Specters of Philosophical Anthropology: Critical- Realism, Non-Humanism and Social Historical Analysis.

 Paul Sutton


This paper endeavours to address the reductive analytical closure engendered by the specters of philosophical anthropology which haunt social-historical discourse through making a non-humanist, critical-realist reading of the Archaeology of Knowledge. In this reading both the discursive and the non-discursive dimensions of Foucault's theory of discursive formation are elaborated. Knowledge of social reality is therefore understood as a series of discourses generated by historically specific discursive formations each with their own formative rules. Social formations are, nevertheless, conceived as real social entities which can only be known as objects of discourse.

Specters of Philosophical Anthropology: Critical Social Realism and Social-Historical Analysis.

              "... a ghost never dies, it remains always to come and to come-back" (Derrida 1994:99).[I]


Specters haunt social historical [II] discourse - specters of philosophical anthropology. The ontological framework of many social-historical discourses remains that of an implicit humanism which positions human beings "as the axis and the source of all social relations" and which essentialises those relations as human "expressive or purposive activity" (Jones 1982:52).

Despite Althusserian and Foucaultian exorcisms this ghost did not die. Indeed, it remained and has, disappointing-ly, made something of a come-back. This paper is an endeavour to "reckon" (Derrida 1994:xx) with the reductive analytical closure I

As Derrida observes, a ghost can never be finally exorcised but always remains. Thus, social-historical analysis must continually attempt to engage and reckon with philosophical anthropology even if it is not currently fashionable to do so.

It is my intention, therefore, re-animate the reckoning with the ghosts of philosophical anthropology through a Critical Realist (henceforward CR) re-reading of an early much overlooked text of Michel Foucault - The Archaeology of Knowledge [III] I have re-read this text in an explicitly non-humanist manner that synthetically combines post-structuralist and Marxian theory. Such a reading enables the analytical closures induced by the specters of philosophical anthropology to be re-addressed.

In this paper I will begin by outlining the meta-theoretical framework of CR and its epistemological consequences. Secondly, I will provide a description of the substantive theoretical framework of CR that is derived from the Archaeology, including a detailed explanation of Foucault's poststructuralist theory of discursive formation. Finally, I will briefly address the relationship between the Archaeology and some of Foucault's genealogical texts describing how Foucault's archaeological/genealogical method is consistent with the realist method delineated by Marx which underpins CR.

My aim in this paper is to indicate how an explicitly theoretical social-historical discourse goes some way to address the analytical closure of humanist modes of thought which do not adequately address the methodological effects of the ontological assumptions upon which they are premised. I argue that it is only when assumptions as to the nature of social reality are made explicit that the possibilities of what can be established about that reality, the most fruitful methods for its investigation, and the type of knowledge produced can be ascertained.


A CR ontology assumes that social reality is constituted by the operation of relatively enduring or intransitive, structures which can only be elucidated through the use of corrigible, or transitive concepts [IV] Concepts are deployed in order to obtain knowledge of the intransitive things (real objects) which constitute social reality.  Such objects are not "human constructs" but are relatively enduring "real structures" (Bhaskar 1975:25) which exist and function independently of human knowledge, experience and the conditions which make them accessible.

In CR meta-theory then, things/objects are visualised as complex social entities, possessing an ensemble of powers and tendencies, which generate the phenomena and events in social reality. Generational relations are, therefore, conceptualized as tendencies rooted in the interaction of "real"  social structures and mechanisms. Such structures and mechanisms are "distinct from the patterns of events that they generate": similarly, events are "distinct from the experiences in which they are apprehended" (Bhaskar 1975:56). 

Social reality can, therefore, be seen as stratified into three overlapping domains -  the real, composed of structures and mechanisms, the social entities which generate the actual, constituted by events; and the empirical, in which events are discursively mediated in experience (Outhwaite 1987:22, Marsden 1993:94).

These domains of social reality are distinct, but they often become conflated in social-historical analysis. It is this conflation which reduces the complex being of social reality and serves to effect analytical closure.

This conflation occurs because social structures do not exist independently of their effects (Bhaskar  1975:246). Hence: 

Society, as an object of enquiry, is necessarily `theoretical' in the sense that, like a magnetic field, it is necessarily unperceivable; so that it cannot be empirically identified independently of its effects. It can only be known, not shown to exist (Bhaskar 1989:82).


The tripartite division of social reality into the distinct domains of the real, the actual and the empirical is crucial to a non-humanist CR mode of analysis. CR obviates this conflation by deploying the metaphysical assumption that the realities it seeks to investigate are ontologically independent, yet nonetheless knowable. 

