together influence social subjects' enunciative capacities (my translation). The analytical benefit of a situational analysis is that its methods of mapping can be used against different theoretical backgrounds. CLARKE (2005, pp.87, 89) underlines that situational analysis does not define the phenomena of empirical interest but leaves this decision to the analyst and his/her particular research interests. [57]

The aim of this account has been to solve the PDA's methodological deficit by constructing its research program. Reflecting the initial conceptualization of a research program's phenomenal characteristics, this account has developed a system of theoretical, methodological and phenomenal concepts, which, in their turn, served as a heuristic framework to identify and develop analytical strategies and methodical instructions for empirical discourse analysis consistent with them. Amongst other things, empirical discourse analysis has been conceived of as requiring the theory-driven co-construction of the research object, necessitating analytical strategies that are consistent with the phenomenal characteristics and constitutive conditions of discourse, and presuppose that scientific methods for generation and interpretation of empirical data are selected and utilized in accordance with the methodological premises of the second-order hermeneutics and methodological holism. The presented research program functions also as a frame of communication because it does not only make it easier for discourse analysts to carry out their individual research projects, but also provides them with a frame of reference to discuss and compare their research results and participate in identification and solving of methodical problems. It must be mentioned that the outlined research program itself has the function of a discourse because it makes it possible to distinguish research interests, methods and scientific practices commensurable and incommensurable with the PDA. Nevertheless, the intention of this account has not been to establish any hegemonic and subsequently incontestable methodical instructions. According to LAKATOS (1970, p.137) the "heuristic power" of a research program depends to great extent on its capacity to animate empirical research. Hence, the validity and usefulness of the outlined research program remains to be assessed by the scientific community. In the best case, this research program inspires similar accounts, which will then involve an ongoing debate on the PDA's methodological position. [58]

I would like to thank Vincent GENGNAGEL and Andreas KALLERT as well as the two anonymous peer-reviewers for their critical and helpful comments on earlier drafts of this article.

1) "Essex School in Discourse Analysis" refers to the distinctive set of theoretical ideas elaborated in the pioneering works of Ernesto LACLAU and Chantal MOUFFE (e.g., 2001 [1985]) and their subsequent adaption in the works of Jason GLYNOS (2001), David HOWARTH (2000), Oliver MARCHART (2007), Martin NONHOFF (2006), Yannis STAVRAKAKIS (2007) and Jacob TORFING (1999). The concept of post-foundational discourse analysis has been used by CEDERSTRÖM and SPICER (2014), MARCHART (2007) and MARTTILA and GENGNAGEL (2015). The more commonly used label of "post-structural discourse analysis" appears inappropriate because the PDA is not located beyond the structuralist paradigm in social sciences. <back>

2) Methods for generation and interpretation of data consistent with the PDA are described further in Sections 5.2.1 and 5.2.2. <back>

Åkerstrøm Andersen, Niels (2003). Discursive analytical strategies: Understanding Foucault, Kosellack, Laclau, Luhmann. Bristol: Policy Press.

Alasuutari, Pertti (1996). Theorizing in qualitative research: A cultural studies perspective. Qualitative Inquiry, 4(2), 371-384.

Alvesson, Mats & Sköldberg, Kaj (2000). Reflexive methodology: New vistas for qualitative research. London: Sage.

Angermüller, Johannes (2007). Nach dem Strukturalismus: Theoriediskurs und intellektuelles Feld in Frankreich. Bielefeld: Transcript.

Angermüller, Johannes (2011). From the many voices to subject positions in anti-globalization discourse: Enunciative pragmatics and the polyphonic organization of subjectivity. Journal of Pragmatics, 43, 2992-3000.

Bernhard, Stefan (2014). Identitätskonstruktionen in narrativen Interviews. Ein Operationalisierungsvorschlag im Anschluss an die relationale Netzwerktheorie. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 15(3), Art. 1, http://nbn-resolving.de/urn:nbn:de:0114-fqs140311 [Date of Access: October 15, 2014].

Bohnsack, Ralf (2004 [2000]). Group discussions and focus groups. In Uwe Flick, Ernst von Kardorff & Ines Steinke (Eds.), A companion to qualitative research (pp.214-221). London: Sage.

Boucher, Geoff (2008). The charmed circle of ideology: A critique of Laclau, Mouffe, Butler and Žižek. Melbourne: Re-Press.

