Resemblance Is Not Structure
René Magritte, La Condition Humaine, 1933, oil on canvas, 100 × 81 cm, National Gallery of Art, Washington, DC.
In René Magritte’s La Condition Humaine (1933), a canvas sits on an easel before an open window, its painted landscape continuing the one visible behind it so seamlessly that the boundary between representation and world dissolves. The painting poses a problem. Is what we see on the easel a resemblance of the world beyond the window, or is it structured by the same principle that organizes that world, so that the correspondence runs deeper than any surface similarity could account for? Magritte offers no resolution. The two landscapes are superimposed without being identical, related without the nature of their relation being declared. It is a painting about the difference between analogy and homology: between a resemblance that illuminates and a structural correspondence that generates. That distinction, apparently technical but in fact fundamental, is what this essay is about.
Historians borrow concepts the way magpies collect objects: indiscriminately, attracted by shine. The discipline has no choice; it raids neighboring fields for instruments adequate to problems that resist homegrown solutions. The danger lies in borrowing without attending to what made the borrowed term precise. A concept evacuated of its technical meaning gives the impression of rigor while performing none of its work. Two such concepts, routinely flattened into near-synonymy, demand more careful handling than they typically receive: analogy and homology. What separates them is a matter of kind, and getting the distinction right changes what a historian can claim, what they must demonstrate, and how far their argument can travel beyond the case immediately before them.
Analogy is the historian’s most natural rhetorical instrument, and perhaps for that reason the most liable to overextension. To argue by analogy is to say: these two things resemble each other in ways that illuminate both. When we observe that the internal organization of a clandestine radical press shares features with the juridical brief it attacks, or that the procedures of a colonial census echo those of the metropolitan police apparatus it partly draws from, we are staging a resemblance and inviting the reader to find it generative. This is legitimate and often indispensable work. Analogy is how historical writing makes the unfamiliar accessible, how it builds connective tissue across cultural and temporal distance, how it opens questions that more systematic inquiry can then pursue.
Analogy also carries a structural limitation worth acknowledging honestly. It is silent on why the resemblance exists, whether it is necessary or contingent, whether it would survive changed conditions, whether it expresses a common generative logic or merely records a convergence of independent processes toward similar-looking outcomes. Analogies are content-dependent: they hold between specific features of specific phenomena at specific moments, and they require continuous renegotiation as those features change. The analogy that holds between the French prefectoral system and the Iranian SAVAK in 1970 may have dissolved entirely by 1979, and the analogical framework carries no leverage for explaining the transformation. The problem arises when the historian writes as though demonstrating a resemblance constitutes explaining a correspondence, as though the staging of similarity were itself an account of why the similarity exists. That conflation is precisely what the concept of homology exists to prevent.
Bourdieu and the structural claim
The word “homology” arrives in social theory from structural linguistics and anthropology, where it carried a precise technical meaning. In structural linguistics, two elements are homologous when they occupy equivalent relational positions within their respective systems, when the role each plays within its own structure corresponds to the role the other plays within its structure. What corresponds is the position and the network of relations it entails, apprehended at the level of structure rather than content. Lévi-Strauss extended this logic into anthropology, showing that myths and kinship systems across cultures could be homologous while sharing nothing on the surface, that what corresponded was the organizing grammar, the deep relational architecture, rather than any visible feature. Bourdieu inherits this usage and both refines and radicalizes it, making homology the hinge concept of field theory: the mechanism by which distinct social fields, each governed by its own specific logic and internal stakes, can be shown to be structurally correspondent while retaining their autonomy from each other and from any common external cause.
His most sustained deployment of the concept appears in Distinction (1979), where he argues for a homology between the structure of social space, class positions distributed according to the overall volume and specific composition of economic, cultural, and social capital, and the structure of cultural space, the distribution of tastes, aesthetic dispositions, and cultural practices across the social field. A preliminary clearing of the ground is necessary here, because the reading of this argument that first presents itself misses almost everything theoretically important about it. The intuitive version runs like this: educated bourgeois consumers prefer Schoenberg and Godard; working-class consumers prefer variety television and football; Bourdieu is mapping that correlation and calling it a homology. On this reading, the term names an empirical regularity, sociologically interesting but theoretically inert. A correlation between two variables is silent on whether their co-occurrence is structurally necessary or historically accidental, and it offers nothing when the specific contents shift, as they invariably do. Bourdieu’s homology claim operates at the level of the relational principle that organizes both social space and cultural space as structured fields, the level at which the generative logic of each becomes visible.
