Social reality persists because relations acquire form. What appears stable—a city, an institution, a classification, a curriculum, an archive—is never simply given; it is the temporary result of heterogeneous forces aligned strongly enough to organize what may happen next. This alignment is the decisive condition. A relation becomes operative when it ceases to be a mere connection and begins to distribute possibility, establishing thresholds, sequences, permissions, exclusions and paths of recurrence through which later action must pass.
Form, in this sense, is neither object nor image. It is stabilized consequence. The street organizes movement before anyone describes its politics; the diagnosis reorganizes a life before treatment begins; the archive determines which past remains actionable; the standard decides compatibility by excluding what does not fit; the classroom arranges authority before a word is spoken. These structures gain force precisely because their construction recedes from view. Once sedimented into routine, infrastructure or professional language, they appear natural, technical or inevitable. They are none of these. They are relations that have hardened. This proposition alters the task of criticism. The objective is no longer simply to interpret visible forms or expose hidden ideologies, but to reconstruct the processes through which dispersed interactions became durable enough to govern subsequent reality. Such reconstruction requires a relational epistemology. Space cannot be separated from the practices and representations that produce it; objects cannot be isolated from the networks through which they acquire agency; ecological systems cannot be reduced to autonomous entities when every organism, institution and technology emerges through interdependence. Even discourse must be approached not as the expression of prior thought but as a field of rules determining which statements, objects and speaking positions can appear. Knowledge is therefore implicated in the reality it claims to describe. Categories do not stand outside the world. They sort it. Metrics do not merely record conditions. They create institutional visibility. Maps, diagrams and interfaces do not passively display a territory. They establish the relations through which that territory becomes manageable, investable or governable. Representation is already construction. Yet relationality must not become a vague celebration of connection. Networks are unequal; some relations support life, while others extract, immobilize or erase. The political question concerns the distribution of capacity within the arrangement. Who may alter the structure? Who remains bound by it? Who can move between scales, and who is confined to the consequences of decisions made elsewhere? A building support may permit future adaptation or permanently restrict it; an educational institution may generate autonomy or dependency; a classification may enable recognition while rendering adjacent experiences administratively impossible. Power is lodged not only in commands but in the architecture of available action. The most consequential decisions often occur before any public decision is announced, embedded in the protocols, formats and defaults that establish what can count as a legitimate option. This is why infrastructures demand inversion: the apparently secondary systems of naming, indexing, maintenance and coordination must be brought to the foreground. Their invisibility is not evidence of neutrality but a measure of their success. A functioning infrastructure disappears into use, while its exclusions remain distributed across bodies, territories and time. The same principle governs cultural and intellectual fields. Value is not contained within the work but produced through relations among institutions, markets, critics, archives and audiences. Citation creates proximity and distance. Publication allocates visibility. Repetition accumulates conceptual mass. A corpus becomes more than an aggregation of texts when its internal relations generate paths of return, comparison and transformation. At that threshold, knowledge behaves spatially. It develops centres, margins, supports and passages; it allows certain concepts to travel while others remain isolated; it can preserve difference or flatten it into searchable equivalence. Digital systems intensify this condition because machine legibility depends on explicit relations, but explicitness can either preserve complexity or convert it into administrative simplification. The challenge is not to resist structure. Structure is unavoidable. The challenge is to construct structures that remain accountable to their own effects. This requires a difficult balance between durability and revision. A field without stable supports cannot accumulate memory, but a field whose supports cannot be questioned turns provisional arrangements into law. The architecture of relations must therefore preserve its own history of formation. Every category should retain the trace of its exclusions; every standard should remain answerable to the bodies it fails; every archive should reveal the principles through which absence was produced; every institution should expose the labour that sustains its apparent coherence. Such transparency does not dissolve power, but it makes power contestable. The strongest consequence of relational thinking is thus not that everything is connected, a proposition too weak to matter, but that every durable form is made through specific connections whose arrangement can be studied, judged and altered. Social reality becomes transformable at the moment its supports cease to appear natural. Critique then acquires an architectural task: to identify load-bearing relations, distinguish those that enable collective life from those that merely reproduce dependency, and design conditions under which new forms may endure without concealing the violence of their stabilization. The field remains open, but not formless. Its future depends on making the mechanisms of durability visible enough that permanence can become a matter of collective choice rather than inherited constraint.