Bourdieu, Pierre (2005). Was heißt sprechen? Zur Ökonomie des sprachlichen Tausches. Wien: Braumüller.

Bourdieu, Pierre (2008 [2002]). Science of science and reflexivity. Cambridge: Polity Press.

Bowman, Paul (2007). Post-Marxism versus cultural studies: Theory, politics and intervention. Edinburgh: Edinburgh University Press.

Carpentier, Nico & Spinoy, Erik (2008). Introduction: From the political to the cultural. In Nico Carpentier & Erik Spinoy (Eds.), Discourse theory and cultural analysis: Media, arts and literature (pp.1-28). Cresskill, NJ.: Hampton Press.

Cederström, Carl & Spicer, André (2014). Discourse of the real kind: A post-foundational approach to organizational discourse analysis. Organization, 21(2), 178-205.

Charmaz, Kathy (2006). Constructing grounded theory: A practical guide through qualitative analysis. London: Sage.

Clarke, Adele E. (2003). Situational analysis: Grounded theory after the postmodern turn. Symbolic Interaction, 26(4), 553-576.

Clarke, Adele E. (2005). Situational analysis: Grounded theory after the postmodern turn. Thousand Oaks, CA: Sage.

Clegg, Stewart R.; Courpasson, David & Phillips, Nelson (2006). Power and organizations. London: Sage.

Costache, Adrian (2011). For post-historicist philosophy of history: Beyond hermeneutics. Logos & Episteme, 2(4), 489-505.

Crowell, Steven Galt (1990). Husserl, Heidegger, and transcendental philosophy: Another look at the Encyclopaedia Britannica article. Philosophy & Phenomenological Research, 50(3), 501-551.

Dean, Mitchell (1999). Governmentality: Power and rule in modern society. London: Sage.

Denzin, Norman K. (1969). Symbolic interactionism and ethnomethodology: A proposed synthesis. American Sociological Review, 34(6), 922-934.

Deppermann, Arnulf (2013). Interview als Text vs. Interview als Interaktion. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 14(3), Art. 13, http://nbn-resolving.de/urn:nbn:de:0114-fqs1303131 [Date of Access: February 5, 2015].

Derrida, Jacques (1977). Signature, event, context. In Jacques Derrida (Ed.), Limited Inc (pp.1-23). Evanston, IL: Northwestern University Press.

Derrida, Jacques (1981a [1972]). Positions. Chicago, IL: The University of Chicago Press.

Derrida, Jacques (1981b [1967]). Writing and difference. London: Routledge & Kegan Paul.

Diaz-Bone, Rainer (2006). Zur Methodologisierung der Foucaultschen Diskursanalyse. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(1), Art. 6, http://nbn-resolving.de/urn:nbn:de:0114-fqs060168 [Date of Access: March 3, 2010].

Diaz-Bone, Rainer (2007). Die französische Epistemologie und ihre Revisionen: Zur Rekonstruktion des methodologischen Standortes der Foucaultschen Diskursanalyse. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 8(2), Art. 6, http://nbn-resolving.de/urn:nbn:de:0114-fqs0702241 [Date of Access: March 3, 2010].

Diaz-Bone, Rainer (2010). Kulturwelt, Diskurs und Lebensstil: Eine Diskurstheoretische Erweiterung der Bourdieuschen Distinktionstheorie (2nd ed.). Wiesbaden: VS Verlag für Sozialwissenschaften.

Diaz-Bone, Rainer (2013). Review Essay: Situationsanalyse—Strauss meets Foucault?. Forum Qualitative Sozialforschung/ Forum: Qualitative Social Research, 14(1), Art. 11, http://nbn-resolving.de/urn:nbn:de:0114-fqs1301115 [Date of Access: June 13, 2014].

Diekmann, Andreas (2014 [1995]). Empirische Sozialforschung: Grundlagen, Methoden, Anwendungen (9th ed.). Reinbek: Rowohlt.

Dyrberg, Torben Bech (1997). The circular structure of power: Politics, identity, community. London: Verso.

Erzberger, Christian & Kelle, Udo (2003): Making inferences in mixed methods: The rules of integration. In Abbas Tashakkori & Charles Teddlie (Eds.), Handbook of mixed methods for the social and behavioural sciences (pp.457-490). Thousand Oaks, CA: Sage.

Flick, Uwe (2009). Sozialforschung: Methoden und Anwendungen: Ein Überblick für die BA-Studiengänge. Reinbek: Rowohlt.