That principle works as follows. In social space, positions are defined relationally: by how much capital an agent commands relative to all other agents, what composition of capital they hold, what trajectory they represent within the overall structure. The same logic governs cultural space. Opera occupies its position because of its location within a relational structure in which cultural goods are “hierarchized” by the same logic that “hierarchizes” social positions, differential access to capital and the perpetual imperative to distinguish oneself from those positioned below while orienting upward toward those above. The homology claim is then this: the principle of differentiation organizing social space and the principle of differentiation organizing cultural space are the same principle. High/low in culture and dominant/dominated in class are both generated by the same underlying logic of distinction and differential capital accumulation, and the correspondence is structural all the way down, anchored at the level of the generative principle rather than the specific contents that populate each space at any given historical moment.
The stability of the homology across historical change follows from this. Photography migrated from low to high cultural status across the twentieth century. The specific goods attached to dominant-class taste shifted considerably. None of this disturbs the homology, because what anchors it is the structural equivalence of relational positions. The dominant fraction of the dominant class will be positioned in cultural space as it is positioned in social space, because the principle distributing positions in each space is the same principle. A historian invoking this argument owes the reader something correspondingly demanding: a demonstration that two fields are organized by the same generative principle, that the relational logic of each corresponds structurally to that of the other, and that this correspondence is produced by their mutual implication within a broader social field. That evidentiary standard is high. It is also what gives the argument its power: it generates predictions, constrains interpretation, and travels beyond the specific case in ways analogy cannot.
Foucault’s challenge
A Foucauldian would find several things to object to in everything above, and the objections sharpen rather than merely complicate the argument.
The phrase that would arrest Foucault first is “the same principle.” His entire methodological stance, worked out most carefully in The Archaeology of Knowledge and the preface to The Order of Things, is organized against this move: the positing of a single underlying logic that secretly generates apparently distinct surface phenomena. Foucault calls this “total history” and opposes it to “general history,” which refuses to assume that the elements of a given historical formation breathe together, that they express a common essence or are animated by a master code. The homological argument, at its strongest, performs exactly the operation Foucault proscribes: it finds beneath two domains a shared generative principle and treats that finding as explanation. Foucault’s response would be that the naming does no explanatory work. You have redescribed a correspondence as a principle, but the redescription accounts for nothing about how the correspondence came to exist or why it takes the form it does.
The deeper objection concerns power. Bourdieu’s framework operates within a fundamentally economic ontology: capital is the substance that flows and converts across fields, and positions in social space are positions of greater or lesser capital. Even when Bourdieu analyzes culture or symbolic goods, the underlying grammar is one of possession and differential distribution. Foucault’s methodological injunction to “cut off the king’s head” targets exactly this grammar, the model in which power is a quantity held by some and withheld from others, flowing downward from those who possess it to those who lack it. Modern power, Foucault argues, is productive; it calls into being the very subjects it subsequently appears to govern. A framework anchored in the distribution of capital forecloses this analysis at the level of its founding assumptions.
The deepest objection is about time. The homological analysis is synchronic: it maps a structural correspondence at a given moment and explains it by reference to a shared principle. Foucault’s genealogical method is an inquiry into how things that present themselves as structural necessities came to be structured in that way, through accidents, struggles, reversals, and sedimentations that left no necessary trace in the current configuration. A genealogy of the taste/class homology would trace how specific historical practices, the spatial organization of schooling, the medical regulation of bodily comportment, the development of museum culture as a technology of hierarchical classification, gradually produced both a field of cultural goods ordered hierarchically and a population of subjects for whom those goods were legible as markers of distinction. The homology is the explanandum: the historical achievement whose production needs to be accounted for.
