> ## Content Index
> Fetch the complete content index at: https://otherwise.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# A Late-Night Discovery
- URL: https://otherwise.ghost.io/a-late-night-discovery/
- Published: 2026-09-16T05:14:12.000Z
- Updated: 2026-10-17T13:39:18.000Z
- Description: Five Books, One Conjecture, and the Open Future
- Author: Khaled Salih
- Tags: Newsletter, #Migrated-1792244264359, #Import 2026-10-17 14:38

16 September 2026

# A Note Before Reading

I did not set out to write this article.

Late one night, while working across five books that I had thought were about different problems, I placed them mentally beside one another and noticed a connection I had somehow missed while writing them.

I began with Kurdistan. From there, the questions took me through borders, states, institutions, doctrines, protected ideas, intelligence, and error-correction. Only afterwards did I recognize that the same question had been following me throughout: **what happens when something created by fallible human beings becomes protected from human correction?**

This essay is my attempt to understand that connection.

I offer it as a conjecture, not a conclusion. It may contain errors of its own. In fact, an essay about the necessity of correction should probably make that admission before asking anyone to read it.

What I think I discovered that night is simpler than the five books themselves: perhaps the most important thing we can preserve is not any particular answer, institution, border, doctrine, or political arrangement, but the possibility of correcting it—and of leaving the future open to better possibilities we cannot yet imagine.

That is where this essay begins.

***A Late-Night Discovery: Five Books, One Conjecture, and the Open Future***

![](https://storage.ghost.io/c/d6/2c/d62cdd13-a476-41c5-8f2c-29270c99d95d/content/images/2026/10/98cd1ac6-c99c-4639-a09a-1810b33f3fbf_1536x1024.png)

---

# A Late-Night Discovery

Five Books, One Conjecture, and the Open Future

Late one night, after working again on the five books, I placed them together and saw something I had somehow failed to see while writing them.

I thought I had written five books about different problems.

I had not.

They were five attempts to understand the same problem.

*Kurdistanî Azad* (co-authored with Ann-Catrin Emanuelsson) began with the struggle of a people to become the author of its political life. *Kurdistan: The Offense of Existing* asked why the continuing existence of that people could itself expose an error in the political map. *Kurdistan and the Open Future* asked whether the choices inherited from history exhaust the possibilities available to us. *When Parochial Error Goes Global* moved from political enclosure to epistemological enclosure. And *BACKSPACE* finally removed Kurdistan, Islam, and the Middle East as necessary starting points and turned the question back on all of us.

Only then did I see the conjecture underneath them.

The five books were about the open future.

And they had all been circling the same danger:

***What happens when something created by fallible human beings becomes protected from human correction?***

That question has followed me through Kurdistan, through states and borders, through doctrines and institutions, through religion and nationalism, and eventually back into the minds of intelligent people—including my own.

The answer I arrived at is not reassuring.

Human beings do not merely make mistakes.

We can become extraordinarily intelligent at protecting them.

## The Open-Future Conjecture

The conjecture I could suddenly see across the five books was this:

**Human beings become trapped when things we ourselves have created—ideas, identities, institutions, borders, doctrines, memories, interpretations, political arrangements—come to be protected as though they were beyond human correction.**

**Freedom begins when we recover the capacity to criticize and revise them without destroying the people who live among them.**

The second sentence matters as much as the first.

I am not arguing for a world without traditions, nations, religions, borders, inherited memories, institutions, loyalties, or commitments.

That would be another fantasy of perfection.

Human beings need explanations. We need institutions. We inherit languages, memories and ways of life long before we are capable of examining them. We become attached to places. We love people. We tell stories about where we came from. We build homes and political communities. We make commitments that cannot be renegotiated every morning.

The problem is not attachment.

The problem is not stability.

The problem begins when a human creation acquires **immunity from human correction**.

There is an enormous difference between treating something as settled enough to act upon and declaring it beyond criticism.

The deepest political and civilizational task therefore cannot be to eliminate error.

Karl Popper taught us why. We are fallible. Knowledge grows not because we finally discover an authority incapable of error, but because we conjecture, criticize, discover mistakes and replace weaker explanations with better ones.