In modes of analysis haunted by the specters of philosophical anthropology these domains are conflated and accompanied by a reversal of the movement from philosophical to empirical analysis which leads to empiricism (see below). Moreover, in such modes of analysis there is the assumption that the mechanisms, events and experiences of social reality are unilinearly and naturalistically linked to one another through the expressive or purposive activity of the human subject. The operation of these mechanisms, the effects of which are social events, can only be known through human experience. Experience is the foundation upon which knowledge of social reality is constructed. Furthermore, thought dominated by an analytic of philosophical anthropology presents an effect of the operation of generative mechanisms, human agency, with a cause. The ontologically humanist and epistemologically empiricist assumptions of such thought therefore limit what aspects of, and the way in which, social reality can be apprehended.

Central to CR is the exorcism of such assumptions. The aim of CR is not the production of knowledge of social reality with an axis and source in the human subject. Rather it is to theorize a sui generis social reality in order that corrigible knowledges can be produced so that particular theoretically defined problems can be elucidated. The analytical distinction between the knowledge generated by CR and social reality is however rigorously respected. For as Woodiwiss (1990: 7-8 emphasis added) argues, there is a necessary distinction:

 between the `analytic' or conceptual and the `concrete' or the empirical ... between the mutually irreducible social scientific things composed of signifiers and their signifieds, the latter including empirical data as well as concepts, and the extra social scientific things to which they are made to refer.

Social-historical discourse does not therefore spontaneously refer to its objects but is rather "made to refer" through the complex interaction of discursive and non-discursive processes.

The analytical distinction between social-historical and extra social-historical objects is precluded by any commitment to the self-evidence of the empirical world. Such a commitment conflates and reduces the independent ontological domain of real social entities and mechanisms to the empirical domain; that is, to the human experience of social-historical events. This conflation/reduction is a consequence of deploying a representationalist concept of language and the maintenance of a distinctly parsimonious attitude to theoretical abstraction. In other words, it represents a failure to see that all language-borne phenomena, including theory is semiological and hence "non-representational" (Woodiwiss 1990). 

Within CR discourse, then, language, is conceived as non-representational. That is, words and concepts etc. do not simply represent non-linguistic phenomena in a natural, transparent or unmediated way: the connections between signifiers and signifieds are "arbitrary" or "unmotivated" (Saussure (1960: 67, 69)). Words are ultimately made to refer or are accepted as referring to concepts and things by the society to which the language belongs. Because the relationship between language and reality is arbitrary and social it is able to change over time. It is the property of a "community of speakers" through whose relations it is aligned with the extra-linguistic world. The alignment and re-alignments of language in relation to extra-linguistic reality are thus the result of "social forces" (Saussure 1960:78). 

Similarly social-historical discourse is aligned with reality by the forces operating in the societal, institutional and situational formations of which it is a dimension. Concepts within this discourse do not, therefore, represent the world but are the property of a community of "theorists". The relationship between concepts and social-historical reality ought not to be taken for granted, therefore, but rather ought to be one of "conscious and controlled reference" (Woodiwiss 1990:9).

Controlled reference can be achieved through the explicit deployment of a CR ontology, which assumes that social reality does not depend for its existence on human consciousness of it, and a CR epistemology that privileges theoretical construction of the object of knowledge over empirical observation. Observation is used "to control abstraction rather than initiate, verify or falsify it" (Woodiwiss 1990:9).

This recognition of the constitutive role of theory and the prioritizing of theory over empirical observation in social-historical discourse does not mean that  epistemology is devalued and the endeavour to make social reality intelligible is relinquished. It is instead to acknowledge, firstly, that epistemologies emerge from particular discursive formations which have particular conditions of existence which limit what is or is not considered valid knowledge (see below). And secondly,  that observation of the empirical dimension of reality is a consequence of a heterogeneous network of discursive, socio-cultural and technological relations. There can be no observation, or indeed observer, anterior to this network. Empirical evidence, therefore, supports the abstractions of theoretical practice but does not initiate them (Smith 1982, Banks 1989). Privileging theory is an acknowledgement not only of the limits of what can be known about social reality, but also of theory's constitutive role in the production of such knowledge. 

In sum, CR is an ontological position which has definite epistemological consequences. It is a position which attempts to overcome the reductive ontology of humanism.

Human subjects, as individuals or as corporate groups, are considered by CR to be neither the axis or origin of social reality nor necessarily the most important source of its transformation. Hence, sociality is not conceived of as simply the result of individual or collective action, but rather as the result of the operation of sui generis social entities and generative mechanisms. CR discourse is not, therefore organised in terms of a pre-critical humanist essentialism or under-theorized analytic categories such as "experience" [V]

Central to thought animated by the analytic of philosophical anthropology is the deployment of the under-theorized  category of experience, common to universal human subjects. Thus, the reality of human experience is represented as immediate and the necessity of the category of experience as self evident. The result is conceptual ambiguity and analytical closure.

