Horkheimer, Habermas, Foucault as Political Epistemologists

Abstract

I

Introduction. In this paper I offer a reorientation of the problematic of political epistemology that puts power at the centre of analysis. I am motivated to do this out of concern that both social and political epistemology tend too often to focus on non-elites, such as anti-vaxxers, rather than elites, where much more serious damage is done to the transmission of knowledge. In order to make this shift I will return to a different problematic about knowledge and power developed in twentieth-century European philosophy, in the works of Horkheimer, Habermas and Foucault. In the conclusion I will briefly discuss how this trajectory of work can shed light on recent attempts to decolonize epistemology.

Clearly, the question of the relationship between politics and epistemology must involve not simply an account of the epistemic vices of non-elites, vices that are easily identifiable by our existing epistemic norms, but a consideration of the connections between those epistemologies that became influential in the mainstream public discourses and broad social injustices, such as predatory capitalism and colonialism. As numerous feminist epistemologists have argued, we need to consider the role that theories of justification and formulations of knowledge have had on the epistemic authority of various groups in society such as manual labourers, women, and colonized peoples, and how this has diminished democracy and secured elite power in our technocratic societies from external critique (Harding 2006, 2008; Code 1991).

I realize that one might take epistemology’s effect on society to be an ‘after-the-fact’ sort of consideration—that is, one might argue that if an epistemology is strongly warranted on philosophical grounds, its political effects are merely extrinsic and not causally determinant. The philosophers I will consider here hold that such a view is simplistic and unwise. My project is to become clear on their arguments, and to show how these provide a distinct approach to political epistemology.

It is important to note that philosophers of the twentieth century were not the first to acknowledge a relationship between knowledge and power, or between epistemology and politics (two different sets of relations). Immanuel Kant argued in favour of transcendental idealism in part on the grounds that it would defeat authoritarianism and provide support for the demands of free speech made by the revolutionaries of his era. Kant believed that an approach to knowledge in which human beings are seen as mere passive recipients, without taking an active role, is an approach that gives support, or cover, for dogmatically held claims (Tiles and Tiles 1993). And he believed that such a dogmatism would commit the human species to a Hobbesian state of nature, in which ‘assertions and claims’ can only be established ‘through war’ (Kant [1787] 1958, a752/b780; see also Shell 1980). In contrast to this version of objectivism, Kant thought that, because our knowledge of objects is based on practices of human reason, transcendental idealism mandates open procedures of disputation for epistemic reasons, as the means to improve belief. Here Kant is offering an analysis of the political effects of epistemologies as a reason to favour a certain approach to justification.

Kant is not the only philosopher who made such arguments. Bacon was concerned with the effect of the market on both the circulation and the justification of knowledge; Locke was motivated to attack the doctrine of innate ideas because of his concern with the growing influence of the Enthusiast movement, which sanctioned women’s speech in the public square; and the logical positivists were forthright in their hopes that a demystification of speculative claims could defeat the growing currents of fascism (Tiles and Tiles 1993; Potter 1993; McCumber 2001; see also Bordo 1987; Mills 1988). Thus the history of epistemology needs to be retold to correct its common representation as above the fray of earthly politics. But the corrected history of epistemology will not be a story that discredits so much as it redeems by establishing that even in the central texts of the European canon one can find a sophisticated political reflexivity from which we can still learn.

However, it remains the case that close theoretical work on the exact nature of the relationship between power and knowledge, and politics and epistemology, did not occur in the West until the twentieth century. Three of the most important thinkers to have worked on this topic are Max Horkheimer, Jurgen Habermas and Michel Foucault. The importance of their work is that each retains, I will argue, a project of epistemic normativity in which epistemic concerns are not replaced by political concerns but illuminated and potentially strengthened by a consideration of the domain of the political.

The political epistemology that emerges from this tradition can be viewed as analogous to the field of political economy. The project of political economy was intended to provide an explanatory account at a meta-level, but one that did not eclipse economic theory so much as trace the interrelationship between economic forces and political formations, to note the political constitution of economic roles and forms of agency, and thus to de-naturalize contemporary processes of production and distribution, as well as create a space for imagining alternative configurations and research questions. The project of political economy is allied with the descriptive sociology of economics. The Nobel Prize-winning economist Paul Krugman, among others, has argued that the 2008 economic crisis requires a sociological analysis to understand how certain methods of analysis and argumentation, despite their shortcomings, escaped critical scrutiny, but his goal is also normative: to learn how economics as a field might be improved. Similarly, the critiques advanced by Horkheimer, Habermas and Foucault have normative aims directed at improving theories of knowledge.

My aim in this paper is to explore the general approach to this topic that Horkheimer, Habermas, and Foucault offer, and to ask in particular of each one, how are epistemic criteria (or criteria aiming at truth) understood in relation to power? Many resist the claim that there is a constitutive relation between power and knowledge, fearful of ending up in the sort of epistemic nihilism associated with Lyotard or Rorty, in which epistemic considerations are rendered null.¹ The three philosophers that I have chosen agree that the autonomous operation of epistemic criteria is a necessity for thought: Horkheimer, in fact, accuses the American pragmatists of forsaking truth in the name of adaptation; Habermas reveals his own orientation to this effect by accusing Foucault of simply replacing knowledge with power; and Foucault himself insisted with some irritation that the idea that he believes knowledge is no more than a ‘thin mask thrown over the structures of domination … is so absurd as to be laughable’ (Foucault 1990a, p. 264). This raises the question of how their normative approaches to knowledge can be squared with the idea of a constitutive relationship between power and knowledge.