David Deutsch pushes that insight further. Problems are inevitable, but problems are soluble. There is no reason to expect the growth of knowledge to reach some final point at which criticism is no longer necessary.

That changes the question.

The task is not:

***How do we build a society that never makes terrible mistakes?***

There is no such society.

The better question is:

***How do we prevent inevitable errors from becoming protected, replicated and increasingly difficult to correct?***

And beneath even that lies a larger question:

***How do we preserve the conditions under which better explanations, institutions, choices and futures—ones we cannot yet imagine—can still be created?***

Once I saw that question, the five books looked different to me.

## *Kurdistanî Azad*: Let It Be

*Kurdistanî Azad* begins with a people.

That sounds obvious. It was not obvious to the political systems that divided Kurdistan.

For much of modern Kurdish history, Kurds were not first asked how they wished to govern themselves. They were required to survive political arrangements made around them and often without them.

Names were changed.

Languages were restricted.

Histories were rewritten.

Borders hardened.

Revolts were crushed.

Villages disappeared.

People adapted because people do what human beings have always done: they try to live.

They marry. Raise children. Work. Remember. Hide memories when remembering becomes dangerous. Speak one language outside and another at home. Find ways around rules. Resist sometimes. Compromise sometimes. Leave. Return. Begin again.

This human world matters to me because political theory can easily turn people into categories.

**Kurds.**

**Arabs.**

**Turks.**

**Persians.**

**Muslims.**

**Jews.**

**States.**

**Minorities.**

**Majorities.**

But nobody wakes up in the morning as a geopolitical category.

People wake up as mothers, fathers, children, lovers, teachers, shopkeepers, farmers, students, believers, doubters, citizens, refugees.

Political arrangements eventually enter all those lives.

That is where the map becomes human.

*Kurdistanî Azad* follows a long movement:

***denial → survival → political agency → self-rule → correction → home***

The crucial turn is near the end.

Self-rule cannot be the final answer.

If Kurdish power merely replaces non-Kurdish power, and then declares Kurdish power immune from criticism, the deepest problem has not been solved.

A people needs more than the ability to govern itself.

It needs the ability to discover that it has governed badly.

It must be possible to criticize leaders without becoming an enemy of Kurdistan.

To **expose** corruption without becoming a traitor.

To **replace** policies.

To **reform** institutions.

To **question** official memory.

To **protect** journalists.

To **allow** courts to restrain power.

To **permit** women and men to challenge inherited authority.

To **lose** an election.

To **admit** failure.

And to **try again**.

That is the **Backspace Key**.

It does not erase history. Nothing can.

The dead do not return when an explanation changes. A war cannot be unstarted. A child who grew up under persecution cannot recover those years. Institutions created around an error do not vanish when somebody finally recognizes the mistake.

Backspace means something more modest and more demanding:

**the preserved capacity to make an explanation genuinely replaceable when criticism warrants its replacement.**

That is why we did not want *Kurdistanî Azad* to end with a perfect Kurdistan.

We do not believe in one.

We wanted it to arrive somewhere more human.

A home.

But a particular kind of home.

**A home with a Backspace Key.**

A free Kurdistan would not be a Kurdistan without Kurdish mistakes.

It would be a Kurdistan in which Kurdish mistakes do not have to become sacred.

## *The Offense of Existing*: Let Existence Question the Map

The second book pushed me into a stranger question.

***Why can the mere existence of a people become politically troublesome?***

Before a demand is made.

Before a border is challenged.

Before a weapon is lifted.

Sometimes existence itself contradicts an explanation.

Imagine a map saying that a political question has been settled.

The border is final.

The nation is one.

The territorial arrangement is natural.

The official history is complete.

Then people continue living in ways that expose what the explanation has left out.

Their existence becomes inconvenient.

This is how I began to understand the title *Kurdistan: The Offense of Existing*.

The offense is epistemological before it becomes political.

Political arrangements are not mountains.

Borders are not geological formations.

States are not laws of physics.

They are human answers to human problems.

**Some** are good answers.

**Some** solve difficult problems remarkably well.

**Some** are terrible.

**Some** survive for centuries.

But **survival is not truth**.