In CR experience is conceived of as problematic, contingent, and dispersed in a whole range of social practices which form the conditions of existence of experience. Experience is formed in "the correlation between fields of knowledge, types of normativity, and forms of subjectivity in a particular culture" (Foucault 1992:4). This exorcism of the humanist category of experience does not result in the annihilation of the subject (Woodiwiss 1991). Rather, it means that social entities are objects of analysis in their own right independent of human subjects. This said, the recognition of the independence of these entities does not mean that their operations are external to human subjects.

As Benton (1988: 213-4) argues there are three methodological principles of explanation which addresses the potential problem of externality. Firstly, as psychoanalysis emphasizes, experience, action and subjectivity have an internal unconscious determination. Secondly, the structural conditions of existence of experience, action and subjectivity are not simply constraints but are the conditions of their possibility [VI].  Thirdly, relatively enduring structural relationships between social agents - individual human subjects or collectivities thereof - and between social agents and objects, can bestow powers on such agents. Social agents are conduits of power, power which both enables and constrains social action. Thus, the humanist category of experience is not conceived of as emerging spontaneously but rather as a correlate of the socio-cultural conditions which make experience possible.

In short, human experience is no longer considered to be the ontological foundation of social reality (Hirst & Woolley 1982): experience is no longer a self evident explanatory concept, but part and parcel of the problems which analysis seeks to render intelligible. In CR the analysis of discursive and non-discursive processes displaces the analysis of experience.

Visualising human experience as the result of positioning and interpellation by discursive and non-discursive processes is not, however, a reductionist move in which the subject becomes merely the bearer of social structures. Rather, it is an attempt to liberate the analysis of social formations from the limiting ontology of humanism. For it is through regarding the social subject in this way that a different light can be shed on social formations as structured by social relations which are conceived of as non-human "entities and relations which may only be `seen' thanks to the beneficence of theory..." Woodiwiss (1990:5).

In sum, knowledge of social reality is appropriated through an explicit theoretical practice which displaces under-theorized notions of the subject and experience in order that the uniquely social dimensions of reality may be understood. As social reality is independent of any knowledge of it, however, it is important to respect the boundary between the transitive concepts, models etc., and the intransitive real social entities and relations to which they are made to refer or are accepted as referring.

A framework for an explicitly theoretical form of practice is contained in the Archaeology of Knowledge.

This text can be fruitfully read as a development of Foucault's attempt in The Order of Things to awaken social-historical discourse from the "anthropological sleep" (Foucault 1970:340).  


The Archaeology can be read as a sustained attempt to exorcise the specters of philosophical anthropology. This text deliberately transgresses the categories deployed in humanist modes of social-historical analysis through the construction of a distinction between "total history" - traditional social-historical discourses in both their idealist and empiricist forms, and "general history" - a non-humanist, open textured analysis of the social-historical.

Foucault (1991:9) defines the project of total history as the reconstitution of the material or mental principle that is embodied in particular societies and the law of their cohesion. This mode of analysis involves three untenable assumptions. Firstly, all social-historical events and phenomena found within a particular spatio-temporal domain are linked by a homogenous system of causal relations which have a central core; secondly, the same type of historicity and historical transformations work upon all the dimensions of sociality - economic, political and ideological; and thirdly, history can be divided into distinct periods with their own unitary laws. Thus, total history "draws all phenomena around a single centre - a principle, a meaning, a spirit, a world view, an overall shape;..." (Foucault 1991:9-10) [VII] The single centre to which all phenomena are drawn in social-historical discourses dominated by an analytic of philosophical anthropology is of course the universal human subject.      

All forms of totalizing history, with their stress upon continuity and their correlative focus upon the sovereignty of the human subject and teleological development, are rejected as they induce analytic closure. Indeed, Foucault endeavoured to open up space for analysis through deploying what he called "general history".

General history is a way of producing historically contingent knowledge characterised by a space of "dispersion" in which social-historical events and phenomena are separated and systematised, divided into series, related to one another not by virtue of some essential principle but through theoretical practice which produces heterogeneous and complex social-historical configurations.

The use of the term "general" is perhaps unfortunate for it may lead to the misconception that one form of generality - total history - has merely been replaced with another which merely substitutes continuity with dispersion. Indeed, the term reinforces the tendency to homogenize and essentialize history and historical knowledge (Wickham 1983, 1990: 44-5). I intend to show, however, that despite the shortcomings of the term "general history" Foucault's text is an effort to rethink the process of social-historical enquiry, liberating it from the ahistorical closure inherent in the discourse of total history.    