Before I begin, I should note that Habermas’s work mainly discusses human interests rather than power, and Horkheimer’s concern is focused on reason rather than knowledge. Still, I will show that their work, like Foucault’s (and frankly, like Kant’s), argues for the epistemic necessity of developing a political reflexivity about epistemologies. However, they cash out these claims in interestingly different and constructive ways.

II

Horkheimer. In his 1947 book, Eclipse of Reason, as well as in several other essays published in the 1930s and ’40s, Horkheimer develops a critique of what he calls ‘traditional empiricism’ in the social sciences. His analysis was based on his academic experiences as a sociologist as well as his work at the intersections of the social sciences and private enterprise both in Germany and in the United States.

Horkheimer’s central criticism of the traditional form of empiricism dominant at the time was that it was operating with a mistaken ontology of truth. What we are describing when we describe the world around us is not a natural creation, but the product of collective human praxis, meaning reflective practical activity.

[H]uman action unconsciously determines not only the subjective side of perception [through, for instance, production of certain kinds of perceptual and measuring tools] but in larger degree the object as well. The sensible world which a member of industrial society sees about him every day bears the marks of deliberate work: tenement houses, factories, cotton, cattle for slaughter, men, and in addition, not only objects such as subway trains, delivery trucks, autos, and airplanes, but the movements in the course of which they are perceived. The distinction within this complex totality between what belongs to unconscious nature and what to the action of man in society cannot be drawn in concrete detail. (Horkheimer 1975, p. 201)

Thus traditional empiricism obscures not only the active role of knowers, as Kant argued, but the powerful role of praxis in producing the phenomena it measures and describes, instead taking social entities to be akin to found objects. Because it also sidelines the role of interpretation in the social sciences, it hinders open debate about the interpretative frames in use, as well as about the social construction of objects like tenement houses and cattle for slaughter. ‘The whole perceptible world as present to a member of bourgeois society and as interpreted within a traditional world-view which is in continuous interaction with that world, is seen by the perceiver as a sum-total of facts; it is there and must be accepted’ (Horkheimer 1975, p. 199). In truth, however, ‘The facts which our senses present to us are socially preformed in two ways: through the historical character of the object perceived and through the historical character of the perceiving organ’ (1975, p. 200). This claim helps to explain why, in the previous quotation, Horkheimer slides the category ‘men’ in with other socially constructed objects. Not only is there a feedback loop between the measurement, description, perception and delimitation of phenomena and the phenomena themselves, but we are also phenomena within this dynamic, as knowers. Not only perceptible social reality, but also the mode or process of perception, interpretation and judgement is the product of a particular historical type of social activity and praxis.

Horkheimer’s overall concern with empiricism, then, is that it tends toward reifying the objects it measures and obscuring the conditions of their production, as well as its own role in this production.²

In their important essay Dialectic of Enlightenment, Horkheimer and Adorno (1947) develop a broad argument about the associations between the rise of modern European sciences and ideas about reason, on the one hand, and the development of capitalism, on the other. The form that empiricism took in this period was what they called ‘instrumental reason’, in which means-end calculations toward the advancement of strategic short-term goals eclipsed theoretical and interpretative debate. Thus reason was instrumentalized by capitalism for its own purposes, and any other use or practice of reason beyond this frame was cast as speculative and unscientific. The search for quantities of data mimics capitalism’s focus on accumulation, and the denial or minimizing of the role of interpretation and judgement in the sciences accords with capitalism’s desire to conceal its harms and its values by recourse to naturalistic claims and absolute laws. Overall, then, in their view, the social and human sciences have been instrumentalized by capitalism with the effect of influencing predominant methods and justificatory theories; the preference for quantitative methods was not arrived at through an internal logic of rationality but by pragmatic considerations and contextual conditions which have received insufficient reflexive critique.

Of course, the idea that theory choice in the human sciences is guided by pragmatic interests is neither new nor subversive of all forms of empiricism, at least since the publication of Quine’s ‘Two Dogmas of Empiricism’ (1951). Positivist characterizations of the sciences or transcendental justifications of theory choice (both of which Horkheimer calls forms of ‘false consciousness’) are no longer hegemonic, which raises the question of how much of his critique would still be relevant for the more contemporary versions of what used to be called radical empiricism. However, although positivist characterizations of empirical evidence have lost ground, what has received less attention is Horkheimer’s insistence that the way in which pragmatic considerations play out in theory choice in the social sciences involves the market, and that an epistemology of the social sciences must analyse their position within a society in which market priorities and market logics have achieved hegemony. For Horkheimer, it is this context that is central to the social production of the ‘knowing individual as such’ (1975, p. 199). I suspect that the strict distinction we still tend to make today between normative epistemology and a purely descriptive sociological approach to knowledge effectively closes off such considerations. Thus it remains instructive to revisit Horkheimer’s arguments that ‘sociological’ concerns are intrinsic to any epistemological assessment of the sciences: in this way, perhaps we can incorporate the necessary reflexivity or capacity for self-knowledge and self-criticism.