A political arrangement does not become true because armies defend it.

It does not become true because international organizations recognize it.

It does not become true because generations of schoolchildren have learned to color it on a map.

Survival proves survival.

Nothing more follows automatically.

That does not mean every border should be abolished.

It means something much less revolutionary and, in another sense, much more radical:

**Every human-made political arrangement remains criticizable.**

This is also where I became increasingly careful about the word *sacred*.

When I call a border, text, memory, place, interpretation or institution sacred, I do not mean that it possesses some objectively detectable property called sacredness.

I mean that human beings have **sacralized** it.

Nor is reverence itself the problem.

People may love a homeland.

They may cherish a text.

They may protect a memory.

They may feel profound attachment to a place.

The danger begins when reverence becomes immunity.

Then criticism is redescribed as desecration.

Disagreement becomes betrayal.

Historical inquiry becomes hostility.

The human creation begins masquerading as something beyond human judgment.

Kurdish existence therefore performs a peculiar epistemological function.

The map says:

*The question is settled.*

The people standing inside the map answer simply by continuing to exist:

*Then your explanation has left something out.*

## *Kurdistan and the Open Future*: Who Wrote the Menu?

The third book moved from maps to choices.

We inherit political menus and then spend extraordinary amounts of intellectual energy deciding which item to choose.

*Independence or integration.*

*Centralization or federalism.*

*State or non-state.*

*Rojava or Damascus.*

*Left or right.*

*Security or peace.*

*Resistance or surrender.*

The arguments can become so intense that the existence of the argument itself feels like freedom.

But there is a question that must come before choosing.

***Who wrote the menu?***

What assumptions generated these alternatives?

Who decided that these were the available choices?

Which possibilities disappeared before deliberation even began?

And the question I now consider most important:

***What better possibility does not yet exist because the knowledge required to create it has not yet been created?***

That is my **Choice-Space Audit**.

It challenges one of the most persistent errors in political thinking: mistaking the alternatives presently visible for the limits of possibility.

Human beings have repeatedly created choices that previous generations did not possess.

Federal arrangements had to be invented.

Constitutional courts had to be invented.

Representative institutions had to be invented.

New technologies create choices that did not previously exist. New knowledge changes what societies can coordinate, produce, cure, prevent and imagine.

Why should politics be exempt?

This is why I recently found myself asking a question about Syria that initially sounded almost too simple:

***What if Syria is the variable?***

For a century, Kurds have repeatedly been treated as the variable.

**Turkey** is given.

**Iran** is given.

**Iraq** is given.

**Syria** is given.

**Kurds** must **adjust**.

Kurds must **integrate**.

Kurds must **compromise**.

Kurds must **fit**.

***But why must the state always be the fixed object?***

States are human creations too.

They can learn.

They can decentralize.

Federalize.

Rewrite constitutions.

Redistribute authority.

Become something their founders did not imagine.

The moment the state itself becomes criticizable, the choice-space changes.

That led me to another proposition:

**Freedom does not consist only in choosing among alternatives history has supplied. It includes the capacity to create alternatives history did not supply.**

We are not merely consumers of history.

We can become authors of possibilities.

## *When Parochial Error Goes Global*: When Error Acquires Guards

The fourth book took me away from maps and into ideas.

***Why do some errors survive criticism so successfully?***

An error by itself is ordinary.

I make them.

You make them.

Scientists make them.

Governments make them.

Religions contain inherited interpretations that may be mistaken.

Constitutions contain errors.

Political movements make errors.

No society escapes this.

The danger begins when an error acquires **protection**.

An explanation becomes attached to an institution.

The institution creates careers.

Careers create incentives.

Identity attaches itself to the explanation.

Sacrifices are remembered through it.

Doubt becomes costly.

Exit becomes dangerous.

Questions that were once ordinary become suspicious.

The system gradually stops asking:

***Is this explanation still adequate?***

Instead it asks:

***How can this explanation survive what just happened?***

Something profound has changed.

The error has acquired more than believers.

It has acquired guards.

That is what I mean by **protected error**.

The danger of error is therefore not merely that it is wrong.

The greater danger appears when it develops mechanisms that interfere with our ability to discover that it is wrong.

