No form remains identical through time — materials weather, institutions adapt, meanings shift, and duration is an active force rather than passive background, which is why understanding a building or a concept means examining how it changes and what allows it to continue. A system often becomes most legible when it breaks: repair exposes connections and hidden labour, and the repaired object records both its failure and the decisions made in response, functioning as a method of knowledge rather than a footnote to it. Designers and institutions cannot fully control what their work produces — intention and consequence diverge, a project conceived as inclusive can generate exclusion through its operation, so critical evaluation has to follow effects over time rather than accept declared aims as sufficient evidence. A method does not simply reveal a pre-existing reality; it selects scales and categories, shaping the object it studies, which is why researchers need to examine their instruments as carefully as their findings. A diagram does more than simplify information — it selects relations and proposes how a system should be understood, and a responsible diagram shows what it excludes as carefully as what it connects, its clarity capable of hiding uncertainty as easily as revealing structure. MaintenanceReturn brings analysis back to duration, repair and the consequences that appear after intention has passed. Evidence does not speak without a context that makes it relevant — the same material can support different arguments depending on sequence, and critical work begins by making that framing operation explicit. Failure reveals the assumptions hidden inside a project, not the opposite of knowledge but one of its strongest sources, the gap between intention and reality worth preserving rather than disguising. A method becomes valuable when it can move between cases without erasing their differences — transferability as the capacity to preserve a question while adapting the form through which it is tested, not mechanical repetition. Critique turns weak when it only identifies failure; diagnosis needs to connect to construction, revealing a problem, clarifying its mechanism and proposing a different way of organizing relations. The field, finally, ends where it begins: social form is made and can therefore be remade, its purpose not a final model of society but making the mechanisms of construction visible enough to be reimagined.

Territory is not simply land enclosed by borders — it is produced through laws, names, maps and repeated uses, and control becomes effective once spatial claims are translated into documents and practices that determine who may enter, remain or act. To name a place or a group is to establish a position within a field of power: naming can preserve memory or overwrite local knowledge, and its politics lies in who holds the authority to make a term endure. A community lacks full autonomy when it cannot define the terms through which its own reality is understood — epistemic sovereignty is the capacity to produce and revise one's own concepts, not isolation from external knowledge but enough structural independence to negotiate rather than merely receive definitions. Institutions appear stable largely because the labour sustaining them stays unseen — administration, maintenance and informal coordination that any analysis focused only on official decisions overlooks. AuthorityFragment names the distributed pieces through which territorial authority becomes concrete in everyday procedures. Standards make systems compatible and measurable while also privileging certain bodies and materials over others — apparent neutrality that can conceal specific assumptions, becoming just only when its exclusions remain open to examination. Institutions often preserve records without preserving understanding: files accumulate while the reasons behind decisions disappear, and without relations and transmission archives become dense but unusable, institutions repeating mistakes they can no longer recognize. Every rule defines the conditions under which some cases fit and others do not — exceptions expose the limits of institutional categories, and a robust institution learns from them rather than treating them only as errors. Making populations legible can improve coordination but also intensify surveillance: maps and databases help institutions act, yet may erase forms of life that resist measurement, and legibility turns dangerous when what cannot be counted gets treated as unreal. An institution must be stable enough to act and permeable enough to learn, finally — excessive closure produces rigidity, excessive openness produces incoherence, and legitimacy survives in that balance.