More contemporary versions of Horkheimer’s basic argument can be found in the research team of Helga Nowotny, Peter Scott and Michael Gibbons, who have diagnosed the methodological deficiencies in current empirical research in Re-Thinking Science: Knowledge and the Public in an Age of Uncertainty (2001). This work includes normative aims that argue for developing a context-sensitive approach to understanding the particular development of scientific practices in the modern era, as well as the relationship between science and economic systems and forms of subjectivity. Similarly, Horkheimer’s project was not to reject normative epistemology, but to bring to consciousness the link between the social production of knowledges in the sciences and the social production of society, and thus to show that the production of knowledge, no less than ‘subway trains and tenement houses’, reflects the current condition of a given culture’s praxis.

The tradition of German critical theory was guided by normative epistemic considerations toward enhancing the possibility of achieving a more reflexively adequate knowledge based on a more adequate and accurate understanding of the dynamic and historical nature of the object of study, as well as of knowledge’s own potential to exert a constitutive effect on that object, changing or substantially forming it. Because of this dynamic at the ontological level, the normative dimension of theory (such as that involving theory choice) cannot be sequestered as a separate or specialized theoretical domain. An adequate reflexivity will acknowledge the implicating dialectical relations between forms of theory and empirical practices and social relations and social subjects, reveal the specific effects in a given domain of inquiry, and incorporate as a standard practice a normative debate over them. If it is impossible to conduct social inquiry that does not have a feedback effect on its objects, then we should open up debate over the constitution of such objects. For example, if the rather large category of depression has an effect on symptoms and behaviour as well as diagnoses, let us have a clear-eyed debate over the pros and cons of these effects, rather than attempting to represent the debate over the category as simply one about reference. Moreover, theoretical approaches that aim to augment social functioning under current conditions should not be able to hide under the guise of neutrality once we have incorporated this level of reflexivity and normative assessment as a routine feature of the justification procedures.

Horkheimer’s conception of critical theory also includes an account of how to approach normative arguments over the ends of inquiry that an adequate reflexivity will engender; this will prove relevant to the later discussion of Habermas and Foucault. Given the fact that inquiry has a role in shaping society and not merely in describing, measuring or explaining it, a sufficiently reflexive epistemology of the social sciences should take a critical approach to the existing social totality. It should uncover the embedded norms and framing assumptions that cloak the dogmatism about ends. Sometimes Horkheimer characterizes this effort as an ‘antagonism between theory and practice’ ([1947] 1987, p. 83), meaning an antagonism between critical theory and the forms of empirical practices currently in dominance. He argues that we should be ‘honest’ about this antagonism of objectives, given that the critical epistemology he advocates for will not appear justified or justifiable from the perspective of traditional empiricism. Rather, because the approach he advocates is open about the need for a values debate, it will appear biased. Taking any position in regard to social ends other than neutrality looks like a biased position, lacking sufficient empirical support, given its critique of the existing emphasis on data.

I think Horkheimer overstates the case a bit here: surely the feedback loop is subject to empirical substantiation, and the non-reflexive or insufficiently reflexive character of many disciplinary methodologies can also be established, as commentators did in regard to the 2008 economic collapse. But Horkheimer is certainly correct that such reflexivity would require a culture shift in the sciences, from pretensions of neutrality to the exposure and debate over the normative commitments that drive framing assumptions. And he makes the interesting claim that the debates over social ends cannot be conducted as hypothetical arguments but as existential ones, by which he means that adequate empirical verification cannot be found in present reality, but requires the experience of a future, differently formed society.

So what are the epistemic standards of justification operating to determine theory choice on such an account? Horkheimer gives us something along the lines of standpoint theory: a theory of justification must be largely procedural and involve a dynamic interaction between the proletariat and critical social theorists (1975, pp. 213–15). Neither are likely to discern the truth alone. In his view, workers need to be central to the social sciences, but not because they have access to an unvarnished experience of present realities, since, as with standpoint theory, the positional-based experience is insufficient by itself. But multiple standpoints are required to develop a critical consciousness in relationship to experience.

The workers, then, are not mere conveyors of data, used like thermometers, as Fricker (2007) put it, whose interpretative analysis has no authority. Anticipating Foucault here, Horkheimer’s argument is that working-class knowledge can provide ideology critique in exposing the specious nature of some existing categories and concepts. Their input is also central to the crafting of realistic alternatives. Today this position would most likely be altered to include the experience of any marginalized groups as a necessary ingredient for reflective social analysis. From the margins, where those left behind in the process of capital accumulation live, one might discern what has been left out of current models, what values remain unaccommodated in existing society, what truths exist below the radar screen of dominant knowledges. It is not the case that only in the margins one will find the truths, nor that those who live on the margins have more reliable analytical knowledge, but that a full assessment of current society requires a reflective account of what remains in the shadows. Thus Horkheimer’s privileging of the importance of basing social theory on the knowledge available to the working class emerges from the emancipatory normative aims that he proposes for critical theory, but is also motivated by his desire to develop a social theory that has more epistemic adequacy than traditional theory can produce.