The Archaeology, then, can be fruitfully read as a non-humanist, non-representationalist, critical theory of social-historical discourse analysis. Its focus is upon discourse as institutionalised social practice, the ways in which social institutions form discursive practice, and the constitutive effects of that practice.

Social-historical reality, therefore, is analysed through the examination of institutionally formed language use rather than the subjects enunciating such language. Thus, the analytical focus of the Archaeology is the social usage and normative power of discourse, not the presence of discourse in the experience or thought of individual or collective subjects. The subject, therefore, is regarded as an "inter-discourse" the site of an ensemble of subject positions given through discourse and by which the subject is interpellated. The fragmented and often contradictory ensemble of subject positions, which result from the open-textured nature of the discursive formations which create their conditions of existence also makes possible the interpellation of the subject through counter-discourses, which make social conflict possible.

The Archaeology seeks to elucidate the rules of formation of discourses by elucidating the discursive and non-discursive social conditions that make their emergence possible. The particular configuration of discursive and non-discursive conditions of existence of discourse Foucault called discursive formations. 

Discursive Formations and their Rules of Formation.

Foucault deploys the theory of discursive formation to analyse the relationship between knowledge and the non-discursive institutions and social practices which are necessary conditions of their existence. He conceives discursive formations as having four domains, that of objects, enunciative modes, concepts and themes, each of which have their own rules of formation.

 1. The Domain of "Objects".

In order to understand what makes it possible for an object to emerge as an important and become known within social-historical discourse the rules that make the existence of such an object possible and which distinguish it from other objects of concern must be described. These rules consist in the elucidation of the:

(i) "surfaces of emergence" (Foucault 1991:41), the "social and cultural spaces in which a particular discursive formation appears" (Sheridan 1990:97). Such spaces are defined by economic political and ideological processes.

(ii) "authorities of delimitation" (Foucault 1991:41-2), the social positions and identities occupied by those who have the right to delimit/pronounce upon objects of discourse.

(iii) "grids of specification" (Foucault 1991:42), the beliefs and conceptual models used to differentiate objects from each other.

Although objects which emerge from discursive formations are identified through theoretical practice they are not simply part of an "architecture of concepts" (Foucault 1991:35). They are events whose conditions of existence lie in the imbrication of the discursive with the non-discursive and which have real effects in the non-discursive. Furthermore the determinants of the rules of formation of discursive formation are broadly speaking social, that is, non-discursive. As Woodiwiss (1990:63-4) argues in his development of Foucault's text:

 ...whilst what discourses might signify is determined by the pertinent sets of linguistic and social relations, what they refer to is determined by those additional discursive and non-discursive structures which determine the institutional settings, the corporate groups and the intellectual templates whereby their objects are defined [VIII]

2. The Domain of "Enunciative Modalities".

The discourses which are made possible by discursive formations are not simply constituted as conceptual architectures, but also as practices. Integral to these practices are the "enunciative modalities" which exist at the interface between the discursive and the non-discursive. That is, discourses are constituted through their modes of enunciation. Different modes or styles of making statements give shape to discursively produced knowledge. The rules governing the formation of these modes are those of:

(i) "who is speaking?"; that is the social position and status of the speaker sanctioned by law, tradition etc.;

(ii) "the institutional site"; that is the institution from which a discourse is legitimated and applied;

(iii) "(t)he positions of the subject"; the possible positions a person can take up within discourse such as questioner, observer or teacher (Foucault 1991: 50-3).

It is important to emphasise that enunciative modes are not regulated by the transcendental human subject of philosophical anthropology but by the anonymous functioning of the ensemble of their rules of formation. This anonymous function is broadly speaking social in nature.

 3. The Domain of "The Formation Of Concepts".

The analysis of concept formation involves the description the "organization of the field of statements where they appeared and circulated" (Foucault 1991:56), that is a description of the interaction of the rules governing:

(i) the "succession" or sequence in which particular concepts appear within a discourse;

(ii) the "coexistence" of different concepts present in a discourse and belonging concomitantly to other discourses, and how these concepts become obsolete.

(iii) " procedures of intervention"; procedures governing how concepts are rewritten, translated and transcribed from one register to another such as from an informal into a formalized language or vice-versa.

As Foucault (1991:63) suggests these rules of conceptual formation:

 operate not only in the mind or consciousness of individuals, but in discourse itself: they operate therefore, according to a sort of uniform anonymity on all individuals who undertake to speak in this discursive field. 

