What a Museum About Repair Taught Me About Building

Part I: The Exhibition

I visited an exhibition called The Great Repair expecting to think about architecture.

The exhibition, curated by ARCH+, presented works by architects, artists, and researchers asking a simple question: what does it mean to repair something rather than replace it?

I expected before-and-after photographs. Restored facades. Case studies in conservation.

What I found was a disagreement that had been arranged in glass cases.

Some installations treated repair as restoration. Return the broken thing to what it was. Honour the original. Preserve the continuity between past and present.

Others treated repair as transformation. Acknowledge that what was lost cannot be recovered. Build something new that carries the memory of the break without pretending the break did not happen.

A third group asked a harder question. Who decides what counts as repair? Whose version of the original is being recovered? And what happens when the system being repaired was never designed to serve the people now responsible for fixing it?

That third question is the one I have not been able to leave alone.

Because it is not really a question about architecture.

It is a question about governance. About inheritance. About what you owe a system you did not choose and cannot simply abandon.

Part II: Two Instincts

The conflict I saw in the exhibition is old.

It runs through every serious attempt to fix something that has been broken at scale. Whether the broken thing is a building, a constitution, a health system, or a set of land rights, the same two instincts appear.

The first is restoration.

Restoration believes the break was a deviation from something that was working. The goal is to recover that prior state of coherence. This instinct is motivated by continuity. By the sense that what existed before the rupture had legitimacy worth preserving. That the way forward is the way back.

The second is reconstruction.

Reconstruction argues that the prior state was not worth recovering. That what looked like order was order built on someone else's exclusion. That honest repair cannot pretend to restore because restoration would mean reinstating the conditions that made the break inevitable.

These two positions are usually presented as opposites.

They are not.

In every real attempt at repair, both instincts operate simultaneously. The person doing the fixing is always choosing, in each specific decision, how much of the original to preserve and how much to transform. The choice is rarely absolute. It is a negotiation, made under pressure, about what the thing was actually for.

The conflict between these instincts is not a problem to eliminate.

It is the condition under which repair happens.

What matters is not resolving the conflict. What matters is being honest about which instinct is driving each specific decision, and whose interests that decision serves.

          THE BREAK
              |
  .-----------|-----------.
  |           |           |
  v           v           v
RESTORE   REBUILD    ACKNOWLEDGE
  |           |           |
Return to  Discard    Mark the break
prior      the prior  visibly. Build
state.     state.     forward from
  |           |       what happened.
  v           v           v
Hides      Destroys   Carries the
tension.   knowledge. tension
Tension    Vacuum     honestly.
returns.   fills.     Durable.

Part III: The Gold in the Crack

Japan has an old practice that addresses this directly.

Kintsugi, which means golden joinery, is the art of repairing broken ceramics using lacquer mixed with powdered gold, silver, or platinum. The cracks are not hidden. They are traced in gold. The break becomes the most visible feature of the repaired piece. The repair is the history. The history is the object.

A kintsugi bowl is not a restored bowl.

It is a more honest one.

The break is acknowledged. The repair is visible. The object carries, in gold lines running through it, the record of what happened to it. It does not pretend to be what it was before. It becomes something that could only have existed after the break.

This is not merely a beautiful idea.

It is a structural principle.

Most institutions that have survived rupture do the opposite of kintsugi. When they repair, they conceal. They adopt new procedures without acknowledging what the old procedures failed to prevent. They write new constitutions that replicate the assumptions of the ones they replace, because the drafters did not examine why the previous document failed.

The result is what you would get from a kintsugi artist who hid the gold.

A bowl that looks whole but carries, invisibly, every structural weakness of the original break.

The break is not the problem. The refusal to acknowledge it is.

The concealment is not neutral. It is a choice. And the choice has consequences. Because a break that is concealed is a break that cannot be examined. A break that cannot be examined is a break that cannot be understood. And a break that cannot be understood will eventually reopen, usually at the worst possible moment, in the most damaging possible way.

CONCEALED REPAIR           ACKNOWLEDGED REPAIR

Break occurs               Break occurs
     |                          |
Crack hidden               Crack traced
Glue applied               in gold
     |                          |
Looks whole.               Looks repaired.
Carries original           Carries record
weakness invisibly.        of break visibly.
     |                          |
Same stress applied        New stress applied
     |                          |
Reopens.                   Holds. Gold is
Worse than before.         stronger than
                           original material.

Part IV: Systems Built Against You

There is a version of this problem that is specific to Africa, and it is harder than the general case.

When a system breaks under normal conditions, the break and the cause are eventually separable. The flood recedes. The war ends. The misguided policy is reversed. What remains is a damaged system that still, in its design logic, was meant to serve the people inhabiting it.