This helps to clarify Horkheimer’s relationship to empiricism. Critical theory must also be empirical, in close relation especially to the workers’ experience. Yet that experience itself, like the objects and processes of perception, requires interpretation and is subject to critical analysis. His position fits loosely with the mid-century radical empiricism that developed out of logical positivism, which incorporates a pragmatist approach to justification along with ontological pluralism and the indeterminacy of theory choice. Yet here Horkheimer develops a concept of what he calls ‘objective reason’ which amounts to a commitment to the determinacy of theory choice.

Although he takes a relatively contemporary view on the indistinguishability of fact and value, in characterizing the knowledge that critical theory seeks, Horkheimer takes a position on the relation between politics and knowledge that is startlingly distinct. In Eclipse of Reason, Horkheimer presents the idea of ‘objective reason’ to counter the ‘subjective reason’ that validates only instrumental means-end calculations and results in a diminished rationality. He describes subjective reason as follows:

The neutralization of reason that deprives it of any relation to objective content and of its power of judging the latter, and that degrades it to an executive agency concerned with the how rather than the what, transforms it to an ever-increasing extent into a mere dull apparatus for registering facts. (Horkheimer [1947] 1987, p. 55)

To rescue reason, we must return to an idea of objective reason, but Horkheimer’s is not quite the same argument as Boghossian (2006) or that of others, such as Williams (2002), who want to rescue truth and reason from what they see as a political reduction. Horkheimer’s argument relies on a Socratic idea of reason as ‘universal insight’ ([1947] 1987, p. 10) in which definite beliefs and lines of action follow directly from the reasoning process. He says:

The philosophical systems of objective reason implied the conviction that an all-embracing or fundamental structure of being could be discovered and a conception of human destination derived from it. They understood science, when worthy of this name, as an implementation of such reflection or speculation. (Horkheimer [1947] 1987, p. 12, my emphasis)

Such a view has been eclipsed, he argues, by a reason put to the service of subjective ends that are themselves incapable of objective justification:

In order to prove its right to be conceived, each thought must have an alibi, must present a record of expediency. Even if its direct use is ‘theoretical’, it is ultimately put to test by the practical application of the theory in which it functions. Thought must be gauged by something that is not thought, by its effect on production or its impact on social conduct, as art today is being ultimately gauged in every detail by something that is not art, be it box-office or propaganda value. (Horkheimer [1947] 1987, pp. 50–1)

One might well take such a critique of instrumentalizing reason to be a denial of the constitutive relationship between knowledge and power. Horkheimer’s view seems to imply that those who would put knowledge to the service of social justice are operating in the same way as those who would put it to the service of capitalism. In both cases, thought is guided by non-thought, or extra-epistemic considerations.

The objective reason that Horkheimer champions, however, is not politically neutral, a mere method subject to various political uses but without a politics of its own. He argues that the intrinsic political implications of reason will emerge from following reason beyond narrow pragmatic concerns once it is no longer truncated to short-term interests and means-ends calculations. Market-driven research forecloses the rational deliberation over ends because it well knows that it cannot win that game. Hence it seeks to pre-empt the objective truths that reason would reveal, such as, one might argue today, the truth that market-driven, growth-oriented economies are in conflict with species survival.

Few of us can muster such faith as Horkheimer that the pursuit of an unencumbered reason over ends will yield an objective normativity. Critical reason alone would seem to provide no guarantee of reaching consensus over the best, most justifiable ends. Horkheimer did not think consensus would be reached, either, at least for the foreseeable future, but he diagnosed this failure as due to market influences, a diagnosis that may still have some plausibility; but that alone is not adequate to explain the insufficiency of reason to direct a normative theory. Horkheimer’s inability to persuasively justify the ends he proposed inspired Habermas to the theory of universal pragmatics, although whether this offered a true solution is still subject to debate.

However, there are other aspects of Horkheimer’s account that retain more plausibility. His critique of the current ideology of reason and the epistemologies of the social sciences yields three corrective prescriptions which are motivated by both political and epistemic concerns.

• (1) First, for reasons that Horkheimer might call ‘materialist’, there needs to be embedded within the philosophy of social science, and within epistemology, some sort of reflexive ideology critique, or an accounting of the contextual influences that have potential to limit the practices of reason. The development of such accounts should be a routine feature of seeking justification. When such a critical contextualization is not undertaken, or derided in some way, the real problems of the hegemony of instrumental reasoning and the eclipse of reasoning over ends are magnified. To correct for this, he calls for a kind of political self-accounting, which would standardize the following kinds of questions for any project in the social sciences. How is our methodology and theory of justification concordant with dominant social conventions of belief and practice? How do they function in favour of these conventions? What is the procedure by which the ends toward which our epistemic activity is aiming are decided upon? Who is involved in the procedure, and who is excluded?

• (2) Second, knowledge in the social sciences needs to be guided by a dialectic between social theory and the working class and other social strata that have been socially marginalized. Some have argued that the production of knowledge needs to be made democratically inclusive so as to correct for shared cultural bias. Horkheimer’s claim is stronger, or perhaps more specific. The base level of social theory needs to be a description of reality as workers experience it, from their vantage point. This knowledge is not available to all, yet it is crucial to the development of social analysis.