Foucault, Michel (1977). Nietzsche, genealogy and history. In Donald F. Bouchard (Ed.), Language, counter-memory, practice: Selected essays and interviews by Michel Foucault (pp.139-164). Ithaca, NY: Cornell University Press.

Foucault, Michel (1988). Technologies of the self. In Martin H. Luther, Huck Gutman & Patrick H. Hutton (Eds.), Technologies of the self: A seminar with Michel Foucault (pp.16-49). Amherst, MA: University of Massachusetts Press.

Foucault, Michel (2009 [1969]). The archaeology of knowledge. London: Routledge.

Gaukroger, Stephen W. (1976). Bachelard and the problem of epistemological analysis. Studies in History and Philosophy of Science Part A, 7(3), 189-244.

Glasze, Georg (2007). Vorschläge zur Operationalisierung der Diskustheorie von Laclau und Mouffe in einer Triangulation von lexikometrischen und interpretativen Methoden. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 8(2), Art. 14, http://nbn-resolving.de/urn:nbn:de:0114-fqs0702143 [Date of Access: April 17, 2008].

Glynos, Jason (2001). The grip of ideology: A Lacanian approach to the theory of ideology. Journal of Political Ideologies, 6(2), 191-214.

Glynos, Jason & Howarth, David (2007). Logics of critical explanation in social and political theory. London: Routledge.

Glynos, Jason & Howarth, David (2008). Critical explanation in social science: A logics approach. Swiss Journal of Sociology, 54(1), 5-35.

Greimas, Algirdas Julien (1990 [1976]). The social sciences: A semiotic view. Minneapolis, MN: University of Minnesota Press.

Guest, Greg; Namey, Emily E. & Mitchell, Marilyn L. (2013). Collecting qualitative data: A field manual for applied research. London: Sage.

Gumbrecht, Hans Ulrich (2004). Production of presence: What meaning cannot convey. Stanford, CA: Stanford University Press.

Halliday, Michael A.K. (1985). An introduction to functional grammar. Sydney: Edward Arnold.

Heidegger, Martin (2008 [1923]). Ontology: The hermeneutics of facticity. Bloomington: Indiana University Press.

Howarth, David (2004a). Towards a Heideggerian social science: Heidegger, Kisiel and Weiner on the limits of the anthropological discourse. Anthropological Theory, 4(2), 229-247.

Howarth, David (2004b). Hegemony, political subjectivity, and radical democracy. In Simon Critchley & Oliver Marchart (Eds.), Laclau: A critical reader (pp.256-276). New York: Routledge.

Howarth, David (2005). Applying discourse theory: The method of articulation. In David R. Howarth & Jacob Torfing (Eds.), Discourse theory in European politics: Identity, policy and governance (pp.316-349). Basingstoke: MacMillan.

Howarth, David (2006). The method of articulation. In Margo van den Brink & Tamara Metze (Eds.), Words matter in policy and planning: Discourse theory and method in social science (pp.23-42). Utrecht: Labor Grafimedia.

Jäger, Siegfried & Maier, Florentine (2009). Theoretical and methodological aspects of Foucauldian critical discourse analysis and dispositive analysis. In Ruth Wodak & Michael Meyer (Eds.), Methods of critical discourse analysis (2nd ed., pp.34-61). London: Sage.

Jørgensen, Marianne & Phillips, Louise (2002). Discourse analysis as theory and method. London: Sage.

Kelle, Udo (2005). "Emergence" vs. "forcing" of empirical data? A crucial problem of "grounded theory" reconsidered. Forum Qualitative Sozialforschung/ Forum: Qualitative Social Research, 6(2), Art. 27, http://nbn-resolving.de/urn:nbn:de:0114-fqs0502275 [Date of Access: March 17, 2013].

Keller, Reiner (2011a). Wissenssoziologische Diskursanalyse: Grundlegung eines Forschungsprogramms (3rd ed.). Wiesbaden: VS Verlag für Sozialwissenschaften.

Keller, Reiner (2011b). The sociology of knowledge approach to discourse (SKAD). Human Studies, 34, 43-65.

Kisiel, Theodore (2010). Hermeneutics of facticity. In Bret W. Davis (Ed.), Martin Heidegger: Key concepts (pp.17-32). Acumen: Durhamn.

