> ## 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. # Critical Rationalism in Action: How My Manuscript Changed, and How I Changed With It - URL: https://otherwise.ghost.io/critical-rationalism-in-action-how/ - Published: 2026-05-16T14:51:13.000Z - Updated: 2026-10-17T13:39:42.000Z - Author: Khaled Salih - Tags: Newsletter, #Migrated-1792244264359, #Import 2026-10-17 14:38 16 May 2026 This is the second essay in my series on the ideas behind my book, When Parochial Error Goes Global. The first essay, “[No Error Is Final](https://khaledsalih.substack.com/p/no-error-is-final-a-deutschian-popperian),” introduced the Open Society, the Universal Explainer, and the moral discipline of error-correction. This second essay turns that principle back onto my own work. It tells the story of how the manuscript changed through criticism — and how I had to change with it. ——— I did not merely write a book about Critical Rationalism. This is the second essay in my series on the Open Society, the Universal Explainer, and the discipline of error-correction. The first essay argued that no error is final when criticism remains possible. This essay turns that principle back onto my own work. I had to let Critical Rationalism rewrite the book — and, in places, rewrite me. That is the most honest way I can describe what happened during the long transformation of my manuscript, *When Parochial Error Goes Global: Islam, Epistemological Closure, and the Defense of the Open Society*. ![](https://storage.ghost.io/c/d6/2c/d62cdd13-a476-41c5-8f2c-29270c99d95d/content/images/2026/10/3e34e043-ecbd-4ccb-8fc6-a00de7049d59_1054x1492-jpeg.jpg) The book began with force. It began with urgency. It began with a conviction that something fundamental had gone wrong in the way modern open societies understand *islām*, religious authority, ideological capture, and the suppression of criticism. I wanted to explain why certain systems of thought do not merely make claims about the world, but make criticism of those claims morally, socially, legally, or spiritually dangerous. That original impulse was not wrong. But it was incomplete. And because it was incomplete, it contained errors. The strange thing about writing a book on error-correction is that the book itself eventually becomes a test of whether the author believes his own argument. It is easy to praise criticism in theory. It is harder to let criticism enter the architecture of one’s own work. It is easy to say, with Popper, that our theories should die in our place. It is harder to let one’s favorite paragraphs die. It is easy to admire David Deutsch’s optimism about knowledge creation. It is harder to admit that one’s current explanation is not yet good enough. This manuscript forced that confrontation. The first draft had power, but also danger My original manuscript had energy. It had moral seriousness. It had many of the central ideas that remain in the final version: the Open Society, the suppression of error-correction, the danger of sacred immunity, the role of apostasy and blasphemy norms, the importance of criticizing doctrines without dehumanizing people. But the first draft also carried risks. At moments, the language risked sounding as if the problem were people rather than systems. At moments, my metaphors — software, code, program, machine, operating system — risked becoming too literal. At moments, the critique of a closed system of ideas could be misread as a critique of the human beings who inherit or inhabit that system. That was not my intention. But intention is not enough. A bad explanation does not become good because the author meant well. A dangerous ambiguity does not become safe because the writer knows what he intended. The reader does not live inside the author’s private mind. The reader encounters only the words on the page. And some of my words needed correction. The first major breakthrough was this: > The Muslim is the person. > > The doctrine is not the human being. > > The organization is not the individual. Once that distinction became explicit, the entire manuscript changed. It was not a minor ethical disclaimer. It became the governing rule of the book. # The person is not the memeplex The deepest correction was the distinction between the human being and the system of ideas. > A person is not a doctrine. A believer is not an institution. A Muslim is not *islām*. A captured mind is not a deleted mind. This may sound obvious, but it is precisely the distinction that contemporary debate often destroys. One side says: because Muslims must be protected from hatred, *islām* must be protected from criticism. The other side says: because *islām* must be criticized, Muslims may be treated with suspicion. Both positions are wrong. Both collapse the person into the doctrine. The Open Society requires a more difficult discipline: > Protect the person. Criticize the doctrine. Test the organization. Restore the Backspace Key. That became the moral spine of the manuscript. It allowed me to write with severity about doctrines, laws, sacred histories, institutional structures, social enforcement systems, apostasy norms, blasphemy pressures, gender codes, and political movements without reducing Muslims to those systems. The Muslim, in the final version of the book, is not the carrier of a defect. The Muslim is a Universal Explainer: a human being capable of conjecture, criticism, creativity, moral