Troels Krarup. Theory, Culture, and Society. Volume 38, Issue 5, February 2021.
Introduction: Why ‘Archaeology’?
In spite of more than 50 years of reception of Foucault’s archaeology, we lack a clear and adequate specification of what is particularly ‘archaeological’ about the methodology and how it relates to the study of knowledge. Certainly, we have at our disposal numerous readings of Foucault’s own account of his ‘archaeology’, but the problem is that he never stated explicitly why the term ‘archaeology’ was adequate. However, through what one could call an ‘archaeology of archaeology’ itself, that is, an analysis of the space within which archaeology emerged as a distinct methodology for the study of knowledge, it is possible to provide a number of specifications and clarifications. Moreover, conducting such an analysis and methodological specification will allow us to draw up the contours of what may be called ‘problem analysis’ in social research today.
Existing literature on Foucault’s archaeology establishes the following points of importance to the present argument: (a) Foucault’s work may be read methodologically, as opposed to theoretically (Koopman, 2013); (b) the French epistemological school and Hegel were influential to the formation of Foucault’s thought (Gutting, 1989, 2010); (c) the history of science must address not only solutions and answers but also problems (Lecourt, 1975; Webb, 2012); (d) archaeology concerns the conditions of controversy in discourse, which are irreducible to social structure and interests (Kennedy, 1979); (e) it opposes the hermeneutical interpretation of meaning or intentions (Hacking, 1986); (f) it is a kind of work of thought upon itself in a new conception of ‘enlightenment’ (Raffnsøe et al., 2015). However, the literature lacks a systematic exposition of the distinct methodological relationships in archaeology between discourse, problems, conditions, and comparisons. Correspondingly, certain areas in our map of the intellectual field from which Foucault’s archaeology emerged remain almost entirely blank, especially the French tradition of the history of philosophical systems of thought, the comparative mythology of Georges Dumézil, and the distinct post-Hegelian conception of ‘problems’. Clarifying and establishing these points will lead us to abandon certain received views. For example, the conventional partitioning of Foucault’s work into different periods may have some value in terms of distinguishing thematically different enquiries, but it fails to account for the continued use of archaeological methodology throughout (although sometimes under different labels). Similarly, we do not find insurmountable dilemmas at the core of archaeology that forced Foucault (and should force us) to abandon the methodology—such as an ontological dilemma of assigning transformative force either to anonymous structures or to ‘the reflective activity of the subject’ (Han, 2002: 189), or a dilemma of static structures versus historical change (Koopman, 2013: 40-1).
To begin with a seemingly simple question: Why did Foucault call his methodology ‘archaeology’? The question has only rarely been posed and has so far not been given a satisfactory answer, arguably because lacking an adequate account of the historical formation of the concept (see Kusch, 1991; Paltrinieri, 2015; Sabot, 2006). It is often stated that archaeology offers a method for the study of the historical emergence of our structures of knowledge, bridging the two disciplines of philosophy and history (Osborne et al., 2015: 24). If that were the case, why then did Foucault not simply call his methodology the history of knowledge? As early as in 1961 in the preface to History of Madness, Foucault stated: ‘My intention was not to write the history of that [lost] language’ in which madness and reason communicated, ‘but rather [to] draw up the archaeology of that silence’ in which only reason speaks and madness is reduced to mental illness (Foucault, 2009 [1961]: xxviii). What is the difference here between history and archaeology? The question is not trivial, since Foucault used the term in the titles of almost all of his works dating from the 1960s and formulated his methodology, The Archaeology of Knowledge, under it (Foucault, 1972a [1969], 2003 [1963], 2005 [1966]). Moreover, Foucault did not abandon the notion, but continued throughout his life to term his method ‘archaeology’ (e.g. Foucault, 1984a).
I argue that ‘archaeology’ furnished Foucault with a response to a methodological problem that he faced when writing History of Madness (Foucault, 2006 [1961]). I demonstrate how, in Madness and in later books, Foucault pushed the boundaries of an existing and well-defined field of research on ‘the history of systems of thought’ beyond its two main currents. These dealt, respectively, with philosophy (the history of philosophical systems) and science (the French epistemological school). Foucault advanced these into a third line of study—that of ‘knowledge’ in the broadest possible sense. The history of systems of thought had developed a methodology for the comparison of contradictory philosophical systems and the French epistemological school had focused on problems of contradiction within scientific systems. With the expansion of focus to knowledge, Foucault could no longer assume that contradictions were rooted in opposing claims to universal truth. In response to this challenge, he adopted from the historian of mythology Georges Dumézil the archaeological methodology devised not for material culture but for written sources. Later, he forged a notion of ‘problematization’ closely aligned with the post-Hegelian correlates (‘problems’ and ‘problematics’) already developed within the French epistemological school.
The argument has more than historical interest as it reopens the horizon of archaeological inquiry for social research today (Dean, 2015). Thus, I conclude by drawing up the contours of an approach to the study of knowledge designated ‘problem analysis’, which may be conducted without resorting to the widespread reading of Foucault as a social theorist of power/knowledge. Problem analysis takes manifest contradictions as access points to the study of the more fundamental problems that are common to the contradictions as their condition of existence and that organize a given system of thought, its historical descent and evolutionary morphology (Krarup, 2019a, 2019b).