Kvale, Steinar (1983). The qualitative research interview: A phenomenological and hermeneutical mode of understanding. Journal of Phenomenological Psychology, 14(2), 171-196.

Laclau, Ernesto (1980). Populist rupture and discourse. Screen Education, 34, 87-93.

Laclau, Ernesto (1990). New reflections of the revolution of our time. London: Verso.

Laclau, Ernesto (1993). Discourse. In Robert E. Goodin & Philip Petit (Eds.), Companion to contemporary political philosophy (pp.431-437). Oxford: Basil Blackwell.

Laclau, Ernesto (2004). Glimpsing the future. In Simon Critchley & Oliver Marchart (Eds.), Laclau: A critical reader (pp.279-328). New York: Routledge.

Laclau, Ernesto & Mouffe, Chantal (1990). Post-Marxism without apologies. In Ernesto Laclau (Ed.), New reflections of the revolution of our time (pp.97-132). London: Verso.

Laclau, Ernesto & Mouffe, Chantal (2001 [1985]). Hegemony and socialist strategy: Towards a radical democratic politics (2nd ed.). London: Verso.

Lakatos, Imre (1968). Criticism and the methodology of scientific research programmes. Proceedings of the Aristotelian Society, 69, 149-186.

Lakatos, Imre (1970). Falsification and the methodology of scientific research programmes. In Imre Lakatos & Alan Musgrave (Eds.), Criticism and the growth of knowledge (pp.91-196). Cambridge: Cambridge University Press.

Lakatos, Imre (1999). Lecture 8: The methodology of scientific research programmes. In Matteo Motterlini (Ed.), For and against method (pp.96-108). Chicago, IL: University of Chicago Press.

Law, John (2004). After method: Mess in social science research. London: Routledge.

Lemke, Thomas (2007). An indigestible meal? Foucault, governmentality and state theory. Distinktion, 8(2), 43-65.

Lemke, Thomas (2015). New materialisms: Foucault and the "government of things". Theory, Culture & Society, 32(4), 3-25.

Luhmann, Niklas (1997). Gesellschaft der Gesellschaft (Vol. I). Frankfurt/M.: Suhrkamp.

Marchart, Oliver (2007). Post-foundational political thought. Edinburgh: Edinburgh University Press.

Marttila, Tomas (2010). Constrained constructivism in post-structural discourse analysis. Sociologia Internationalis, 48(1), 91-112.

Marttila, Tomas (2013a). Culture of enterprise in neoliberalism: Specters of entrepreneurship. London: Routledge.

Marttila, Tomas (2013b). Whither governmentality research? A case study of the governmentalization of the entrepreneur in the French epistemological tradition. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 14(3), Art.10, http://nbn-resolving.de/urn:nbn:de:0114-fqs1303106 [Date of Access: October 1, 2013].

Marttila, Tomas (2015/Forthcoming). Post-foundational discourse analysis: From political difference to empirical research. London: Palgrave MacMillan.

Marttila, Tomas & Gengnagel, Vincent (2015). Post-foundational discourse analysis and the impasses of critical inquiry. Journal for Discourse Studies / Zeitschrift für Diskursforschung, 3(1), 50-69.

Maynard, Douglas W. & Clayman, Steven E. (1991).The diversity of ethnomethodology. Annual Review of Sociology, 17, 385-418.

Nadel, Siegfried F. (1962). The theory of social structure (2nd ed.). London: Cohen & West Ltd.

Nonhoff , Martin (2006). Politischer Diskurs und Hegemonie: Das Projekt Soziale Marktwirtschaft. Bielefeld: Transcript.

Nonhoff , Martin (2007). Politische Diskursanalyse als Hegemonieanalyse. In Martin Nonhoff (Ed.), Diskurs, radikale Demokratie, Hegemonie: Zum politischen Denken von Ernesto Laclau und Chantal Mouffe (pp.173-193). Bielefeld: Transcript.

Phillips, Nelson (2003). Discourse or institution? Institutional theory and the challenge of critical discourse. In Robert Westwood & Stewart Clegg (Eds.), Debating organization: Point-counterpoint in organization studies (pp.220-231). Malden, MA: Blackwell.

Purvis, Trevor & Hunt, Alan (1993). Discourse, ideology, discourse, ideology, discourse, ideology... The British Journal of Sociology, 44(3), 473-499.

Rappert, Brian (2007). On the mid range: An exercise in disposing (or minding the gaps). Science, Technology & Human Values, 32(6), 693-712.