agency, and error-correction. That matters because the book’s purpose is not to condemn the person. It is to expose the cage. # Metaphors had to be disciplined The second major correction concerned metaphor. I use strong diagnostic metaphors in the book: software, code, operating system, factory reset, panopticon, cage, ledger, kill chain, memeplex, Backspace Key. These metaphors are useful because they make invisible systems visible. They help the reader see how doctrines can replicate, how institutions can preserve themselves, how social enforcement can operate without overt violence, and how inherited ideas can be protected from criticism. But metaphors are dangerous. A metaphor that begins as illumination can become distortion. If I describe an ideological system as software, the reader may begin to imagine human beings as machines. If I describe social conformity as programming, the reader may begin to imagine believers as hard drives. If I describe religious restorationism as a factory reset, the reader may begin to forget that people are not devices. That would be a moral and philosophical failure. So the manuscript had to build its own safeguards. The book now repeatedly insists that these are diagnostic metaphors, not descriptions of human essence. The software metaphor applies to the system of ideas, not to the human being. The human being is not software. The human being is the one who can debug, reject, rewrite, escape, and create. That correction was essential. Without it, the book would have risked becoming the very thing it criticizes: a system that flattens the person into the structure. # The great conceptual correction: parochial error The most important intellectual breakthrough came later. At first, the book risked sounding as if *islām*’s problem was unique in origin. But the more I worked through the manuscript, the clearer it became that this was too crude. *islām* is not the only tradition with parochial beginnings. All civilizations begin parochially. Judaism begins in a particular sacralized geography, with a particular people, covenant, law, memory, exile, temple, and text. Christianity emerges from Jewish apocalyptic expectation, Roman imperial violence, Greek metaphysical language, and the memory of Jesus interpreted through universal claims. African relational cultures arise from concrete kinship worlds, ancestral memory, land, community, obligation, and the deep grammar of personhood. Dharmic civilizations emerge from layered ritual, metaphysics, caste, renunciation, plurality, sacralized geography, and philosophical argument. Modernity itself did not descend from nowhere as pure reason. It emerged from Greek philosophy, Roman law, Jewish and Christian moral universalism, medieval scholasticism, Reformation conscience, Enlightenment criticism, scientific method, industrial power, liberal constitutionalism, and the long struggle to separate conscience from coercive authority. No civilization begins as pure universality. Every civilization begins somewhere. The decisive question is not whether a tradition begins locally. They all do. The decisive question is whether it develops mechanisms to correct its local beginning — or whether it sacralizes that beginning as final law. That is the concept I came to call parochial error. A parochial error is the mistake of treating a local regularity, inherited custom, tribal heuristic, historical contingency, or environment-specific rule of thumb as if it were an objective universal law. A parochial beginning is not itself the problem. The problem begins when a local answer says: > I am final. I am universal. I am sacred. I may not be corrected. I must govern all future minds. That insight changed the manuscript. It moved the book from polemic into Critical Rationalism. > The question was no longer: “Why is *islām* uniquely bad?” The question became deeper: > What happens when a historically bounded rule of thumb becomes immune to correction and then travels globally as law, conscience, identity, politics, and civilization? That is the real subject of the book. *islām* as case study, not biological destiny This correction also sharpened the book’s treatment of *islām*. Islam should not be reduced to Arabness or Middle Easternness. What later became *islām* emerged from a broader late-antique religious matrix: Jewish, Christian, Syriac, Aramaic, Byzantine, apocalyptic, imperial, scriptural, and monotheist. Arabic became the sacral interface. The *qur’an* crystallized in Arabic. But the mature memeplex quickly exceeded Arab ethnicity. It moved through Persian, Turkic, Kurdish, Berber, African, South Asian, Malay-Indonesian, European, and modern diasporic carriers. The stable object is not Arabness. The stable object is error-protection. That was another important refinement. It prevented the book from becoming an ethnic or regional critique. The object of analysis became clearer: not Arabs, not Middle Easterners, not Muslims as people, but a portable system of sacred authority, legal classification, prophetic finality, scriptural immunity, and resistance to root-level correction. That is why the book’s central case is *islām*, but its deeper warning is universal. > Every civilization can become closed. Every tradition can mistake its origin for reality itself. Every institution can protect its bad explanations. Every society can confuse loyalty with