The Field of the History of Systems of Thought in France around 1960
When Foucault was appointed to a chair at the prestigious Collège de France in late 1969, he chose for it the title ‘History of Systems of Thought’. This placed him firmly in a tradition already well established at the Collège. For several years, senior professors—especially Jean Hyppolite, Jules Vuillemin and Georges Dumézil—had worked to pave the way for Foucault’s entry there (Eribon, 2011: 336). These represented also some of the most influential teachers and mentors for the young Foucault in the 1950s and 1960s. Jean Hyppolite (1907-1968) had been Foucault’s teacher of philosophy in high school, his contemporary director of the École Normale Superieure, then member of his thesis board. Jules Vuillemin (1920-2001) had served as the head of the philosophy department at the University of Clermont-Ferrand, where Foucault got his first teaching position in 1960; the two had befriended each other, and Vuillemin officially presented Foucault’s candidature at the Collège. Georges Dumézil (1898-1986) had ensured Foucault his post as French cultural diplomat at Uppsala in central Sweden in 1955—a post he had himself occupied some 20 years earlier. Here, Foucault gained access to the rich Carolina Rediviva Library and wrote his thesis (published as Madness in 1961) based on a hitherto unstudied collection of medical writings from the 16th through the 18th centuries.
The three figures already mark out a certain diversity in the field of the history of systems of thought in France around 1960. In its most traditional form, this current emerged from the history of philosophy, but it also encompassed the history of science with the French epistemological school. When Foucault obtained his chair at the Collège de France in 1969, he had become the most important proponent of a third direction of study within the field into the history of ‘knowledge’. Figure 1 illustrates the field and positions its most important personalities. Needless to say, any schematization implies choices and cannot be considered complete. For other purposes than the mapping of a methodological field, figures such as Blanchot, Saussure, Benveniste, Nietzsche and Heidegger could have been included. For example, as Webb (2012) has shown, Heidegger’s influence on Foucault was thematic rather than methodological. Another caveat to Figure 1 is that the field is embedded in the much broader fields of Parisian intelligentsia (Riley, 2010) and French social theory from Saint-Simon onwards (Gane, 2003). However, the aim of archaeology is not to trace the filiation of ‘precursors’ but to determine the limits of a discursive space (Lecourt, 1975). For example, Dumézil’s fascination at a young age with Durkheim’s sociology (which took ancient myths to represent social reality) is irrelevant to the present analysis given his later explicit denial—on archaeological grounds—of a ‘hypothetical reconstruction’ of ‘religion, social organization, and politics as they might really have functioned’ (Dumézil, 1949: 29). Similarly, it is not enough, for example, to observe that Saint-Simon (1975) also talked about religious and scientific ‘systems’ and of ‘series of comparisons’ in order to include him in the field. These are only ‘moments’ and not in themselves an ‘epistemological structure’, which may have a plane of simultaneity very different from historical time (Foucault, 2008: 117, 167). Consequently, the analysis of the field of the history of systems of thought may go as far back as around 1800 (Kant, Gérando, Hegel) in some respects, and halt sometime around 1930 (Dumézil) in others.
Figure 1. The field of the history of systems of thought in France around 1960
History of Philosophical Systems or Philosophizing History of Philosophy
Vuillemin and Guéroult were defining figures for the history of philosophical systems. In What Are Philosophical Systems? Vuillemin (1986) writes that such systems must be considered immediately different and even irreconcilable because philosophy by definition claims universality and coherence. Thus, philosophical systems are mutually contradictory (Vuillemin, 1986: vii). This ‘plurality of philosophies’, Vuillemin (1986: ix) continues, ‘makes the concept of philosophical truth inadequate and inappropriate, at least if the word truth is used in its ordinary sense’, giving way instead to a kind of enquiry based not on methodological validity and guarantees, but only on ‘a prudence fraught with risks and further queries’.
In this, Vuillemin follows closely his predecessor at the Collège de France, Martial Guéroult (1891-1976) (Guéroult, 1979: 64). In 1951, Guéroult had entitled his Collège chair ‘History and Technology of Philosophical Systems’ (clearly echoed in Foucault’s later title). Guéroult exercised a formative influence on philosophy in Paris up to the early 1960s, including on Foucault as attested in Archaeology (Libera, 2016: 16). In Philosophy of the History of Philosophy (published only posthumously but first drafted in the late 1930s), Guéroult (1979) addressed the apparent impossibility of a ‘history of philosophy’ given the immediate contradiction between ‘history’ (in which philosophical systems were understood as different ‘facts’) and ‘philosophy’ (in which the claims of each system to universality, coherence, and truth would be assessed). Guéroult sought a way out of the dilemma by beginning with the immediate ‘experience’ of the fecundity of working on the ‘internal reconstruction of [philosophical] doctrines according to their own organization’ (Guéroult, 1979: 50-2).
In so doing, Guéroult explicitly continued the line from Joseph-Marie de Gérando (1772-1842), in particular his Comparative History of Philosophical Systems (Gérando, 1804). Gérando distinguished the history of philosophical systems both from encompassing but sterile narrative history and from the opinionated subsuming of all systems under one’s own preferred standpoint. Instead, he formulated a ‘comparative history’ based on a new ‘method of experiences’ that ‘converts the facts [of the different philosophical systems], through a succession of parallels, into so many experiences on the journey of the human spirit’ (Gérando, 1804: xxiv). Comparative history was to proceed in two steps: first, as a presentation of the different philosophical systems in a simple classification and chronology; second, in a ‘critical analysis’ of the systems taken ‘as grand experiences, as so many grand givens that can—through methodological comparison of motifs, proofs and effects produced—shed new light on the fundamental question [of philosophy] itself’ (Gérando, 1804: xxi-xxii).