 The anonymous operation of these rules has therefore both discursive and non-discursive dimensions.

 4. The Domain of "The Formation of Strategies".

Discursive formations make possible specific organizations of concepts and objects, specific types of enunciation which form discursive strategies or themes. These strategies are dispersed within discourse according to the rules governing their:

(i) "points of diffraction of discourse" (Foucault 1991:65). A single theme may be taken up in more than one type of discourse. For example, a series of economic concepts which concern value may become inserted into a discourse which deploys a strategy based on exchange or on a discourse which deploys a strategy based on remuneration (Dreyfus and Rabinow 1982:71). These different strategies are related to:

(ii) "the economy of the discursive constellation" (Foucault 1991:66). Themes in a discourse have determinate relationships to themes in other discourses to which they are related. 

The formation of discursive strategies enables social transformation to take place as a discursive formation does not fill all the possible space opened up by the configuration formed by its objects, enunciations, and concepts: a discursive formation is therefore "essentially incomplete owing to the system of formation of its strategic choices" (Foucault 1991:67). The incomplete condition of a discursive formation means it can be "taken up again, placed and interpreted in a new constellation" which reveals "new possibilities" (Foucault 1991:67).

Only certain objects, concepts etc., emerge from discursive formations and become inserted in particular discourses which are themselves open, loose ensembles of statements. The constellations of statements which form themes and structure discourse when inserted into other discourses and articulated with other discursive and non-discursive structures function in different ways depending on:

(iii) "the function the discourse under study must carry out in a field of non-discursive practices"(Foucault 1991:68). Discourse has a constitutive function in the economic, political and ideological dimensions of sociality. This general function also includes particular rules concerning the "process of appropriation of discourse", which people have the right to authorize the institutionalization of strategies; the "possible positions of desire in relation to discourse" (Foucault 1991:68), which is not marginal to discourse, but rather is a "formative element" of the way in which a discourse functions in non-discursive practices.

In sum, the aggregated rules of formation regulate one another and determine what a discourse signifies. Nonetheless, "the determinants of these rules are social in the general sense" (Woodiwiss 1990:63). That is, in CR the rules are seen as formed through their relationship with the structures of non-discursive social reality which determine what a particular discourse refers to.

Unfortunately, due to Foucault's somewhat oblique, prolix, and tortuous style and his habit of introducing a bewildering array of concepts which are never fully elaborated, the analysis of the relationship between discursive and non-discursive practice, which is central to the Archaeology, has often been misunderstood. Foucault's exhaustive catalogue of what discourse is not and his continual emphasis upon the specificity of discourse, often obscures the crucial role of the positive relation of discourse to the non-discursive. CR maintains that the imbrication of the discursive and non-discursive is an integral part of the rules of formation of discursive formation. For example, the formation of objects of discourse requires a non-discursive surface of emergence, that is, a social and/or an institutional site. Similarly an enunciative mode requires not only subjects to speak in particular styles but institutional sites from which speech can be spoken.

The relationship between the discursive and non-discursive in the Archaeology has been criticised by Dreyfus and Rabinow (1982). Whilst acknowledging that the text possesses "a concrete and social notion of the non-discursive background" (Dreyfus & Rabinow 1982:77), they argue that this background has a nominal status, it is a passive domain determined by discourse. They interpret Foucault's attempts to specify the effectivity of discursive practice as merely a version of structuralist theory in which there is "the strange notion of regularities which regulate themselves" and "a strange alliance between rules as descriptive regularities and as prescriptive operative forces" (Dreyfus and Rabinow 1982: 84, 81).

Dreyfus and Rabinow, however, fail to fully acknowledge that the rules of formation of discursive formation have, as indicated above, integral non-discursive as well as discursive conditions of existence. Regularities in discourse do not therefore regulate themselves but are rather the product of the interaction both discursive and non-discursive relations. Foucault, it is true, did not establish a "hierarchy" of effectivity in the relations between objects, subject positions concepts and strategies, or between the discursive and the non-discursive. This, however, can be seen as a deliberate theoretical move made in order to open up the space of analysis rather than close it down through a preordained causal hierarchy. The precise articulation of the relations within a discursive formation and of the discursive to the non-discursive can only be established through substantive analysis, as exemplified in the genealogies Discipline and Punish and The History of Sexuality Volume 1 (see below).   