Rorty, Richard (1979). Philosophy and the mirror of nature. Princeton, NJ: Princeton University Press.

Rose, Nikolas & Miller, Peter (1992). Political power beyond the state: Problematics of government. The British Journal of Sociology, 43(2), 173-205.

Saussure, Ferdinand de (1959 [1916]). Course in general linguistics. London: Peter Owen.

Silverman, David (1997). Interpreting qualitative data: Methods for analysing talk, text and interaction. London: Sage.

Somers, Margaret S. (1994). The narrative constitution of identity: A relational and network approach. Theory & Society, 23(5), 605-649.

Somers, Margaret S. (1995). What is political or cultural about political culture and the public sphere? Toward a historical sociology of concept formation. Sociological Theory, 13(2), 113-144.

Soulliere, Danielle; Britt, David W. & Maines, David R. (2001). Conceptual modeling as a toolbox for grounded theorists. The Sociological Quarterly, 42(2), 253-269.

Stäheli, Urs (2001). Die politische Theorie der Hegemonie: Ernesto Laclau und Chantal Mouffe. In Andre Brodocz & Gary S. Schaal (Eds.), Politische Theorien der Gegenwart II: Eine Einführung (pp.194-223). Opladen: Leske + Budrich.

Stäheli, Urs (2004). The competing figures of the limit. In Simon Critchley & Oliver Marchart (Eds.), Laclau: A critical reader (pp.226-240). London: Routledge.

Stavrakakis, Yannis (2007). The Lacanian left: Psychoanalysis, theory, politics. Edinburgh: Edinburgh University Press.

Strauss, Anselm L. & Corbin, Juliet M. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage.

Teubner, Günther (2006). Rights of non-humans: Electronic agents and animals as new actors in politics of law. Journal of Law and Society, 33(4), 497-521.

Tiles, Mary (1984). Bachelard: Science and objectivity. Cambridge: Cambridge University Press.

Titscher, Stefan; Wodak, Ruth; Meyer, Michael & Vetter, Eva (1998). Methoden der Textanalyse. Opladen: Westdeutscher Verlag.

Torfing, Jacob (1999). New theories of discourse: Laclau, Mouffe and Žižek. Oxford: Blackwell.

Torfing, Jacob (2005). Discourse theory: Achievements, arguments and challenges. In David R. Howarth & Jacob Torfing (Eds.), Discourse theory in European politics: Identity, policy and governance (pp.1-32). Basingstoke: Palgrave Macmillan.

Weber, Max (1949 [1904]). Objectivity in social science and social policy. In Edward A. Shils & Henry A. Finch (Eds.), The methodology of the social science (pp.49-112). New York: Free Press.

Wodak, Ruth & Meyer, Michael (Eds.) (2009). Methods of critical discourse analysis (2nd ed.). London: Sage.

Žižek, Slavoj (1990). Beyond discourse analysis. In Ernesto Laclau (Ed.), New reflections of the revolution of our time (pp.249-260). London: Verso.

Žižek, Slavoj (1995). Looking awry: An introduction to Jacques Lacan through popular culture. Cambridge, MA: MIT Press.

Zienkowski, Jan (2012). Overcoming the post-structuralist methodological deficit: Metapragmatic markers and interpretative logics in a critique of the Bologna process. Pragmatics, 22(3), 501-534.

Tomas MARTTILA, born 1977, PhD in sociology, Master of Social Science. 1997-2004 study of political science, economic history, history and economics at the University of Lund, Blekinge Institute of Technology (Sweden) and University College Dublin (Ireland). From 2005 to 2008, fellow of the Graduate School "Markets and Social Systems in Europe" at the University of Bamberg (Germany). PhD completed in November 2010, and PhD thesis published by Routledge (New York) in 2013 ("The Culture of Enterprise in Neoliberalism: The Specters of Entrepreneurship"). His habilitation project dealing with the methodologization of the post-foundational discourse theory will be published in October 2015 by Palgrave MacMillan (Basingstoke) ("Post-Foundational Discourse Analysis: From Political Difference to Empirical Research"). Tomas MARTTILA has been a researcher since 2008 at the Department of Sociology, esp. Sociological Theory at the University of Bamberg. His main research interests include empirical discourse analysis; methods of empirical social research; theories of science; economic sociology; sociology of education. In FQS, Tomas MARTTILA published the article Whither governmentality research: A case study of the governmentalization of the entrepreneur in the French epistemological tradition.