Thus, from its very inception, the history of the systems of thought was not only a discipline with a specific object of inquiry, but also a distinct comparative methodology. However, in Gérando’s formulation, the history of philosophical systems remained predominantly philosophy because the purpose of such a ‘meta-analysis’ of philosophy was to ‘fix our choice of system on the one that turns out to be the best’ (Gérando, 1804: xxii). This motivation can be found as late as in the works of Octave Hamelin (1856-1907), such as Descartes’s System (published posthumously and, incidentally, with an introduction by his friend Durkheim; Hamelin, 1911). Nonetheless, precisely this assumption about history as a golden route to truth (philosophical or scientific) was rejected by Guéroult, who insisted that the immediate experience of the fecundity of the history of philosophy entailed not a priori specification of what that fecundity consisted in, nor any teleology of where it would lead. Nonetheless, the aim remained to raise the question of ‘the condition of possibility of the system of experience’ through the comparison of historically specific philosophical systems (Guéroult, 1979: 54). Without a priori positing the existence of a latent universal human reason to be discovered through the analysis of history, Guéroult maintained that the apparent opposition between history and philosophy could be overcome by ‘the spirit’ (l’esprit) through the study of how it was ‘created and revealed to itself’ as spirit by history (Guéroult, 1979: 36). (As we shall later see, this post-Hegelian argument anticipates a key point in Foucault’s reflection on enlightenment as thought’s reflexive process of freeing itself from itself.) According to Guéroult, this implied a new concept of ‘reality’, one that does not rely on a priori criteria for ‘true philosophy’ but considers all philosophies as ‘equally true’ insofar as they exist (Guéroult, 1979: 254). Thus, ‘philosophizing thought dedicates itself to the “reality” of the “thing”’, that is, to the philosophical doctrines it ‘experiences’ (Guéroult, 1979: 95).
A Kantian Cul-de-sac
We shall see how the methodology of comparing contradicting systems became important to Foucault, but let us first consider his own sole account of the origins of his use of ‘archaeology’. Foucault points to Kant in whose definition he claims that ‘archaeology’ designates ‘the history of that which renders necessary a certain form of thought’ (Foucault, 1971: 60). Foucault does not refer to the supposed text of Kant—in fact, such a reference, let alone discussion of Kantian archaeology, is absent from his entire oeuvre (McQuillan, 2010: 39). According to a number of scholars (Kusch, 1991; Libera, 2016; McQuillan, 2010; Paltrinieri, 2015), Foucault’s reference goes to Kant’s employment of the term ‘archaeology’ in the ‘jottings’ (löse Blätter) for his unfinished essay on the Progress of Metaphysics (Kant, 2002 [1804]). Here, Kant uses the term just once to designate what the editor (in the title he gave the jottings, sometimes mistaken for Kant’s own) called a ‘philosophizing history of philosophy’. The question Kant raises (in just eight rather heavy-going lines) under the label of ‘philosophical archaeology’ concerns the origin of human questioning about the beginning and end of the world and of itself. This question he defines as ‘rational’ rather than ‘historical or empirical’, as it ‘does not borrow [its facts] from historical narrative, but draws them from the nature of human reason’ (Kant, 2002 [1804]: 417).
It is not obvious how this adds up to ‘the history of that which renders necessary a certain form of thought’, although there is a degree of resemblance to Gérando’s position. Indeed, Guéroult echoes the Kantian notion precisely when he references Gérando, saying that the history of philosophy ‘appears to philosophizing reason as an always relevant and fecund experience’ (Guéroult, 1979: 50, emphasis added). However, Kant does not provide the same methodological specification as Gérando, nor does he seem concerned with the same problem (that of co-existing but mutually contradictory systems of thought). Elsewhere, Kant provides some additional elements, but remains far from describing an archaeology of philosophical systems, let alone one of knowledge. In Anthropology from a Pragmatic Point of View, the ‘archaeology of nature’ is mentioned in passing as an already-established discipline, treating natural phenomena as signs from which unseen nature can be inferred (‘if there’s smoke there’s fire’) (Kant, 2006 [1798]: 87, see also 227). In Critique of the Power of Judgement, Kant described the archaeology of nature in relation to living organisms as ‘comparative anatomy’ in search for a ‘system … regarding the principle of their [the different species] generation’ (Kant, 2000 [1790]: 287). Both works struggle with the problem of characterizing the human being as simultaneously a product of nature, which blindly dictates the pursuit of egoistic pleasure, and of its own understanding (Vernuft), which gives it freedom and makes it ‘capable of perfecting [itself] according to ends that [it] adopts’ (Kant, 2006 [1798]: 226). Kant’s explicitly teleological characterization of the human being is thus: ‘to bring about the perfection of the human being through progressive culture, although with some sacrifice of his pleasures of life’ (Kant, 2000 [1790]: 226).
Foucault translated Anthropology as his ‘minor thesis’ along with Madness, conducting ‘structural comparisons’ between it and Kant’s three Critiques (1781, 1788, 1790) and identifying a ‘rupture’ with the old ‘problematic’ of metaphysics and the formation of a new one—that of the human being (‘man’) as a finite being in the world (Foucault, 2008 [1961]: 75, 105, 117). However, this new problematic was not clear to Kant himself—nor to the long tradition of phenomenological philosophy following him. Rather, the notion of a vivifying ‘spirit’ (Geist) of ‘ideas’ in the human being is ‘secretly indispensable to the structure of Kantian thought’ as its ‘epistemological structure’ (Foucault, 2008 [1961]: 65, 116). The analysis follows very closely (but does not cite) that of Heidegger (1998 [1929]) and clearly anticipates the broader one in The Order of Things (Defert et al., 2008: 7). However, there is no explicit methodology for the archaeology of thought in Kant or Heidegger—nor can one be derived from the new problematic of the human being, which Foucault calls ‘illusory’, announcing its replacement by Nietzschean philosophy (Foucault, 2008: 123-4). The ‘birth’ of archaeological methodology is not with Kant but must be found elsewhere.