• (3) In particular, we need to analyse how our mode and object of study may be the product of a certain praxis. What we are measuring is what the current political economy has, to a large extent, produced, such as recidivism rates linked to race, or wage scales linked to gender. The sphere of the measurable has also been socially produced. This calls for a kind of reflexive accounting that would ask, not just how such phenomena are produced socially, but how they are produced epistemically. How do the objects of empirical study manage to fly under the radar of reflexive critique?

To summarize Horkheimer’s account, we might say that existing knowledges and epistemologies are inherently political, in that their content as well as their procedures are the product of social forces. These are contingent relationships, but what is not contingent on his view is the fact that a reflexive critical reason will lead to determinate beliefs and action. In the era of climate catastrophe, this approach appears plausible. Horkheimer maintains that, freed of instrumentalism, reason carries within itself the potential elements of a political salvation. To a large extent, his account relies on familiar epistemic virtues such as openness, self-critique, public accountability, and fallibilism, and he remains hopeful that contextual distortions of inquiry can be identified and disempowered. But his concern with the social construction both of both knowers and the known leads him to argue that we must take responsibility for the effects of the process of inquiry itself, and thus engage politically with social transformation. For Horkheimer, then, an epistemic concern with human knowledge will aim higher than mere truth.

III

Habermas. I am most interested here in the early arguments Habermas made in his 1968 publication of Knowledge and Human Interests, in which he pursues epistemological concerns entirely through an accounting of the history of philosophy.³

Habermas reminds us that the original meaning of theory involves an interested cultivation of self.

The theoros was the representative sent by Greek cities to public celebrations. Through theoria, that is through looking on, he abandoned himself to the sacred events. (Habermas 1968, p. 301)

The theorist aims to bring himself into accord with the event, and opens himself to its effects, through which he will bring ‘himself into accord with the proportions of the cosmos’ (Habermas 1968, p. 302). We can understand this in the alethic tradition of seeking understanding. In this tradition, theory has a twofold movement, first toward the cosmos, but also toward the self, whose conduct and form must be moulded to reflect the ordered motion it discerns in Being so that it can achieve understanding. Thus the Platonic conception connects theory even in its pure form with the conduct of life, and affords an internalist version of justification, insisting that a particular formation of and among theorists is necessary for successful science.

This conception of theory has a certain affinity with positivism, as Husserl ([1936] 1970) pointed out. Both assume that the theorist must free himself ‘from dogmatic association with the natural interests of life and their irritating influence’, (that is, everyday particular interests), and both intend to describe the universe ‘just as it is’ (Habermas 1968, p. 303). Husserl finds this approach to be naive precisely because the constitution of the objects of consciousness is concealed and thus exempted from analysis. Only a phenomenological reduction will bring the life-world to consciousness and definitively free ‘knowledge from interest’ (1968, p. 305)

It is on this point that Habermas marks his departure from Husserl and makes his claim to be the true inheritor of the Greek tradition. In his view, it is only in acknowledging the connection of knowledge to human interests, rather than claiming transcendence from them, through which theory works properly. Both positivism and Husserlian phenomenology promised to free the theorist from the irritations of everyday life and create the conditions for a pure perception of the cosmos. Yet this breaks with the classical approach to theory, Habermas points out, because ‘the release of knowledge from interest was not supposed to purify theory from the obfuscations of subjectivity but inversely to provide the subject with an ecstatic purification [only] from the passions’ (1968, p. 306). It is not neutrality that will ensure emancipation, but an interested engagement with universal human interest. From this basis, then, Habermas intends to take up the classical tradition once again in order to refashion a modern post-positivist epistemology.

From here Habermas elaborates his tripartite division, which judges knowledge not on the basis of whether it is interested but on the basis of which interests it pursues. The empirical sciences are based on a technical interest in the prediction and control of our environment, the historical-hermeneutic sciences incorporate a practical interest in social intercourse and the communication of meaning, and the critical sciences incorporate the interest in emancipation. But does this division enable the instrumental reason Adorno and Horkheimer criticized? After all, it might be argued, our technical interest in prediction will not be satisfied by an approach obscured by wishful thinking or political agendas; and thus once again it is the transcendence of narrow interests on Habermas’s view that is more likely to provide both reliable claims and an effective satisfaction of our needs.

Habermas’s account offers the following answer to this concern. The first part of his answer once again involves the ontology of truth, or the character of what can be known. Here he works toward a similar conclusion to Horkheimer’s, but he goes about it in a different way. Habermas explains that his difference with Husserl is not simply over the epistemic pretensions Husserl accords to the phenomenological reduction, but also over Husserl’s acceptance of the account of Being given in the Platonic ontology. The classical tradition’s aim of ‘describing the universe theoretically in its lawlike order, just as it is’ (Habermas 1968, p. 303) is connected, in Habermas’s view, with the bifurcation the Greeks made between logos and doxa, reserving only to logos the ‘realm of Being purged of all inconstancy and uncertainty’ and leaving to doxa all that which is ‘mutable and perishable’. It is when the aim of knowledge is conceptualized as logos that the interested character of human processes of knowing become epistemically problematic. How can an interested procedure, mired in the immanent features of a localized, inconstant lived human experience reach Being when the latter is conceptualized as beyond doxa? ‘Representations and descriptions are never independent of standards’, he argues, and these standards ‘cannot be either logically deduced or empirically demonstrated’ (1968, p. 312). To the extent that standards have a foundation, it is in the ‘natural history of the human species’ (1968, p. 312; compare this view with other historicist philosophers of science such as Lakatos). Given that justification procedures rely on representations, or objects under a description, justification processes themselves cannot be completely disentangled from human interest. But this does not compromise our ability to reach truth unless we accept the bifurcation between doxa and logos. In other words, we must fallibilize or contextualize truth.⁴