From a CR perspective Dreyfus and Rabinow's second failure consists of the underestimation of the constitutive effect of theory and the privileged role it has within Foucaultian analysis (Foucault 1991:26). Through theoretical practice Foucault attempts to de-familiarize knowledge production so that the specters of philosophical anthropology can be exorcized from this process. The explicitly theoretical construction of the rules of formation of objects, enunciative modes, concepts and strategies shows that these rules actively constitute the discourses through which the world is known. Knowledge, therefore, is the result of the relatively autonomous discursive dimension interacting with and through the non-discursive dimensions of social formations. The rules of formation do not, therefore, prescribe a reality which is independent of discourse, but describe the way in which that reality becomes known through discursive practice.

Tranquilized by the anthropological sleep Dreyfus and Rabinow are unable to "see" what Foucault tried to accomplish. Their humanist theoretical framework impels them to interpret the strange exorcism of the ghost of philosophical anthropology from the Archaeology as a failure, whereas this is its success. The theoretical framework of the Archaeology would not, as Dreyfus and Rabinow argue, be abandoned in future texts, but would be developed through Foucault's use of the idea of genealogy. Indeed, CR maintains that the critical theoretical programme of the Archaeology continued to inform the less abstract register of the genealogical histories and that archaeology/genealogy is consistent with the realist methodology delineated by Marx.


In contrast to such diverse commentators as Dreyfus and Rabinow (1982:102), and Smart (1983:74), I do not see Foucault's deployment of the concept of genealogy as a reversal of the priority of theory over practice. This dichotomization of theory and practice seems to me to be a misunderstanding of the idea of theory as a form of discursive practice. As Lemert & Gillan (1982:120) observe, "(f)or Foucault, theory is practice, and practice is theoretical because knowledge is in power and power in knowledge". Theory, then, is a form of political practice. Furthermore, I do not endorse Dews' (1979:150) and Dean's (1994:14) assertions that there is a break between the Archaeology and later texts. The critical theoretical programme of the Archaeology continues to inform the less abstract register of the genealogical histories.

In CR genealogy is understood as an attempt to conceptualize the singularity of social-historical events whilst recognizing that their facticity is constituted, in part, by discourse and that any knowledge of events is necessarily contingent [IX]. As Foucault (1977:157) argues, genealogy gives "equal weight to its own sight and to its objects". This is an acknowledgement of the constitutive effects of a theoretically informed gaze. It is neither a claim for objectivity nor relativity. Rather, archaeology/genealogy seeks to transgress such dichotomies and epistemological categories, seeks to make possible multiple regimes of plausible but corrigible knowledges of social reality.

Whilst denying there is any form of meta-theoretical break between Foucaultian archaeology and genealogy I do, however endorse Smart's (1983:79) assertion that there is a post-Archaeology shift in which the focus upon relations of power, knowledge and the body becomes more explicit. This does not, however, represent a "decentering of discourse" (Fairclough 1992:49) in favour of the analysis of power. For "it is in discourse that power and knowledge are joined together" (Foucault 1990:100). The analysis of discourse as specified in the Archaeology, therefore, remains of crucial importance to genealogical analysis. Whereas archaeology elucidates discursive practice, genealogy elucidates the imbrication of those practices with non-human power/knowledge relations (Foucault 1987:27-8). Thus, genealogy does not displace archaeology but rather constitutes a broadening of its analytical horizon (Davidson 1986:227).

Part of the broadening of the Archaeology's analytical horizon is through a refusal to prejudge the causal effectivity of either discursive or non-discursive practices. Indeed,  Foucault (1981:6) attempted to "lighten the weight of causality" [X] by an increasing analytic polymorphism in which the complex imbrication of discursive and non-discursive elements. Thus, in place of the "uniform, simple notion of assigning a causality" Foucault (1978:13) substitutes a "polymorphous cluster of correlations". Correlations between intra, inter and non discursive elements, relations and domains of reference, results in a "plethora of intelligibilities" (Foucault 1981:7).

This Foucaultian theory/methodology is consistent with the realist method delineated by Marx in the introduction to the Grundrisse and developed in Capital (Callinicos 1982). Marx (1973:101) privileged the analytical movement from "the abstract to the concrete" as this is the way in which "thought appropriates the concrete, reproduces it as the concrete in the mind". The concrete therefore is the result not the point of departure of knowledge production. That is not to say that thought creates the concrete for it "retains its autonomous existence outside the head" (Marx 1973:101). This Marxian conceptualization of realism provides the methodological framework for CR.

By contrast methodologies premised upon the humanist ontology of philosophical anthropology, which are either epistemologically empiricist or idealist and which deploy a representationalist concept of language, act to conflate the "concrete" and the "concrete in the mind", to conflate the real, the actual and the empirical as discursively mediated in experience.