Dumézil and the Archaeology of Knowledge
When Foucault worked on Madness in the late 1950s, he simultaneously encountered a methodological problem and its solution. As Canguilhem (1998: 26) points out in his thesis report, the novelty of Madness lies not only in a philosopher doing original archival work, but also in his submitting other forms of knowledge, such as psychiatric knowledge, to philosophical (or, rather, philosophizing) examination. With the expansion to ‘knowledge’ in a broad sense came deep methodological considerations. As Foucault later put it in an interview on Order, it ‘means that one will take up Don Quixote, Descartes, and a decree by Pomponne de Bellièvre about houses of internment in the same stroke’ (Foucault, 1996: 14). However, we cannot expect to find in Don Quixote or Bellièvre the same kind of claims to universality and coherence as in philosophical systems. Nonetheless, Foucault upheld the comparison of contradictory elements as a central methodological principle—and he called that methodology ‘archaeology’ (Foucault, 1972a).
Dumézil’s ‘capital importance’ to Foucault’s intellectual formation is well-documented but almost never discussed in any methodological detail (Eribon, 2011: 130, see also Eribon, 1992, 1994; Séglard, 2007). Foucault probably began reading Dumézil’s main works systematically upon his arrival in Uppsala and met him in person for the first time a few months later (Eribon, 2011: 132). Dumézil is best known for his theory of a single prehistoric Indo-European civilization, traces of which can be seen in the characteristic tripartite structure common to the mythologies of later Indo-European civilizations as far apart as the Vedic (India) and the Old Norse. More importantly, in his work, Foucault discovered a strict comparative methodology (Poitevin, 2003; Eribon, 1992). He made Dumézil’s importance for his work explicit already in the preface to Madness (2006: xxxv) and more elaborately in his inaugural address at the Collège some ten years later:
It is he who taught me to analyse the internal economy of discourse quite differently from the traditional methods of exegesis or those of linguistic formalism. It is he who taught me to refer the system of functional correlations from one discourse to another by means of comparison. It was he, again, who taught me to describe the transformations of a discourse, and its relations to the institution. (Foucault, 1972b [1970]: 235; see also Foucault, 1998a)
In Indo-European Heritage in Rome Dumézil states that in response to the liberties taken by historians in interpreting the past in order to reinvigorate it ‘we are obliged to specify, next to the archaeology of objects and sites, an archaeology of representations and comportments’ more attentive ‘to the concepts and systems of concepts as they are found deposited in language, in legends and in institutions’ (Dumézil, 1949: 43-4). Thus, archaeological methodology is an attempt at a rigorous comparative study of structures understood as networks of relations in a written material (Dumézil, 1949: 31; 1955; compare Foucault, 1972a: 138-9, 2006: xviii).
Dumézil often references comparative linguistics as an exemplar of archaeological methodology. Comparative linguistics does not make analogies and interpretations but conducts series of structural comparisons in a search for correspondences (phonetical, grammatical, or semiotic) based on which common descents and trajectories of diffusion and differentiation can be established (Dumézil, 1948: 112). To comparative linguistics, it is not in itself interesting that the Latin word pater resembles the English equivalent father. However, interest surges when a ‘series’ of ‘corresponding’ p-f replacements occurs in fish (piscis), foot (pes), fire (pyr), feather (pluma) etc., so that ‘the ensemble strengthens each element’, indicating a common descent and a mechanism of differentiation between the two languages (Dumézil, 1949: 35, see also Dumézil, 1941: 29). Observing a corresponding p-f replacement in German (Vater, Fisch, Fuß, Feuer, Feder), we can further establish that it must have taken place before the differentiation of German and English into different languages. Similarly, in Madness Foucault searched for the ‘origin’ where madness was first distanced from reason in Western culture (Foucault, 2006: xxviii) and in Order he localized ruptures between epistemes across different domains of knowledge (life, wealth, speech). It is sometimes claimed that archaeology cannot account for change, but it can certainly locate and describe it, even in the absence of explicit statements about change taking place in the available sources.
In archaeology, comparison does not take place after the isolated facts have been established—rather, it is only through comparison that ‘facts’ can be established (Dumézil, 1941: 31, 1955: 427). ‘Comparative facts’—to which Foucault (1972a) devotes an entire chapter of his Archaeology—are ‘simultaneously a given and a product’ (Dumézil, 1949: 43). Consequently, archaeological inquiry cannot be limited to a predefined domain (such as ‘philosophy’ in Gérando and Guéroult). Moreover, it cannot (as is often done in social research) proceed by grouping (coding) material by listing similarities and differences between predefined instances. Instead, it must conduct systematic assessment of possible correspondences and ruptures between structures (webs of relations) across otherwise separate and heterogeneous archaeological sites, utilizing ‘the entire domain accessible to comparison without making arbitrary choices within it’ (Dumézil, 1949: 36-7; see also Foucault, 1984b: 77).
Dumézil sometimes goes so far as to say that the comparative study of myths exposes ‘the entire physiology and anatomy’ of past societies (Dumézil, 1941: 16). However, he stresses that ‘unity does not imply uniformity’ (Dumézil, 1941: 13) and that the goal is not a ‘hypothetical reconstruction’ of a language, myth, or society, but strictly a structural study of descent (common origin or not) and evolution (differentiation) (Dumézil, 1949: 30-1). Dumézil’s work is thus not a search for something ‘outside’ the myths themselves, such as a uniform collective ‘social facts’ (Durkheim) or the universal structures of human culture (Lévi-Strauss) (Eribon, 1992; Poitevin, 2003). Correspondingly, Foucault emphasizes that comparison should be ‘limited and regional’ and remain ‘at the level of discourse itself’ (Foucault, 1972a: 48, 157).