So the first part of his argument concerning the role of interest in processes of justification is that the object of knowledge is not unchanging Being but a reality closer to home, as it were. The second part of his argument is that the interests that influence knowledge are, ideally, universal across the species and objectively determinable. Each of the three interests that the sciences pursue—the predictive, communicative, and emancipatory, which are associated with the three main aspects of human life: work, language, and power—can be supported and studied objectively. Human physical needs and interactive needs, corresponding to our predictive and communicative interests, are fairly straightforward, but Habermas’s originality is in the empirical support he finds even for our emancipatory interest within language. His argument for this claim is similar to Horkheimer’s: that an unencumbered reason will be inevitably anti-conventional, robustly reflexive, critical, and in this sense liberatory.

By characterizing these three interests as having an objective basis in human life, Habermas can then argue that the obfuscations and distortions produced by particular interests can be rightfully rejected. Knowledge does not follow from just any given human interests, but only the collective, objectively determinable interests. These alone are expansive enough to generate the undiminished openness of argument that reason requires. And the possibility of objectivity, in turn, is dependent on a reflexive and accurate assessment of the fundamental interests operating in thought (Habermas 1968, p. 311). The substantive character of truth claims is not compromised by the necessary presence of human interest because, he argues, human interests are objectively determinable, objectively justifiable, and will bring us to objective truth (in the sense of a contextual, fallible truth defined as one that does not require a bifurcation between Being and doxa).

Habermas thus redeems the pragmatic interests that Horkheimer repudiated by arguing that they are more expansive than he allowed, even the interest in prediction and control. Like Foucault, Habermas also points out that our knowledge transforms us. Unreflective consciousness ‘can be transformed … [and] to this end a critically mediated knowledge of laws cannot through reflection alone render a law itself inoperative, but it can render it inapplicable’ (1968, p. 310, my emphasis) Human agency has a transformative effect on objective reality.

Habermas thus argues for an intrinsic connection between knowledge and power which then mandates a reflexive relationship between epistemology and politics. Justification requires a pursuit of emancipatory interests in order to create the conditions for the possibility of a meaningfully collective and reflexive dialogical processes.

The ontological illusion of pure theory behind which knowledge-constitutive interests become invisible promotes the fiction that Socratic dialogue is possible everywhere and at any time. (Habermas 1968, p. 314)

This fiction allows philosophy to refuse to take responsibility for the real, existing external conditions which ensure our descent into ideology. Against such practices of pure theory, Habermas promotes an epistemology that will recognize that it does not provide the foundation for itself, that in fact its own foundation—its own legitimacy—lies in the pursuit of the emancipatory interests that will strengthen communicative practice and throw off its current constraints. In this way, Habermas establishes objective and epistemological justifications for the emancipatory values pursued in critical theory. Even when the working class votes conservative, social democratic values can be defended.

Where Horkheimer critiques instrumental rationality wholesale, Habermas critiques only the hegemony of the pursuit of prediction and control. Connected to our other species-wide interests, the empirical sciences need not be oppressive. But he too calls for the political reflexivity and acknowledgement of praxis that Horkheimer elaborated. Horkheimer argued that traditional theory, which consists of instrumental, pragmatic rationality, needs to be replaced with critical theory, and that the traditional epistemic virtues of reflexivity and so on will then work toward political emancipation. Habermas provides a more substantive account which accords the pursuit of prediction and control its place, but also argues for an epistemic interest in creating a form of society conducive to the operation of epistemic virtues. In other words, where Horkheimer calls for a critical theory which can elaborate and defend a transformative social vision, Habermas’s epistemology contains this vision within itself, and provides it with an internally generated justification.

But one might view Habermas as simply arguing that the effective pursuit of knowledge requires democratic conditions, making the relationship between knowledge and power extrinsic rather than intrinsic. Anyone who wishes to pursue knowledge has an inherent interest in promoting the political conditions that will optimize knowing, but the processes of justification can then advance unhindered by political concern. And the interests that organize human knowledge do not constitute that knowledge. The truth about what will meet our physical, communicative and emancipatory needs is itself not shaped by those needs.

The fact that Habermas emphasizes the need to separate power from knowledge in his critiques of Foucault lends further credence to this interpretation of his view. His view contrasts with that of Horkheimer, who calls for a dialectical method that is based on a particular class of knowledge. Even in regard to critical theory, or knowledge guided by emancipatory interests, Habermas would privilege no particular class: justification is guaranteed by the open procedure of undistorted communication, and its norms are given in intersubjective communicative praxis, without attendance to any particular agenda of human oppression.