truth. Every open society can become afraid of its own critical tools. *islām* is the case study. Epistemological closure is the problem. # The manuscript became more humane by becoming more severe One of the paradoxes of the revision process is that the manuscript became both more severe and more humane. It became more severe because the target became clearer. I stopped diffusing the argument. I stopped trying to soften the analysis of systems that make criticism dangerous. I became more willing to say: this doctrine, this institutional structure, this sacred prohibition, this blasphemy norm, this apostasy rule, this gender code, this totalizing claim — these must be criticized. But the book also became more humane because the human being was separated more carefully from the system. That distinction gave me permission to criticize more directly without dehumanizing. The person must be protected precisely because the doctrine must be criticized. This is difficult for many people to hold in mind. But it is the only morally serious position. > Hatred says: the person is the cage. > > Surrender says: the cage must be protected for the sake of the person. The Open Society says: the person is not the cage; the cage must be criticized because the person can be free. That sentence became one of the deepest summaries of the book. # The structure had to stop expanding — and the tools had to be sequenced Another lesson was harder for me personally. At some point, a manuscript no longer needs more material. It needs discipline. During the process, I accumulated many possible additions: historical examples, AI audits, Substack reflections, contemporary cases, source debates, further comparisons, more quotations, more clarifications, more safeguards, more arguments. The same problem appeared in the book’s terminology. At one stage, I had gathered too much of the conceptual architecture into a long epistemological lexicon at the front of the manuscript. The intention was good: I wanted to give the reader the tools before the journey began. But the structure itself had become too heavy. It asked the reader to carry the whole conceptual system before the argument had generated the need for it. That too required correction. The solution came from the same tradition that had guided the book from the beginning. David Deutsch’s The Beginning of Infinity does not force every concept into one large front-loaded glossary. It lets each chapter create the problem, then introduces the relevant terminology where it becomes meaningful. I realized that my book had to do the same. So the manuscript changed again. The long front lexicon became a short Reader’s Compass: only the essential moral and methodological distinctions needed before departure. The detailed conceptual vocabulary moved into chapter-end Conceptual Debriefings, and the full lexicon was moved to the back as a reference appendix. This was not merely a formatting improvement. It was a philosophical correction. The old structure said, unintentionally: learn the system first, then read the book. The new structure says: encounter the problem first, then receive the tool that helps you see what happened. That is more faithful to Critical Rationalism. Concepts should not demand submission before inquiry begins. They should become useful instruments after a problem has made them necessary. Many of them were interesting. Some were even true. But not everything true belongs in the book. That was a difficult lesson. A manuscript can become weaker by becoming larger. A book can be buried under its own intelligence. The desire to protect every claim, answer every critic, include every example, and cover every possible objection can become its own form of intellectual anxiety. At a certain point, expansion is no longer improvement. It is fear disguised as diligence. Popper’s critique of utopian social engineering helped me understand this even at the level of the manuscript. A book about fallibilism must not try to become uncriticizable. It must not answer every possible objection before entering the world. It must remain finite, exposed, corrigible. Otherwise, the manuscript itself becomes a kind of intellectual total system: always expanding, always protecting itself, always postponing criticism until it has become too large to handle. So the final discipline became: > No more expansion. No more conceptual architecture. No more new chapters. No more new metaphors. No more inserting every interesting article. No more protecting the book from all possible criticism by making it unfinishable. The book had to become finite so that the conversation around it could begin. That too is Critical Rationalism in action. That is why serialization now feels right. It turns the book from a monument into a process of public error-correction: article, response, clarification, chapter, criticism, reflection, next chapter. Not final book, final authority, reader submission — but conjecture exposed to criticism. A conjecture must eventually enter the world. The book became a better example of its own method. It no longer merely argued for error-correction; in its structure, sequence, and pedagogy, it began to practice it. The manuscript now has a clearer architecture. It begins with moral and epistemological guardrails: person versus doctrine, Universal Explainer versus memeplex, criticism versus hatred, diagnostic metaphor