It is particularly important not to impose assumptions about the non-contradictory character of the original myth under study and on that basis ‘correct’ sources that contradict themselves. Instead, one must view contradictions as ‘forming a whole’, that is, ‘perhaps not the result of a fortuitous addition of elements, but on the contrary the principle of their organization’ (Dumézil, 1948: 48). This last point is decisive, as we shall see, since ‘contradictions’ are allocated an entire chapter in Archaeology, presenting arguments that clearly link it to another important notion—that of ‘problems’.
Comparisons, Correspondences, and Contradictions
Foucault stresses in Archaeology that discourse analysis ‘is not an interpretative discipline’ and does not seek to define the ‘thoughts, representations, images, themes, preoccupations that are concealed or revealed in discourses’ (Foucault, 1972a: 138-9). Instead, archaeology focuses on ‘those discourses themselves … as practices obeying certain rules’ (Foucault, 1972a: 138). Like Dumézil’s ‘systems’ and ‘structures’, ‘rules’ are not the unifying determinants of consensus, but a historically specific ‘system of differentiation’ to be mapped (Foucault, 1972a: 50). In other words, Foucault’s archaeology is developed to account for how co-occurring but different and even contradictory enunciations are simultaneously possible, that is, ‘to formulate their law of division’ (Foucault, 1972a: 32-3). Thus, archaeology searches for the system and its rules that would account for the ‘conditions of existence (but also of coexistence, maintenance, modification, and disappearance)’ of enunciations (Foucault, 1972a: 38).
In Archaeology, the expression ‘conditions of existence’ replaces that of ‘conditions of possibility’ in his earlier works. While most commentators hold on to the latter, ‘conditions of existence’ is in fact a more appropriate term, since it focuses on the ‘law of coexistence’ of enunciations at the level of discourse itself (Foucault, 1972a: 116). As noted by an editor, Foucault implicitly contrasts this to the speech act theory of Chomsky and Austin with its ‘finite body of rules that authorizes an infinite number of performances’, whereby the speaking subject is a priori attributed agency rooted outside discourse itself (Foucault, 1972a: 27, 107, see notes in Foucault, 2015: 1436-41). Insisting on ‘conditions of existence’ implies treating research materials not as ‘documents’ of agency—or of historical circumstances and events—outside discourse. It requires constructing an archaeological account of the coexistence of given discursive ‘monuments’. The danger of a priori distinguishing a realm of possibility from that of discourse is to subsume the diversity of discourses under the non-contradiction of history or subjectivity (agency)—of ideology, culture, Zeitgeist, etc.—possibly hidden only by muddy or conflictual practice (Krarup, 2019a). Instead, archaeology seeks ‘to divide up’ the diversity of discourses ‘into different figures’ (Foucault, 1972a: 159-60). Thus, ‘for archaeological analysis, contradictions are neither appearances to be overcome, nor secret principles to be uncovered … [but] objects to be described for themselves’ (Foucault, 1972a: 151). This requires special attention to every manifestation of contradiction as potentially the manifestation of a rule-bound system of differentiation. We shall examine examples of this in Foucault’s works later, but first consider how contradictions relate to problems (or problematics, problematizations) as adopted by Foucault from the French epistemological school.
The French Epistemological School: Problems, Problematics, Problematizations
The junction of structure and transformation was not Foucault or Dumézil’s innovation. It was widespread in the history of systems of thought, as evidenced in such notions as ‘evolution’ (Guéroult, Canguilhem), ‘genesis’ (Hyppolite), ‘ruptures’ (Bachelard, Althusser) and ‘genealogy’ (Foucault). While also making occasional use of the term ‘archaeology’ (Kusch, 1991: 6), the French epistemological school developed a post-Hegelian notion of ‘problems’ (or ‘problematics’, ‘problematizations’) in order to theorize this junction, complementing and elaborating that of ‘contradictions’. The notion of ‘problematic’ appears to have come to France through Heidegger, but in France Heidegger’s universal ‘problematic of being’ was quickly converted into a post-Hegelian—always historically specific—dialectic of problems and responses (Vuillerod, 2020). Gaston Bachelard (1884-1962) was the first to employ the notion of ‘problematic’ in published works to characterize not a theory but its structure, and Louis Althusser (1918-1990) developed it further. Others talked of ‘problems’, such as Georges Canguilhem (1904-1995), the Annales historian Lucien Febvre (1985 [1937]), and the philosopher Pierre Aubenque (2013 [1962]). Foucault used both notions interchangeably, as we shall see, and increasingly complemented them with a third alternative, ‘problematization’.
Bachelard opposed the empiricism of logical positivism. He argued that the problem of knowing is not simply about ‘external obstacles … such as the complexity and transience of phenomena’ or ‘the weakness of the senses or of the human spirit’, but resides ‘in the very act of knowing’ in which ‘sluggishness and disturbances arise by some kind of functional necessity’ (Bachelard, 2002 [1938]: 24, translation modified). In Gutting’s (1989: 28-9) rephrasing, scientific observation ‘is based on a theoretical redescription of the object that characterises it in terms of very different categories (e.g., “chemically pure sample,” “orientation of surface molecules”) from those of untutored experience’. Thus, according to Bachelard (2002: 25), ‘everything is constructed’—not socially constructed in the sense of Berger and Luckmann, but insofar as knowledge always occurs in a ‘system of thought’ (Bachelard, 2002: 27). The history of science conducts theoretical redescription of its own in order to identify the ‘epistemological obstacles’ inherent in a given system of thought, as well as to identify the ‘epistemological breaks’ at which the whole system of thought is fundamentally transformed.