And here another distinction becomes indispensable:

**Protect the person. Criticize the explanation.**

I have returned to that distinction again and again because human beings are so easily fused with the ideas associated with them.

**A Muslim** is not identical to an interpretation of *islām*.

**A Kurd** is not identical to a Kurdish party.

**An Israeli** is not identical to an Israeli government.

**An Iranian** is not identical to the Islamic Republic.

**An American** is not identical to American foreign policy.

**A scientist** is not identical to a theory.

**A person** is always more than the explanation currently occupying part of that person’s mind.

This matters morally.

It also matters epistemologically.

If abandoning an idea means abandoning myself, my family, my community, my ancestors, my dignity or my belonging, criticism becomes existentially expensive.

Then changing my mind can feel like destroying myself.

A decent society should make the opposite possible.

**Ideas should be able to die without people having to die with them.**

That may be one of the most important moral functions of an Open Society.

## *BACKSPACE*: Removing the Comforting Word “Them”

And then came *BACKSPACE*.

Only after writing it did I fully understand what it does to the previous four books.

It removes one extraordinarily comforting word:

**them.**

Protected error is not something only other people do.

It is not confined to dictatorships.

It is not confined to fundamentalism.

It is not confined to the Middle East.

It is not confined to societies we regard as backward.

Intelligent societies can become trapped.

Educated societies can become trapped.

Democracies can become trapped.

Universities can become trapped.

Corporations can become trapped.

Iran can.

Israel can.

America can.

Europe can.

Kurdish institutions can.

And I can.

This is where one of the most important lessons I have taken from David Deutsch becomes decisive:

**Intelligence is not rationality.**

Intelligence can discover errors.

But intelligence can also invent astonishingly sophisticated explanations for why an error must not be abandoned.

This explains something I have watched many times.

A policy fails.

The intelligent system does not necessarily reconsider the policy.

It produces a better explanation of why the failure does not count.

A strategy produces consequences its advocates did not expect.

The intelligent organization does not necessarily reconsider the strategy.

It develops a more sophisticated account of why implementation, circumstances, enemies, subordinates, timing or communication caused the problem.

Sometimes those explanations are correct.

That is precisely why the problem is difficult.

Reality does not tell us in advance at which level the error resides.

The employee may be **wrong**.

The general may be **wrong**.

The minister may be **wrong**.

The implementation may be **wrong**.

The strategy may be **wrong**.

The doctrine may be **wrong**.

The institution itself may be **wrong**.

A rational system must preserve the possibility that the error could reside at any relevant level.

The danger begins when one level becomes innocent **before inquiry begins**.

Then an **Error Ceiling** appears.

Below the ceiling, criticism can be furious.

Above it, silence.

## Where Does the Error Go?

This became one of the questions that now follows me everywhere:

***Where does the error go?***

When something fails, where is the failure allowed to travel?

Can bad news reach the top?

Can a junior employee tell a chief executive that the strategy itself is failing?

Can an officer question a doctrine rather than merely an operation?

Can a citizen question the policy rather than simply its implementation?

Can a believer question an interpretation?

Can a Kurd question a Kurdish leader without having Kurdish loyalty questioned in return?

Can an Israeli criticize an Israeli strategy without becoming morally identified with Israel’s enemies?

Can a Muslim criticize an Islamic doctrine without being treated as though the criticism were hatred of Muslims?

Can a European criticize an inherited consensus without immediately being expelled into one of the ideological camps already waiting to claim the criticism?

These are not identical cases.

They do not carry identical moral weight.

They do not have identical histories.

Methodological symmetry does not mean moral equivalence.

It means that **no person, doctrine, state, institution or cause—including those I support—receives immunity from criticism**.

That is the discipline.

The relevant question is not whether a society contains argument.

Many closed systems contain enormous amounts of argument.

The question is:

***Can criticism reach the explanation responsible for the error and make that explanation genuinely replaceable?***

Noise is not necessarily criticism.

Criticism is not necessarily correction.

## When Intelligence Protects Error

There is a darker possibility.

Imagine a system governed by a flawed explanation.

That explanation creates problems.

But the system is intelligent.