Repair in that case is difficult. But the goal is at least clear.

African states inherited a different situation.

The institutions handed over at independence were not broken systems.

They were working systems.

They worked exactly as designed. They extracted resources efficiently. They enforced compliance cheaply. They administered territory in the interest of a government that was not accountable to the people living in it.

The police, the administrative structures, the borders, the legal frameworks, the tax systems: none of these were designed to serve African populations. They were designed to manage them. The distinction is not subtle.

And those systems were handed, intact, to governments that then had to build legitimacy through them.

Mahmood Mamdani's argument in Citizen and Subject is precise on this point. African independence transferred sovereignty without transforming the colonial state. The structure it transferred was institutionally bifurcated: citizens governed by statutory law, subjects ruled through chiefs exercising state-sanctioned customary authority. At independence, that bifurcation did not dissolve. It was inherited. In many cases it deepened. The result was postcolonial governments trying to build democratic legitimacy through institutions whose operating logic remained control, not accountability. The signs on the doors changed. The doors themselves did not.

This is concealed repair at institutional scale.

The restoration instinct says: recover what existed before colonialism. Return to pre-colonial governance forms. Find the legitimate prior order.

But that prior order no longer exists in the context that made it function. A governance form designed for one world cannot simply be transplanted into another.

The reconstruction instinct says: discard the colonial inheritance entirely. Build fresh institutions with no debt to what came before.

But the knowledge embedded in existing institutions, however badly designed, is real. Dismantling them creates a vacuum. Vacuums are filled not by better institutions but by whoever moves fastest.

Neither instinct is adequate.

This is precisely what makes African institutional repair the hardest version of the problem the museum was staging.

ORIGINAL DESIGN LOGIC
(Colonial Administration)
         |
         Built to:
         - Extract resources
         - Enforce compliance
         - Administer for a
           non-resident power
         |
         Handed intact at independence
         v
  INHERITED INSTITUTION
  +--------------------------------+
  | Procedures designed for control|
  | Borders designed for extraction|
  | Law designed for compliance    |
  +--------------------------------+
         |
    .----+----.
    |         |
    v         v
RESTORE    RECONSTRUCT
    |         |
Prior       Discard
context     knowledge.
gone.       Vacuum fills.
    |         |
    '----+----'
         v
    SYNTHESIS
    Work with existing dynamics.
    Apply pressure at leverage points.
    Change arrives later than planned.
    It holds.

Part V: The Artifacts

There is one place where this argument becomes impossible to ignore.

The debate over the return of African artifacts from European museums.

The Benin Bronzes. The Maqdala treasures taken from Ethiopia. The skulls of Namibian Herero and Nama people held in German institutions for over a century. The Egyptian antiquities distributed across European collections during the colonial period.

The official response from most European institutions has followed a consistent pattern.

The artifacts are better preserved here. The collections are shared with the world. Returning them would set a precedent. The legal frameworks governing ownership are complex.

These are not arguments about artifacts.

They are arguments about whose knowledge systems deserve institutional protection.

Because what is being held in those museums is not simply cultural property.

It is evidence.

It is the material record of civilisations that European colonial ideology classified as primitive, as having no history worth preserving, as requiring the intervention of more advanced peoples to be brought into the modern world.

The civilisations that produced the Benin Bronzes did so centuries before Europe had comparable craft. The institutional sophistication required to create them, to commission them, to use them as part of a functioning system of governance and ritual, was erased by the ideology that justified their removal.

Keeping them in European museums is not preservation.

It is a continuation of the original argument.

It says: we decide what is worth keeping. We decide what history means. We decide whose knowledge systems deserve the protection of institutions.

The request for return is not sentimental.

It is a structural argument about repair.

It says: if you are serious about fixing what was broken, start with what you can actually give back.

The restoration instinct, when it comes from the institutions that caused the rupture, is rarely about restoring what was taken. It is about preserving what was gained.

Part VI: Hegel's Patience

There is a philosophical framework for why none of this resolves quickly.

Hegel's idea of dialectic is often reduced to a tidy formula: thesis meets antithesis, and together they produce synthesis. This makes it sound like a negotiation, a splitting of differences, a designed compromise.

That is not what Hegel meant.

Synthesis, in Hegel's actual account, is not designed.

It is forced.

It is what becomes necessary when two irreconcilable positions have been held in tension long enough that neither can sustain itself, and the pressure of their conflict generates a structure that neither side anticipated or would have chosen.

The synthesis does not make both sides happy. It makes both sides obsolete.

The distinction matters because it changes what the builder's job actually is.