Similarly, Canguilhem, Bachelard’s successor at the Sorbonne and mentor to Foucault, sought to write the history of medical science not in terms of progress and failures but of problems—including now-forgotten problems—and of the different responses given to them, especially in the form of theories (Canguilhem, 1990). According to Canguilhem, the emergence of new theories may hide ‘the persistence of the problem at the core of the solution that we thought to have given it’ (Macherey, 1964: 66). However, Canguilhem criticized the history of ‘predecessors’ identified, for example, by the sole occurrence of the same words in widely different epistemological contexts (a critique implicitly adopted by Foucault [2015: 1446]). Even mathematics, as argued by Jean Cavaillès (1903-44), does not simply ‘discover’ logical truths but constructs its objects in fundamentally unpredictable ways (Cavaillès, 2019 [1947]; see also Webb, 2012: 50).
However, it was Althusser, a student of Bachelard and a teacher of Foucault, who developed the notion of problematic in full (Vuillerod, 2020). In Reading Capital, published shortly before Order and Archaeology, Althusser (1970 [1965]) famously identified an epistemological break between the works of the young and the old Marx. Less widely appreciated is Althusser’s distinct ‘philosophical’ mode of reading. Althusser argues that Marx reads the works of classical political economy (Smith, Ricardo, etc.) as a philosopher when he identifies an often-occurring shift in political economy (sometimes within a single sentence) from (a) a question posed about the value of the products of ‘labour’ as a human activity to (b) an answer given about the price paid for ‘labour power’ as a commodity in the labour market. The incapacity of classical political economy to distinguish between labour and labour power ‘led it into inextricable contradictions’ (Marx quoted in Althusser, 1970: 21). Such contradictions emerge from what Althusser calls the ‘problematic’ of classical political economy as a system of thought: ‘the absolute determination of the forms in which all problems must be posed, at any given moment in the science’ (Althusser, 1970: 25). ‘Philosophy’ is no longer metaphysics laying the ground for valid empirical knowledge, but the ‘reading’ for and theoretical redescription of problematics in the history of (scientific) systems of thought. This work focuses on contradictions as manifest responses and hence as access points to the problematics structuring a given discourse. Although a demanding work of textual analysis, the result is not dull exegesis but a display of ‘the action of a real drama, in which old concepts desperately play the part of something absent which is nameless [i.e. only latently present in the text], in order to call it onto the stage in person’ (Althusser, 1970: 29).
Foucault’s Problems
While the notions of ‘problems’, ‘problematics’, and ‘problematizations’ are not afforded systematic discussion in Archaeology, the book mentions their centrality to existing archaeological analyses, for example, that the French title of Order, ‘Words and Things’, was ‘the entirely serious title of a problem’ (Foucault, 1972a: 49, see also 61). We may trace the use of these notions across the entire work of Foucault and see how he provides the same definition as for ‘contradictions’ in Archaeology, making them serve a corresponding but more elaborate methodological function.
Anthropology located a rupture between two fundamentally different ‘problems’ or ‘problematics’ in Kant. Correspondingly, Madness describes the birth of modern psychology in terms of the formation of a new ‘problematic’ about the moral truth of the human being (‘man’) (Foucault, 2006: 325, 528). Order describes the transition to the classical age as ‘the appearance of a new problem, unknown until then’ (Foucault, 2005: 47). As in Althusser, a problem or problematic is a fundamental question that discourse cannot cease to concern itself with and must constantly return to, continuously producing novel responses without solving it. For example, in the modern age, such a dynamically organized discursive production results in a manifest ‘system of contradictions’ and ‘theoretical conflicts’ around the new tripartite structure of anthropological knowledge (‘man, his madness and his truth’) that substituted the binary one of the classical age (‘truth and error’) (Foucault, 2006: 522). For each debate and domain of knowledge, ‘One must reconstitute the general system of thought … [that] renders an interplay of simultaneous and apparently contradictory opinions possible … [and] defines the conditions that make a controversy or problem possible’ (Foucault, 2005: 83, emphasis added). For example, Hume and Condillac present ‘solutions’ that are ‘strictly contradictory’ but which ‘respond to the same problem’ (Foucault, 2005: 78, translation modified). An episteme is not a coherent mentality, but a distributed and dynamic space of concerns, of problems and responses (Foucault, 1994: 581).
The problematic itself is never explicit, yet it is there like a ‘path’ whose ‘exact geography was never clearly mapped out for itself by the thought of the nineteenth century, but it was constantly travelled’ (Foucault, 2006: 526). Thus, for the archaeology of knowledge, contradictions (and conflicts, tensions, uncertainties) form privileged sites of comparison—for establishing comparative facts about the structure of problems that organizes discourse. Fast forward to Foucault’s allegedly new term ‘problematization’ towards the end of his life with History of Sexuality:
To one single set of difficulties, several responses can be made. … [W]hat must be understood is what makes them simultaneously possible … the soil that can nourish them all in their diversity and sometimes in spite of their contradictions…. [T]he work of a history of thought would be to rediscover at the root of these diverse solutions the general form of problematization that has made them possible—even in their very opposition; or what has made possible the transformation of the difficulties and obstacles of a practice into a general problem. (Foucault, 1998b [1984]: 118, emphasis added)
Partitioning or Transformation
Thus, in 1961, 1969, and 1984, Foucault states the same argument about problems, responses, and contradictions. But what about Foucault’s so-called ‘genealogical’ middle period of the 1970s? Here, Foucault is often said to expand his focus from science to institutions and technologies of power. However, while the contrast between Order and Discipline and Punish (1975) may be striking in this regard, contrasting the latter with Madness, which treats asylums and other institutions of confinement, is much less convincing. Certainly, Foucault began to use the Nietzschean term ‘genealogy’ in the 1970s and for some years he spoke only rarely about ‘archaeology’. However, his call for a Nietzschean philosophy is explicit and elaborate already in his introduction to Anthropology and in Madness. Decisively, the methodology practiced under the label of genealogy is not distinguishable from the archaeological one.