versus dehumanizing metaphor. But it no longer asks the reader to master the whole conceptual system before the journey begins. The old heavy lexicon has been replaced by a short Reader’s Compass. The detailed terminology now appears where it belongs: after the chapters have made the concepts necessary, in short Conceptual Debriefings. The book then defines the internal architecture of the anti-rational memeplex. It traces how closure becomes origin, law, extraction, surveillance, and social control. It examines civilizational overwrites in Africa, Anatolia, and the Indus Valley. It turns toward the modern Open Society and diagnoses its category errors: suicidal empathy, ideological camouflage, abandonment of dissenters, and the confusion between protecting persons and protecting doctrines. Finally, it becomes constructive: intellectual hygiene, restoration of error-correction, and a manifesto for the Open Society. That structure was not present in the beginning. It emerged through correction. The manuscript became more Popperian because it allowed its own bad explanations to die. It became more Deutschian because it stopped front-loading vocabulary as authority and began letting problems generate the concepts needed to understand them. # What changed in me The manuscript changed, but so did I. I became more suspicious of my own force. I learned that intensity is not the same as accuracy. I learned that moral clarity without category discipline can become dangerous. I learned that severe criticism must be paired with equally severe protection of human dignity. I learned that metaphors must be watched carefully, because they can illuminate one moment and mislead the next. I learned that the Open Society is not defended by politeness alone, but also not by rage. I learned that compassion without criticism can become surrender, but criticism without compassion can become cruelty. I learned that the author of a book about error-correction must be willing to be corrected. Most importantly, I learned that the sentence “captured minds are not deleted” is not only a claim about Muslims inside *islām*. It is also a claim about myself, my manuscript, and every writer who discovers that his first explanation was not good enough. No error is final. But only if we let it be corrected. This did not stop when the manuscript became more stable. It continued into the articles around the book. After I drafted an essay on Sweden, Islam, Islamism, and intelligence without error-correction, a friend read it and gave me a criticism that was both generous and exact: the essay was too front-loaded and too generalizing. I had begun too high in the philosophy. I had allowed the conclusion — intelligence without error-correction — to arrive before the reader had fully encountered the cases. The criticism was right. So I rewrote the article in the same spirit as the book itself. I narrowed the claim. I made clear that by “Sweden” I did not mean the Swedish people, or every Swedish institution, or the whole Swedish intellectual tradition. I meant a recurring pattern in Swedish public handling of Islam, Islamism, Muslim representation, religious organizations, minority politics, and public funding. I moved the philosophical conclusion later. The article now begins with the concrete problem, moves through the three cases, and only then names the deeper pattern. That correction taught me something important: even a true concept can arrive too early. If the reader is asked to accept the abstraction before seeing the problem, the argument becomes less open. Critical Rationalism is not only about having corrigible ideas. It is also about sequencing ideas in a way that invites correction rather than demanding premature agreement. # Why this matters beyond the book This process matters because it demonstrates something larger. > If criticism can improve a manuscript, it can improve a mind. If it can improve a mind, it can improve an institution. If it can improve an institution, it can improve a society. If it can improve a society, it can improve a civilization. That is not naïve optimism. It is not the belief that progress happens automatically. It is the opposite. Progress happens only when errors can be found, named, tested, and removed without destroying the people who carried them. That is the great moral insight of Popper and Deutsch: > Let bad explanations die in our place. This applies to politics. It applies to religion. It applies to education. It applies to publishing. It applies to the Open Society. It applies to *islām*. It applies to my own book. The goal is not to be right from the beginning. The goal is to build systems — intellectual, moral, institutional, and civilizational — in which being wrong is not fatal. **Another thought** My manuscript began as a warning about *islām* and epistemological closure. It became something larger: a case study in the difference between a closed system and an open one. > A closed system protects its errors. An open system exposes them. A closed system treats correction as betrayal. An open system treats correction as loyalty to truth. A closed system makes doubt dangerous. An open system makes doubt useful. A closed system asks human beings to die for bad explanations. An open system lets bad explanations die in our place. That is what happened to my manuscript. And, in some measure, that is what happened to me.