The contrast between their accounts can be drawn in the following way. Horkheimer’s is, I would argue, an account of reason that aims toward an objective truth in relation to universal human interests, whereas Habermas’s account describes a reason that is more particular. Horkheimer’s account of reason is immanent because it does not carry within itself its own normative justifications, but he rests assured that it can achieve these through an open and dialectical process. Its normative aims are to be found in the future and by current lights, as Adorno ([1973] 1981) argued, it will look unjustified. Habermas, by contrast, claims that both the form and content of the normative aims that provide the justifying foundation for his procedural account of reason are already present, as they have always been, in communicative practice. David Ingram has argued that despite this, Habermas’s account of reason is not transcendental,

if one means by ‘transcendental’ conditions whose necessity and universality can be ascertained with reflexive certainty. Since these norms are regulative for public communication, their proper interpretation and confirmation can occur only in actual discourse, not in private introspection. (Ingram 1990, p. 226)

Yet Habermas has claimed precisely this, that is, to have ascertained the norms through an a priori analysis of the pragmatic conditions of linguistic communication. On his theory of justification, ultimate validity rests with the community of theorists, but this does not abrogate the essentially transcendental character of the norms guiding this general approach to justification. Discourse is thus required for validity, but the very discourse that engages with this question will be structured and judged in light of the norms themselves, and Habermas does not historicize the norms embedded in communication. Given that these norms are necessary for any successful communication to occur, they are indeed transcendent of particular contexts. This means that Habermas’s communicative reason is not open to historicist change or fallibilism. In contrast, Horkheimer’s account has the advantage, in my view, of historicizing the operations of reason itself, and thus allowing it to remain open to the epistemically relevant developments of cultural encounters and political transformation.

IV

Foucault. In this section, I want to consider how Foucault’s accounts of knowledge and of epistemology maintain a normative element, despite their constitutive relations to what he calls ‘power’. Foucault’s work can be read as an extended set of case histories concerning the feedback loop of modes of inquiry, legitimated knowledge, and forms of experience and subjectivity. From these case histories he advances the general view that both truth and knowledge are in some sense constituted by power relations. This is also true of epistemology. The psychiatrist’s claim of knowledge about our deep self is simultaneously a claim of authority, and thus a kind of power move, and the human sciences generally operate to structure various dispersed fields of possibilities for action, which is Foucault’s definition of power. Power is an institutional precondition for the elaboration of a form of knowledge, as Dews (2007) observes, but Foucault complicates the causal assumption his account seems to imply when he argues that there is ‘no power relation without the correlative constitution of a field of knowledge’ (Foucault 1979, p. 27). He says ‘power and knowledge directly imply one another’, and moreover that power and knowledge are also interconnected with pleasure: pleasure spreads power, power anchors and institutionalizes pleasure, and these relations between pleasure and power operate through knowledge (1979, pp. 27, 45, 48).

Because these connections are so ubiquitous, Foucault abandons the ideology/science distinction as well the interested/disinterested distinction (1979, p. 28). Given that all knowledge seems to operate in relation to power, it seems to follow that Foucault would have to reject the distinctions Horkheimer and Habermas invoke between more or less encumbered, limited, or non-reflexive forms of knowing. Foucault has certainly been read as collapsing all knowledge to power, in which case all knowledge reduces to ideology. However, what complicates this picture of Foucault as an epistemic nihilist is his vigorous and irritated protestations against such readings, as in the following:

I know that, as far as the general public is concerned, I am the guy who said that knowledge merged with power, that it was no more than a thin mask thrown over the structures of domination … [This] is so absurd as to be laughable. If I had said, or meant, that knowledge was power, I would have said so, and, having said so, I would have had nothing more to say … (Foucault 1990a, p. 264)

This comment suggests that Foucault would agree with those (like Habermas) who find epistemic nihilism incoherent. If knowledge is reducible to strategy, and known to be so, it cannot have its intended strategic effects. In the same interview, Foucault goes on to explain:

[If I had] made [power and knowledge] identical, I don’t see why I would have taken the trouble to show the different relations between them … What I set out to show was how certain forms of power that were of the same type could give rise to bodies of knowledge that were extremely different both in their object and in their structure … We have, then, power structures, fairly closely related institutional forms—psychiatric confinement, medical hospitalization—that are bound up with different forms of knowledge, between which it is possible to draw up a system of relations based not on cause and effect, still less on identity, but on conditions. Those who say that for me knowledge is the mask of power seem to me to be quite incapable of understanding. It is hardly worth answering them. (Foucault 1990a, pp. 264–5)

The mistake is in thinking that power is operating causally ‘behind’ knowledge. Since power is ubiquitous across the social field of human action and interaction, knowledge is never outside power; it is produced in this field. Thus the inescapability of the relation between power and knowledge that Foucault argues for is simply to follow Habermas himself in asserting something similar to the ubiquity of interests in human knowledge. Foucault asserts, not the ubiquity of objective human interests, but the ubiquity of detached, anonymous interests operating alongside or within strategies of power wherever knowledge appears.

It is true that Foucault repudiated the distinction between science and ideology, and that he famously argued that ‘It’s not a matter of a battle “on behalf” of the truth, but of a battle about the status of truth and the economic and political role it plays’ (1980, p. 132; see also 1980, pp. 117–19, 131–3). However, Foucault’s view was never that power controls knowledge, but that it is ‘power-knowledge, the processes and struggles that traverse it and of which it is made up, that determines the forms and possible domains of knowledge’ (1979, p. 28). This is not power speaking as if it were knowledge, but a circular relation of partial mutual constitution, a relationship in which each part is necessary but never sufficient for the other. Though the formulation of truth is dependent on power, this argument in no way entails that truth is mere illusion, or arbitrary, or unconnected to a shared, extra-discursive reality. Thus Foucault can be read as the sort of historicist post-positivist epistemologist who sees descriptive science as historically contextual and fallible but simultaneously efficacious in discerning features of the perceptible world.