Department of Sociology, esp. Sociological Theory
University of Bamberg
Feldkirchenstr. 21, D-960 52 Bamberg (Germany)

Marttila, Tomas (2015). Post-Foundational Discourse Analysis: A Suggestion for a Research Program [58 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 16(3), Art. 1,
http://nbn-resolving.de/urn:nbn:de:0114-fqs150319.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License

Discursive relations	Phenomenal characteristics
Contrariety	Distinction between a discourse and the totality of other discourses.
Antagonism	The identity of the discursive element A (e.g. social security) is opposed by the discursive element B (e.g. free trade).
Dissociation	Discursive elements A (e.g. social security) and B (e.g. free trade) are mutually contrary because while A is subjectively perceived, social subjects misconceive the very presence of B.
Incommensurability	Discursive elements A (e.g. social security) and B (e.g. free trade) are mutually incommensurable with regard to C (e.g. welfare state), which represents their logic of incommensurability.
Representation	C (the representative) (e.g. social market economy) symbolizes the point of commonality shared by the discursive elements of A (e.g. social security) and B (e.g. free trade) (the represented).
Difference	Discursive elements A (e.g. social security) and B (e.g. free trade) are mutually distinctive and combinable elements, whose logic of commonality is represented by C (e.g. social market economy).

Phenomenal categories	Phenomenal characteristics
Values
Ethical ideals	Paramount values and ideals embodied by nodal points.
Antagonistic others	Opponents and threats that endanger the attainment or sustainment of the ethical ideals.
Subjectivity
Protagonist	Champions and protectors of the paramount ethical ideals.
Opponents	Subjects that endanger or obstruct the attainment of the ideal social order.
Helpers	Subjects designated a supportive role vis-à-vis the attainment and maintenance of the ethical values.
Destinators	The assumed originator or instigator of a course of action.
Receivers	The individual or collective subject on behalf of whom social subjects conduct actions and interactions.
Activity
Actions	Activities associated with subject roles and institutions.
Interactions	Interactions and processes that interlink the activities undertaken by social subjects and institutions.
Objects	Objects social subjects and institutions act upon and which they manipulate during their actions.
Resources	Resources that support social subjects and institutions to accomplish their actions.
Strategies	Strategies and means that social subjects and institutions utilize for the sake of achieving certain effects and outcomes.

Analytical stages	Operationalization
1. Co-construction
Thematic conceptualization	Choice and specification of the research theme and object.
Contextual conceptualization	Definition of the spatio-temporal coordinates of the research object.
Epistemological break	Deletion of the epistemic influence of common sense conceptions, more or less explicit normative standpoints and our theoretical perspective contradicting research traditions.
Deconstruction	Deconstruction of socially conceived implicitness or necessity related to the constitution of the research object.
Theoretical conceptualization	Operationalization of the theoretical framework into a heuristic framework that informs us about the general conditions of constitution of the research object.
2. Reconstruction
Synchronic analysis	Step 1: Compilation of the corpus of data: Identification of the set of practices of articulation that our background knowledge indicates to have been responsible for the constitution and sustainment of the social meaningfulness of our research object in the surveyed spatio-temporal context.
	Step 2: Contents-related analysis of discourse: Reconstruction of the discourse produced in the selected sample of practices of articulation.
	Step 3: Reconstruction of the discursive materiality: Identification of the subject roles and institutions that regulate social subjects' enunciative possibilities and constitute the material prerequisites of the reconstructed structure of discourse
Diachronic analysis	Step 1: Compilation of the corpus of data: Identification of practices of articulation that either enforced the conceived social change or constituted and sustained the discourse that was affected by the social change.
	Step 2: Contents-related analysis of discourse: Contents-related analysis of the contested discourse, which was problematized and changed by counter-hegemonic acts, and the new hegemonic discourse, which was installed by hegemonic acts.
	Step 3: Reconstruction of the material context and consequences of discursive change: Identification of the subject roles and institutions contested by counter-hegemonic acts and installed by hegemonic acts. Identification of the subject roles and institutional affiliations of the hegemonic agents that can explain the credibility and subsequent social acceptance of their counter-hegemonic and hegemonic acts.
3. Collocation	Identification of the phenomenal aspects of the research findings for the sake of enhancing their subsequent reception and paving the way for their comparison with the research conducted on similar phenomena