This results in the attribution of the "particular epistemological concept" of experience with a "general ontological function" and the tendency to ignore the social conditions in which experience is "epistemically significant" (Bhaskar 1975:28). The conflation of a condition of knowledge (ie. experience), with the actual being of the real results in the analytical closure of the "epistemic fallacy" (Bhaskar 1975:36): statements about social being are reduced to experientially based statements concerning the knowledge of that being (Bhaskar 1975: 36, 187, 198). [XI]

In sum, I have argued that the non-humanist theoretical/methodological framework of CR, derived from a synthesis of Foucaultian and Marxian texts enables the eradication of such conflation and closure from social-historical discourse.


This paper has attempted to reckon with the specters of philosophical anthropology through the development of a CR ontology, which assumes that the existence of social reality does not depend on human consciousness of it, and a CR epistemology that privileges theoretical construction of the objects of social-historical discourse over empirical observation.

I have argued that Foucaultian theory of discursive formation of the Archaeology can be read as a non-humanist, non-representationalist form of CR which attempts to elucidate both the discursive and the non-discursive dimensions of social reality; and that the methodology of archaeology/geneaology is consistent with the realist methodology advocated by Marx.

The invocation of the work of Marx in this paper is deliberate. For as Derrida (1994:13) argues the reading and rereading of Marx's work is "more urgent today" than it has ever been. Indeed, CR endorses Derrida's (1994:13) position that it is "a fault not to read and reread and discuss Marx ... a failing of theoretical, philosophical, and political responsibility".

It has been my contention that the "beneficence" (Woodiwiss 1990:5) of a CR theoretical framework makes it possible to see, situate and identify (Derrida 1994: 132) the specters of philosophical anthropology, to obviate the conflation and closure effected by these ghosts and to re-engage in the perennial struggle to exorcise them from social-historical discourse.


[I] Although Derrida's text acted as a stimulus to the writing of this paper I do not mean to imply that this text can be assimilated within a CSR framework. This paper, like the work of Marx, is grounded in a "pre-deconstructive-ontology" (Derrida 1994: 170), that is, a realism, which acknowledges that the assumption of the real as present is a metaphysical but, nonetheless, fruitful meta-theoretical strategy. For Derrida, although not false, a realist strategy is not radical enough as it remains beguiled by the "metaphysics of presence".

[II] This paper aims to "elucidate the question of society and that of history, questions that can be understood only when they are taken as one and the same: the question of the social historical" (Castoriadis:1987: 167).

[III] As Dean (1994: 175, 195) observes even Foucault's later works on governmentality seek to avoid "relapsing into a humanist framework" and continue his "rejection of the founding or constitutive subject of philosophical humanism".

[IV] I understand ontology as defined by Arthur (1986:153) as "that set of fundamental categories through which the character of the social sphere is delimited and the general framework for theory construction established." As the author argues a priori arguments do not "establish the necessity of these categories", however, all research programmes presuppose "a commitment to some ontology."

[V] On the discursive character of experience and the politics of its construction see Scott (1991).

[VI] As Dean (1994:56) argues critical social-historical discourse is "concerned with the variation of forms of specification of human beings and their capacities under definite social, ethical, political and cultural conditions" and with the examination of: the way in which such forms of specification have particular discursive conditions presupposed by and necessary to various forms of social, legal, political and economic practices and techniques.

[VII] Compare Vattimo's (1992) Ch.1 passim compelling postmodern critique of unilinear, teleological forms of historical discourse.

[VIII] The theoretical and methodological position developed in this paper owes what Derrida (1994:92) would call an "unefaceable and insoluble debt" to the text inscribed under the proper name of Anthony Woodiwiss - Social Theory After Postmodernism. The procedure recommended for the production of discursive analyses of social-historical knowledge in Foucault's The Archaeology of Knowledge animates Social Theory After Postmodernism. The latter text, however, contains an explicitly sociological appropriation of the Archaeology which this paper endeavours to build upon.

[IX] Interestingly, Althusser thought his former student Foucault, an exceptional thinker who destroyed the obviousness of facticity (1979: 44,45,). He thought Foucault's texts were immense efforts "in abstraction" which attempted to identify and construct the object of social-historical enquiry. This effort Althusser (1979:103) applauded as it was:

... antipodal to the empirically visible history in which the time of all histories is the simple time of continuity...Instead of these categories, continuity and discontinuity, which summarise the banal mystery of all history, we are dealing with infinitely more complex categories in which new logics come into play...