If synthesis can be designed, the builder's job is to design it. Find the compromise. Build the bridge. Get both sides to the table.

If synthesis emerges under pressure, the builder's job is different.

Hold the tension. Resist the temptation to resolve it prematurely. Maintain the conditions under which both positions can continue to press against each other until something new becomes necessary.

This is not passivity.

It is a specific kind of discipline.

Because the most common failure in institutional repair is not the failure to act.

It is premature resolution.

A government that forces a settlement before the underlying tensions have been genuinely worked through does not produce stability. It produces a frozen argument. The tensions go underground. The system looks repaired. And then, years or decades later, the original conflict re-emerges with all the force of the time it spent suppressed.

This pattern is visible across post-colonial African governance. The constitutions written at independence often imposed forms of reconciliation that the underlying social and institutional tensions did not support. The ethnic coalitions, the geographic federalisms, the power-sharing arrangements: many of these were solutions applied before the actual problem had been fully inhabited.

They looked like synthesis.

They were compression.

You cannot shortcut your way to a durable structure.

The structure has to be demanded by the situation.

And that takes longer than anyone involved would prefer.

Part VII: Attrition as Design

This leads to the most counterintuitive part of what the museum was actually arguing.

Attrition is not failure.

In the standard account of institutional reform, attrition is what happens when reform does not happen fast enough. The timeline slips. The opposition holds. The reformers lose energy or lose power. What was supposed to be a transition becomes an indefinite delay.

This reading treats complex systems as if they respond to reform the way a switch responds to a finger.

They do not.

Complex institutions do not transform under direct force. They erode, adapt, resist, and gradually reconfigure. The transformation is often invisible until it has already happened. The period that looks, from inside, like failure is often the period during which the system is absorbing enough change to make larger change possible.

The builder who understands attrition does not try to change everything at once. They identify the points in the system where consistent pressure produces cumulative effect. They understand that the system's own dynamics will eventually do more of the work than any single intervention.

In engineering, this is fatigue analysis. Apply load below the threshold of immediate failure, repeatedly, and the material eventually changes its properties.

In governance, it is the recognition that durable change usually happens when the old equilibrium runs out of the energy required to sustain itself, not when a reforming force defeats it in direct confrontation.

Consider what this looked like in practice. Botswana's Directorate on Corruption and Economic Crime, established in 1994, did not transform the country's governance overnight. It operated quietly, prosecuting cases, building precedent, training prosecutors, establishing norms of accountability within a civil service that had been built on very different assumptions. No single moment declared the transformation complete. But over two decades, the accumulated effect of consistent pressure at specific institutional points produced something that direct confrontation with the inherited system would likely have destroyed: a functioning accountability mechanism with genuine legitimacy. The change was invisible until it had already happened.

Institutions built for extraction do not transform through proclamation. Renaming a ministry does not change its operating logic. Rewriting a constitution does not change the training, incentives, and habits of the people administering it.

What changes institutions, over time, is the accumulation of different decisions at specific points. New training. Different incentive structures. Accountability mechanisms that work at the level of daily practice rather than annual audits. Leadership that models something different from what the institution was built to model.

None of this is fast.

None of it is visible until it has already happened.

But it is the only kind of change that does not produce new fragility immediately.

Attrition as design means accepting that the timeline of institutional repair is longer than any individual career, any electoral cycle, and often any single generation.

It means building anyway.

Part VIII: The Reflection

The artifacts debate, the governance failures, the constitutional arrangements that look stable and are not: all of them share a common structure.

Someone is insisting on a repair that conceals something.

The European museum that keeps the Benin Bronzes insists that preservation is a form of repair. The colonial administrative structures renamed as national ministries insist that renaming is a form of reform. The constitutions that imposed power-sharing on populations whose underlying tensions had never been addressed insisted that agreement on paper was a form of resolution.

In every case, what is being avoided is the gold.

The visible acknowledgment that the break occurred, that it was not accidental, that it had architects, and that genuine repair requires starting with that fact in plain sight.

Hegel would say this avoidance cannot continue indefinitely.

The pressure of what is unacknowledged accumulates.

The synthesis that everyone is designing around eventually becomes necessary.

Not because anyone chose it.

Because the tension finally demanded it.

The builder's work is not to wait for that moment.

It is to create the conditions in which, when the moment comes, something worth building is actually possible.

The metro essay in this archive ends with a question: whether the institutions we inhabit are running on engineered reliability or on hope. This essay adds a harder question underneath it. Before you can ask whether an institution is reliable, you have to ask what it was built to do. Because an institution running perfectly on hope is still running. The question is whose hope. And whose cost.

Questions I'm Still Thinking About

Books & References