So it solves them.

A policy creates an unintended consequence.

Create another policy.

A bureaucracy becomes inefficient.

Create another agency.

A military doctrine generates a new vulnerability.

Develop a technological workaround.

A political arrangement produces resentment.

Build a security apparatus to manage the resentment.

Each intervention may be intelligent.

Each may solve a genuine problem.

And every successful downstream solution can reduce the pressure to reconsider the upstream explanation.

This leads to one of the conjectures I find most disturbing:

**Intelligence can postpone fundamental correction by making the consequences of protected error temporarily manageable.**

The smarter the system becomes, the longer the original error may survive.

Stupid errors may collapse quickly.

Intelligent errors can acquire lawyers, engineers, bureaucracies, universities, technologies, doctrines, budgets and moral vocabularies.

They can build worlds around themselves.

This is why intelligence without Backspace is not merely insufficient.

Under some conditions, it can make protected error more durable.

## Correction Debt

But the protection has a price.

An uncorrected error does not remain where it began.

It changes the world.

Institutions are created.

Budgets allocated.

Careers formed.

Borders fortified.

Children educated.

Enemies identified.

Alliances made.

Families displaced.

People sacrifice.

People die.

Memories harden around what happened.

The error becomes more difficult to correct because more human life now depends upon the world created around it.

This is **Correction Debt**.

Correction Debt is not merely the suffering already caused by an error.

It is the **future cost of correction created by previous failure to correct**.

A mistake that could once have been acknowledged with embarrassment may later require institutional reconstruction.

A policy that could once have been reversed may become entangled with livelihoods.

A border that could once have been negotiated may become surrounded by graves.

A grievance becomes a memory.

The memory becomes an identity.

The identity becomes a political institution.

The institution develops interests in preserving the explanation that produced it.

Time has not merely passed.

The choice-space has changed.

That is why early correction matters.

Not because early critics are necessarily right.

They are fallible too.

It matters because alternatives are usually cheaper while more doors remain open.

## The Foreclosed-Future Conjecture

And this is where the late-night discovery took me somewhere I had not quite reached before.

For years I had been thinking about protected error.

But sitting with the five books, another possibility became visible.

Protected error does something worse than preserve a bad answer.

**It can consume the future.**

That led me to a second conjecture nested inside the first:

**The Foreclosed-Future Conjecture**

**The greatest danger of protected error is not simply that it keeps a false idea alive. It progressively narrows the possibility of discovering and creating alternatives.**

Every uncorrected error changes the environment in which the next attempt at correction must occur.

It produces **institutions.**

**Interests.**

**Identities.**

**Laws.**

**Borders.**

**Dependencies**.

**Grievances**.

**Memories.**

**Fears.**

**Retaliations.**

And explanations for why all these things must now continue.

The next generation therefore does not inherit merely the original error.

It inherits the accumulated consequences of previous non-correction.

That changes what can be done.

Some doors become more expensive to open.

Some become dangerous.

Others disappear.

And people begin defending arrangements they never created because their livelihoods, identities, security, memories or sense of belonging have become attached to them.

This may be the deepest tragedy of protected error.

**Error accumulates not only consequences. It accumulates foreclosure.**

A border that cannot be questioned closes political possibilities.

A nationalism that cannot be questioned closes possibilities of belonging.

A sacred interpretation that cannot be questioned closes explanatory possibilities.

A leadership that cannot be criticized closes institutional learning.

A binary that cannot be escaped closes alternatives.

And a society that loses its Backspace Key can eventually force future generations to inhabit decisions they never made.

This is why correction matters so much.

Not because correction will finally make us right.

It will not.

We will make another mistake.

And then another.

Correction matters because it protects something larger than today’s answer.

**It protects tomorrow’s possibility of finding a better one.**

## Do Not Destroy the Means of Error-Correction

Underneath all five books I now find a principle Brett Hall formulated with extraordinary economy:

**Do not destroy the means of error-correction.**

I increasingly think of this as a preservation principle for civilization.

Free speech matters because error must be speakable.

Political opposition matters because rulers cannot reliably supply all the criticism their own explanations require.