According to Discipline, the ‘problem’ of Antiquity was ‘to render accessible to a multitude of men the inspection of a small number of objects’, whereas the modern age ‘poses the opposite problem: “To procure for a small number, or even for a single individual, the instantaneous view of a great multitude”’ (Foucault, 1977 [1975]: 216). Foucault’s analysis of panopticon is famous, but the archaeological structure of the argument is rarely noted. Panopticon is a ‘solution’ to a specific form of that modern ‘problem’, namely: ‘What intensificator of power will be able at the same time to be a multiplicator of production?’ (Foucault, 1977 [1975]: 208). More broadly, ‘a whole problematic then develops’, namely, that of institutional architecture (of prisons, schools, factories, hospitals, etc.) ‘that is no longer built simply to be seen … or to observe the external space … but to permit an internal, articulated and detailed control’ (Foucault, 1977 [1975]: 172).
Foucault’s analysis of ‘power’ in Discipline is in fact rather of a problem—not that of power as a universal concept but of disciplinary power as distinct to the modern age: how to order the multitude of free individuals in ways that meet three criteria: (a) lowest possible cost; (b) maximum reach and intensity; and (c) optimizing without harming productivity (Foucault, 1977 [1975]: 218). This problem was already contoured in Madness (Foucault, 2006: 437-9). While disciplinary power sought to standardize individuals, it was not uniform at the level of discourse. For example, the question of paid forced labour was heatedly debated throughout the first half of the 19th century, as payment seemed necessary to build moral subjects but contrary to the notion of punishment (Foucault, 1977 [1975]: 240-3). Thus, several and even contradictory responses were simultaneously made to the problem of disciplinary power. The problem is a latent rule organizing disciplinary discourse as a space of differentiation whose existence Foucault infers on the basis of countless structural comparisons between discourses about the army, prisons, schools, hospitals, and asylums.
Dreyfus and Rabinow’s (1983) influential partitioning of Foucault’s oeuvre must be abandoned, at least when it comes to methodology, and certain misunderstandings pertaining to it corrected. One of Rabinow’s students claims that ‘archaeology asks about what has existed in the past, and without concern for how that which existed came into being’ (Koopman, 2013: 40). Observing that Archaeology implicitly cites Sartre’s critique of Order as a ‘magic lantern’ that only knows ‘movement by a succession of immobilities’, Koopman erroneously comments that ‘Foucault seems unsatisfied with his own response in admitting: “But there is nothing we can do about it”’ (Koopman, 2013: 40-1). Instead, the cited passage is conditional (‘it seems’, ‘as if’) and presents Sartre’s critique in order to tackle it, closing with: ‘But the problem must be examined in greater detail’ (Foucault, 1972a: 167). In the rest of the chapter—entitled ‘Change and Transformation’ and immediately following those on ‘Contradictions’ and ‘Comparative Facts’—Foucault explains that archaeology, rather than abandoning temporality, replaces that based on the continuity of consciousness and language with one that is specific to each discursive formation (Foucault, 1972a: 167). Thus, enunciations that are distant in historical time may be close in discursive space. Similarly, Foucault’s (1998a [1972]) ‘return to history’, which according to Koopman marks the transition to the genealogical concern with continuity together with ‘Nietzsche, Genealogy, History’ (Foucault, 1984b [1971]), cites and defends Dumézil’s method extensively, although it does not mention the term ‘archaeology’.
With the notions of contradictions and problems adopted from the French epistemological school, Foucault affords an important elaboration of Dumézil’s archaeological methodology in relation to discursive change and transformation. Contradiction, he states, relates to transformation because it ‘functions throughout discourse, as the principle of its historicity’ (Foucault, 1972a: 151; see also Foucault, 2005: 83). Obviously, there is constant change at the level of speech in so far as new enunciations are produced, new theories developed, and new controversies form—but a novelty in speech is not necessarily a novelty in terms of discursive rules (Foucault, 1972a: 153). Transformation at the level of the discursive formation is more rare and can be observed only indirectly when enunciations ‘are governed by new rules of formation’ (Foucault, 1972a: 173). By contrast, what Sartre and many others do is to define a principle of transformation independently of the specific transformation itself (as inferred from the series of concrete structural comparisons)—usually in terms of the continuity of consciousness, actions, or causes (Foucault, 1972a: 8). When Foucault opposes this view, it is not due to a ‘structuralist’ counter-ontology, but the result of his many observations of the painstaking struggles and often long and hardly visible (let alone explicit and conscious) historical processes it has taken to abandon one problematic and replace it by another. However, while Foucault shows in Anthropology and Order that the anthropological philosophy of human action and consciousness revolves around problems it cannot itself resolve, this does not rule out the work of scrutinizing our own thought retrospectively and questioning its bases.