One further objection, more specific but no less pointed, concerns the subject who inhabits Bourdieu’s social space. The habitus, that structured set of generative dispositions acquired through practice and operative below the level of explicit calculation, is Bourdieu’s attempt to escape both voluntarism and mechanical determinism, and within field theory it largely succeeds. But a subject with a habitus is still a subject who exists prior to the fields in which it acts; the habitus is something agents have, organizing their orientation within a social space they already, in some sense, occupy. Foucault’s question cuts at a different level: how the very form of “a subject with tastes who occupies a social position and can be classified accordingly” gets historically produced as a possible mode of being. That subject is a historical achievement, the product of specific power/knowledge regimes, of documentary practices, administrative categories, and classificatory systems that made certain forms of subjectivity thinkable. Bourdieu begins where Foucault wants to end.
What survives, and why it matters
The point of staging this confrontation is to clarify rather than adjudicate. Foucault’s objections are serious; they are also objections to what homological analysis alone can accomplish. The homological argument establishes the structural logic of a configuration at a given moment, demonstrating that two fields are organized by the same relational principle and can be analyzed as aspects of a single structured domain. The genealogical analysis asks how that configuration became historically possible, through what struggles and contingencies the current structure was produced, and what it systematically forecloses. These answer genuinely different questions, and the historically adequate account often requires both.
For historians working on archives, documentation, and knowledge-production specifically, Foucault’s concept of the dispositif, the heterogeneous ensemble of discourses, institutions, spatial arrangements, regulations, and administrative procedures through which power/knowledge operates, opens territory that homology leaves unmapped. It is useful precisely for cases where two documentary practices are mutually constitutive, where each practice produces the categories the other operates with, where the correspondence is generative in both directions simultaneously. When a state surveillance apparatus and a political movement’s self-documentation are described as “co-produced” within the same field of political contestation, the claim reaches beyond structural correspondence. The police dossier and the activist bulletin participate in constituting the political subject each claims merely to record; each makes the other’s categories operative; neither can be read in isolation without falsifying what it actually does. That argument requires discourse analysis, attention to the materiality of documentary practice, and a theory of subject-constitution, the full Foucauldian apparatus.
The two frameworks address different problems and produce different kinds of knowledge. Knowing which to deploy, and when to deploy both, is a matter of analytical precision rather than theoretical allegiance.
A note on the word itself
A practical caution to close. “Homologous” sometimes drifts in humanistic writing toward meaning something like “deeply analogous” or “structurally parallel,” similar in some important-seeming way, without the full theoretical weight that Bourdieu’s usage carries. The drift is understandable; the cost is real. If you intend the strong claim, shared generative principles and structural correspondence all the way down, it is worth signaling that you do, because readers trained in field theory will hear the strong claim by default and will hold you to its evidentiary demands. If you want to suggest a significant resemblance, “analogy” is the correct term for that argument, and there is no loss of precision or ambition in using it. The discipline does not suffer from insufficient theoretical vocabulary. It suffers from insufficient precision in the vocabulary it already has. The line between analogy and homology is one place where more careful habits would cost nothing and clarify a great deal.
Further reading
Pierre Bourdieu, Distinction: A Social Critique of the Judgement of Taste, trans. Richard Nice (Harvard University Press, 1984); The Logic of Practice, trans. Richard Nice (Stanford University Press, 1990). Michel Foucault, The Archaeology of Knowledge, trans. A. M. Sheridan Smith (Pantheon, 1972); The Order of Things: An Archaeology of the Human Sciences (Vintage, 1970); “Truth and Power,” in Power/Knowledge: Selected Interviews and Other Writings, 1972–1977, ed. Colin Gordon (Pantheon, 1980). On the Bourdieu–Foucault relationship, see David Swartz, Culture and Power: The Sociology of Pierre Bourdieu (University of Chicago Press, 1997), ch. 3; and Didier Eribon, “Foucault et Bourdieu: Les malentendus d’une rencontre,” in Réflexions sur la question gay (Fayard, 1999).