Independent courts matter because power needs constraints it cannot simply switch off when those constraints become inconvenient.

Scientific openness matters because theories must remain replaceable.

Elections matter because governments must be removable without civil war.

Dissent matters because today’s dissenter may be carrying information that tomorrow’s majority desperately needs.

None of this means that every **criticism** is **good**.

Some criticism is **foolish**.

Some **malicious**.

Some **badly informed**.

Some criticism is itself an attempt to protect another error.

An Open Society is not one in which every critic is right.

It is one in which no authority has the final right to decide that criticism itself must end.

This also means that openness cannot require permanent hesitation.

Governments must govern.

Scientists must use theories.

Generals must choose strategies.

Parents must make decisions.

People must live.

Rationality does not mean refusing to commit until certainty arrives.

Certainty will not arrive.

The better principle is:

**Commit strongly enough to act. Remain open enough to correct.**

That is *Backspace*.

## Five Books, One Conjecture

I can now see a sequence that I never consciously designed.

From *Kurdistanî Azad*:

**LET IT BE.**

A people must be allowed to become the author—and the corrector—of its political life.

From *Kurdistan: The Offense of Existing*:

**LET ITS EXISTENCE QUESTION THE MAP.**

No inherited political arrangement becomes true merely because it has survived.

From *Kurdistan and the Open Future*:

**LET THE FUTURE REMAIN OPEN.**

The choices history has handed us are not the limits of possibility.

From *When Parochial Error Goes Global*:

**LET NO IDEA BECOME IMMUNE TO CORRECTION.**

Error becomes dangerous when it acquires protection against criticism and mechanisms for replication.

And from *BACKSPACE*:

**KEEP THE KEY THAT ALLOWS US TO CHANGE OUR MINDS.**

Every society needs institutions, habits and traditions that allow it to reconsider its own explanations before the cost of protecting them becomes catastrophic.

Only now do I understand the braid.

And perhaps I had misunderstood, while writing them, what these five books were ultimately about.

They are not ultimately about Kurdistan.

They are not ultimately about *islām*.

They are not ultimately about the Middle East.

They are not even ultimately about error.

Their common subject is something more human and more universal.

**The open future.**

## The Question I Am Left With

The question running through all five books is therefore larger than any of the histories that gave rise to it:

***How do fallible human beings build societies in which today’s explanations do not become tomorrow’s prisons?***

I do not have a final institutional design that answers that question.

I distrust the idea that there could be one.

Every institution we build will contain errors.

Every political settlement will create new problems.

Every theory we currently consider excellent may eventually encounter a better explanation.

Every generation will know things its parents did not know.

And our descendants will know things that we cannot know.

This is not a defect in human knowledge.

It is its promise.

Late that night, looking at the five books together, I realized that I no longer saw five conclusions.

I saw five attempts to preserve one possibility:

**that the future should not belong completely to the errors of the present.**

That is what the *Backspace Key* ultimately protects.

**Not perfection.**

**Not certainty.**

**Not a final political arrangement.**

The possibility that being wrong need not become permanent.

**That** an idea can die without the person dying with it.

**That** a government can be replaced without the country being destroyed.

**That** a Kurdish institution can be criticized without Kurdistan being betrayed.

**That** a religious interpretation can be challenged without the believer being hated.

**That** a state can reconsider its assumptions without treating reconsideration as surrender.

**That** a map can be questioned without turning the person questioning it into an enemy.

**That** today’s menu need not contain tomorrow’s possibilities.

And that future generations need not spend their lives protecting arrangements they never chose simply because we made correction too costly for them.

Perhaps that is also why the discovery had to come late.

I could not have begun with it.

I had to begin with Kurdistan.

**With people whose lives were divided by borders they had not drawn.**

**With survival.**

**With political agency.**

**With the right to make one’s own mistakes.**

Then I had to follow the problem from maps into institutions, from institutions into ideas, from ideas into intelligent societies, and finally back into myself.

Only then could I see what had been there all along.

The **purpose** of criticism is not to make us ashamed of being wrong.

The **purpose** is to keep wrongness from becoming destiny.

I do not know the final answer.

That is precisely the point.

**The future must remain open enough for someone to discover a better one.**