Moreover, Foucault offered an epistemic defence of his celebrated category of subjugated knowledges alongside a political defence (see Alcoff 1996). He argues that ‘global theories’—that is, those that seek universal, generic application—are useful only at the price of curtailing, dividing, overthrowing and caricaturing non-global discourses, and thus have proved to be, as he says, a ‘hindrance to research’ (Foucault 1980, p. 81) The problem with global theories lies not only with their political effects but with their dismissive approach toward concrete, particular events that cannot be reduced or adequately included in their terms. Hegemonic knowledges always work through distortions and omissions at the local level in order to enable the reductionist move of containment. Subjugated non-hegemony-seeking knowledges do not require the amount of violence, distortion and omission that hegemonic knowledges require. Note that this is both a political and an epistemic concern.

Foucault characterizes his own genealogical strategy toward knowledge as an

attempt to emancipate historical knowledges from that subjection, to render them, that is, capable of opposition and of struggle against the coercion of a theoretical, unitary, formal and scientific discourse. It is based on a reactivation of local knowledges … in opposition to the scientific hierarchisation of knowledges … (Foucault 1980, p. 85)

Dews (2007) charges Foucault with making power-location the single most important normative criterion by which to judge knowledge, meaning that political rather than epistemic criteria determine Foucault’s evaluations. I think this misunderstands what Foucault means by location. Even if Foucault were to use power-location as the ultimately most important criterion, this criterion itself is not extra-epistemic. Foucault does not propose a strategy of the political against the epistemic, but of one set of power/knowledges against another. He privileges subjugated knowledges because they resist the universal pretensions of global theories which would forcibly subsume particulars under a generalized rubric. In itself this is insufficient for an epistemic analysis, since it does not provide the means to adjudicate competing subjugated knowledges, but it is an epistemic analysis.

Because Foucault’s arguments include independently operating epistemic criteria (independent conceptually, even if never pragmatically), I take Foucault to have believed in some substantive version of truth. After all, he never called for a universal retreat from truth claims (as Wendy Brown (2005) and Allen (1995), for example, claim), but for ‘detaching the power of truth from the forms of hegemony, social, economic, and cultural, within which it operates at the present time’ (Foucault 1980, p. 133).

In another passage from his late work, The Use of Pleasure, Foucault says the following:

I opted for [a hermeneutics of the self], reasoning that, after all … what I have tried to maintain for many years is the effort to isolate some of the elements that might be useful for a history of truth. Not a history that would be concerned with what might be true in the fields of learning, but an analysis of the ‘games of truth’ … and error through which being is historically constituted as experience; that is, as something that can and must be thought. (Foucault 1990b, pp. 6–7, my emphasis)

I suggest that this account is truly in the spirit of Horkheimer and Habermas—both of whom emphasized the analysis of the effects of praxis—both epistemic praxis and institutionalized, material forms of praxis. Foucault explains his project as analysing the mechanism through which ‘being’ is historically constituted as experience. The mistake that Habermas himself makes is to see Foucault’s approach as rendering social reality into mere illusion or chimera, as if the knowledge one gets is ‘merely’ human-made, or discursively constructed, in short, idealist. It seems to me that all three philosophers bypass idealism, and are best characterized, not as epistemic nihilists, or idealists, but as working with some form of Putnam’s internal or immanent realism—that is, a realism on the ground, among us, offering contextual and thus fallible formulations in the sense of time and space, but in this way a more accurate rendering of truth itself.

I would argue, then, that Foucault holds something like the views Horkheimer and Habermas hold about the object of knowledge: for them it is the product of praxis, for Foucault it is the product of discursive formation. But if one can grant to Horkheimer and Habermas the possibility of continuing to believe in a substantive version of truth despite the constitution of perceptual data, then we must grant to Foucault the same possibility. The goal of Foucault’s work is not to replace a concern with truth, but to pursue it with a more adequate understanding of the social elements of its production.

V

Conclusion. The project of a political epistemology, then, needs to address in a much more expansive and rigorous way the social context in which both knowledges and knowing are validated, and include how this occurs within philosophy itself. A political epistemology cannot argue for the independence of epistemologies from political preconditions and political effects, no more than we can imagine an economic process without such preconditions and effects.

This should demonstrate that political epistemology is vital to the decolonial turn. But we need to ask a new set of questions within a larger context of colonial history and ongoing colonizations, and especially within the forms of extractive capitalism still ongoing that operate with normative conceptions of race, cultural advancement, development, and progress to clear their instrumental pathways. We need to ask how the frameworks, concepts and questions of modern European epistemology emerged in its historical context, coterminous with European conquest and colonial expansion. Horkheimer and Habermas suggest ways in which traditional empiricism functioned for capital; Foucault considers the validation of the social sciences in relation to heterosexism and the production of a docile citizenry. Their approaches help to make political epistemology a formidable project against technocracy and epistemic elitism. Yet none of them raised the issue of colonialism. That should be the task ahead.

References

Footnotes