[X] Compare Vattimo's (1992:71) commitment to a "lightening of being". Vattimo's (1992:118) "ontology of reduction...carries to its natural conclusion the Heideggarian idea of being that can never be given as full presence...".  Even the CSR concept of relatively enduring structures would be considered by Vattimo as a form of nostalgia. See note 1 above

[XI] See Outhwaite (1987:32), who observes that both idealism and empiricism "reduce ontology to epistemology" whilst "retain(ing) an implicit ontology of the empirical world".    



 Althusser, L. & Balibar, E. (1979) Reading Capital, London: Verso.

 Arthur, C. (1986) Dialectics of Labour. Marx and his Relation to Hegel, Oxford: Blackwell.

 Banks, J.A. (1989) "From Universal history to historical sociology", British Journal of Sociology, 40, 521-543.

 Bennett, T. (1979) Formalism and Marxism, London: Metheun.

 Benton, T. (1988) The Rise and Fall of Structural Marxism. Althusser and his influence, London: Macmillan.

 Bhaskar (1975) A Realist Theory of Science, Leeds: Leeds Books.

               (1989) Reclaiming Reality: a critical introduction to contemporary philosophy, London: Verso.

Callinicos, A. (1982) Is There A Future For Marxism?, London: Macmillan Press Ltd.

Castoriadis, C. (1987) The Imaginary Institution of Society. Oxford: Polity.

 Cousins, M. (1987) "The practice of historical investigation", in D. Attridge et al (eds.) Post-stucturalism and the question of history", London: Cambridge University Press.

 Cousins, M. & Hussain A. (1984) Michel Foucault, London: Macmillan.

 Davidson, A. (1986) "Archaeology Genealogy Ethics", in D.C. Hoy (ed.) Foucault: A Critical Reader, London: Blackwell.

 Dean, M. (1994) Critical and Effective Histories Foucault's Methods and Historical Sociology, London: Routledge.

 Derrida, J. (1994) Specters of Marx. The State of the Debt, the Work of Mourning, & the New International, London: Routledge.

 Dews, P. (1979) "The Nouvelle Philosophie and Foucault", Economy and Society, 8, 127-171.

 Dreyfus, H.L. & Rabinow, P. (1982) Michel Foucault: Beyond Structuralism and Hermeneutics, London: Harvester Wheatsheaf.

 Foucault, M. (1970) The Order of Things: An Archaeology of the Human Sciences, London: Tavistock.

                      (1977) Language, Counter-Memory, Practice. Selected Essays and Interviews, New York: Cornell University Press.

                        (1978) "Politics and the study of discourse", Ideology & Consciousness, 3, 7-26.

                         (1980)  Power/Knowledge Selected Interviews and other Writings 1972-1977, C. Gorden (ed) Brighton: Harvester.


                           (1981) "Questions of Method: an interview with Michel Foucault", Ideology & Consciousness, 8, 3-14)          

                           (1987 edition) Discipline and Punish: The Birth of the Prison System, London: Peregrine.

                                                  Sexuality Volume 1, London: Penguin

                           (1991 edition) The Archaeology of Knowledge, London: Routledge.


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

Flynn, T.R. (1991) "Foucault and the Spaces of History", The Monist, 74 (2), 165-186.

Hirst, P, & Woolley, P. (1982) Social Relations and Human Attributes, London: Tavistock.

Jones, K. (1982) Law and Economy. The Legal Regulation of Corporate Capital, London: Academic Press.

Lemert, C. C. & Gillan, G. (1982) Michel Foucault Social Theory and Transgression, New York: Columbia University Press.

Marsden, R. (1993) "The Politics of Organizational Analysis", Organization Studies 14 (1), 93-124.

Marx, K. (1973) Grundrisse. Foundations of the Critique of Political Economy, Trans. Nicolaus, M. London: Allen Lane.

Outhwaite, W. (1987) New Philosophies of Social Science, London MacMillan.

Poster, M. (1982) "Foucault and History", Social Research, 49, 116-142.

Prior, L. (1989) The Social Organisation of Death. Medical Discourse and Social Practices in Belfast, London: MacMillan.

Saussure, F. (1960) Course in General Linguistics, London: Peter Owen.

Scott, J.W. (1991) "The Evidence of Experience", Critical Inquiry, 773-797.

Smart, B. (1983) Foucault, Marxism and Critique, London: Routledge.

Smith, D. (1982) "Social history and sociology - more than just good friends", Sociological Review, 30, 286-308.

Wickham, G. (1983) "Power and power analysis: beyond Foucault?", Economy and Society, 12 (4), 468-498.

                      (1990) "The Currency of History for Sociology", in S. Kendrick et al (eds.) Interpreting the Past, Understanding the Present, London: Macmillan.

Woodiwiss, A. (1990) Social Theory After Postmodernism: Rethinking Production, Law and Class, London: Pluto.