Post-Hegelian Enlightenment
Foucault (1984a) re-defines enlightenment from Kant’s liberation from ‘self-incurred tutelage’ through ‘sapere aude’ (dare to know) into a certain form of historical inquiry. Rather than revealing rationality in history or the human being, such inquiry studies ‘the present limits of the necessary, that is, what is not or no longer indispensable for the constitution of ourselves as autonomous subjects’ (Foucault, 1984a: 43). In contrast to Kant’s transcendental inquiry, it is ‘archaeological in its method’ (Foucault, 1984a: 46). This talk of freedom and enlightenment late in his life does not mark a new era in Foucault’s thought. The potential of transformation from within a discursive formation was already explicit in 1961. For example, Kant’s ambiguous role between classical and modern problematics is the result of the ‘insistence’ of his Anthropology that finally deploys a new problem, namely ‘the bond between truth and freedom’ (Foucault, 2008: 104). Similarly, the ambiguous status of Freud as simultaneously the termination of the modern and the invocation of a new (structural) period in Madness (e.g. Foucault, 2006: 339) and Order (Foucault, 2005: 393) concerns his pushing the old problematic to its limit. Indeed, the difficulty of analysing such transitional figures is that two different discourses (e.g. classical and modern) may occur side by side, even within single sentences (just as in the political economists, according to Althusser). The difficulty of pointing out precisely where the transition from one organizing problem structure (discursive formation) to another occurs is intrinsic to archaeological methodology, but it does not exclude change in the manifest responses, nor transformations of discourse itself.
It is possible that Kant paved the way for an archaeology of thought in conjunction with a novel notion of enlightenment. However, as everyone in the field of the history of systems of thought in France in the 1960s was well aware, it was Hegel (1770-1831) who formulated its basic principles and first sought to carry it through. With The Phenomenology of Spirit (1807), Hegel produced a work that was itself a formative history of thought (and not just a historical overview and critique of prior forms of thought), consisting of a long filiation of contradictions that have only been overcome (aufgehoben), one after another, upon fierce struggle. The importance of Hegel’s Phenomenology is massive, as it breaks with the very foundations of classical Western metaphysics, which takes its absolute beginning in the claim that contradictions can have no subject, meaning, and substance (they are nothing) (e.g. Baumgarten, 1757: §7). By contrast, Hegel not only affirms that contradictions can be substantial as problems, that is, as latent dynamic structures of concern in thought, but he places problems at the centre of an inherently dynamic philosophy that has been referred to as ‘problematology’ (Vuillerod, 2019; see also Hyppolite, 1997 [1953]).
Foucault may abandon Hegel’s notion of absolute knowledge (with its teleological connotations of progress) while adopting the role of contradictions, problems, and conflicts in his philosophy (Gutting, 2010: 24; Muldoon, 2014: 109). Already Guéroult (1979: 24-5) credits Hegel for being the first to observe that the object of philosophy cannot be defined a priori and that, instead of beginning with an ‘object’ (human reason), philosophy chooses as its starting point a ‘problem’. Following his translation of Hegel’s Phenomenology in 1939, Hyppolite (1974 [1947]) published his influential companion entitled Genesis and Structure of Hegel’s Phenomenology of Spirit. Hyppolite broke with the humanist reading of Hegel dominant in France at the time by leaving ‘no transcendence outside historical becoming’ (Hyppolite quoted in Canguilhem, 1948: 293). Foucault paid tribute to his old teacher by entitling his minor thesis Genesis and Structure of Kant’s Anthropology (Eribon, 2011: 187). In his thesis report on Madness, Canguilhem asserted that ‘Foucault maintained from beginning to end a dialectical vigour that comes in part from his sympathy with the Hegelian vision of history and from his familiarity with the Phenomenology [of Spirit]’ (Canguilhem, 1998: 26). Indeed, to the extent that Foucault can be said to have sought to overcome Hegel, he made clear (in a tribute to Hyppolite) that it is necessary to do so via Hegel in order ‘to determine the extent to which our anti-Hegelianism is possibly one of his tricks directed against us’ (Foucault, 1972b: 235).
Concluding Remarks: Towards Problem Analysis
I have presented a methodological reading of Foucault’s archaeology by situating it in the broader tradition of the history of systems of thought in France around 1960. Moreover, I have accounted for both the necessity and elaboration of that methodology following Foucault’s broadened focus on ‘knowledge’ (extending far beyond philosophy and science). The immediate source of inspiration for the methodological conception of archaeology is not Kant, as often assumed, but Foucault’s mentor Dumézil. Foucault made a considerable contribution by forging archaeological methodology with the post-Hegelian notion of problems as the conditions of existence and organizing rules of discourse as a distributive (rather than coherent) system. Specifically, through structural comparisons of manifest contradictions—and, more broadly, oppositions, controversies, struggles, tensions and so on—between different responses, the archaeology of knowledge seeks to locate and specify their conditions of existence: the latent but organizing problems in discourse, functioning as ‘systems of differentiation’ or rules. These points are central for an adequate appreciation of Foucault’s work and deployment of his approach.
Archaeological methodology may inspire social researchers to initiate new types of investigations of philosophy, science and, in particular, knowledge in a broad sense. Social research deploying archaeological methodology may be referred to as ‘problem analysis’ (Krarup, 2019a, 2019b). While problem analysis may seek to account for the descent and evolutionary morphology of a problem, it can also be deployed in relation to contemporary societies. In so doing, social research would have to avoid pre-defined limitations of the object under study because the determination of limits would result precisely from its endeavours—just as, for example, the specification of ‘Indo-European languages’ marks the result rather than the outset of ‘archaeological’ inquiry in comparative linguistics. However, while it may eventually end in analyses of the fundamental epistemic problems in long historical epochs, problem analysis (and its potential for enlightenment) begins more moderately. Whenever we realize that we had not only not got the solution right, but also the problem we were trying to solve, the condition of existence has also been formed for new and—perhaps—more appropriate responses that may eventually help overcoming the problem.