Explore · Reader Guide
Critical Rationalism
This book does not begin by explaining or defending Critical Rationalism. It begins closer to the point where a reader familiar with Popper may already be standing.
You may already accept that knowledge has no authoritative source; that observation, reason, tradition, expertise, and institutions can all contribute knowledge without becoming foundations that certify it; that objective truth is compatible with human fallibility; that criticism matters not pedigree; and that progress depends on finding and eliminating error rather than justifying beliefs as finally secure.
The book's question is what happens when those ideas are allowed to travel to an explanation of a person. What do they imply about you as a singular person: the place where inherited knowledge can become a problem, where a conjecture can appear, where criticism can change what is thinkable, and where new knowledge can be created?
Where the book begins
Start with 00 - Universality.
The chapter introduces several pieces of machinery used throughout the book: objective knowledge, biological knowledge, abstraction, compression, problems, tentative solutions, error elimination, successor problems, computation, recursion, ancestry, derivation, and explanatory creativity.
Some of this will be familiar. The important move is not the familiar schema itself:
The important move is to treat the epistemology personally, as well as universally. It must constrain explanations of people, institutions, technologies, and the knowledge embodied in them.
The first question to carry forward is therefore:
If knowledge grows without authoritative foundations, what must be true of the singular person in whom problems, conjectures, criticism, and new explanations occur?
The first movement: from populations back to you
Read 01 - Genetics, 02 - Populations, and 03 - Societies as a single movement. These chapters repeatedly separate two things that are easy to collapse:
- knowledge about differences, regularities, classifications, groups, or populations;
- a causal explanation of one particular person.
The point is not that population knowledge is unreal or useless. The point is that a representation can contain real knowledge without containing the causal history of the person represented by it.
For a reader already suspicious of epistemic authority, this is the first extension to watch carefully. A measurement, category, diagnosis, social role, or statistical relation can become authoritative in practice even when it does not explain the individual to whom it is applied.
01 - Genetics develops the distinction through personality genetics and heritability. 02 - Populations generalizes it into the problem of transporting population knowledge into an individual causal explanation. 03 - Societies then asks what happens when socially inherited knowledge, institutions, norms, and authority relations enter the individual's problem situation.
The successor question is:
If you cannot be identified with the classifications that describe you, what positive explanatory structure remains?
The second movement: the solitary epistemic person
Continue through 04 - Solitude, 05 - Openness, and 06 - Abstractions.
Here the book sharpens the individual standpoint. Objective knowledge can exist outside you, communication can connect you to other minds, and institutions can coordinate people, but nobody occupies a collective subjective standpoint from which society is simply known as a whole.
That does not make the individual isolated from society. The opposite follows. Language, theories, norms, technologies, institutions, and classifications can all become parts of your problem situation. They are encountered, reconstructed, retained, criticized, misunderstood, or replaced from a singular standpoint.
05 - Openness develops the consequence that your future epistemic organization is not exhausted by your present one. 06 - Abstractions then turns attention toward the representations through which reality becomes manageable, and toward what compression can conceal.
The question becomes:
If every representation leaves something out, what keeps different explanations answerable to the same reality rather than dissolving into private interpretations?
Consilience, criticism, and the creation of new problems
Read 07 - Consilience, 08 - Criticism, and 09 - Problems together.
These chapters take three familiar Critical Rationalist commitments and make them less comfortable.
First, if explanations travel, they collide with explanations developed elsewhere. 07 - Consilience asks what it means for different explanatory structures to constrain one another rather than merely coexist.
Second, criticism is not treated only as a procedure for rejecting error. 08 - Criticism asks what criticism can create: distinctions, decompressions, new problem situations, and successor problems that were not previously available.
Third, 09 - Problems turns the familiar epistemic cycle back on its first term. A theory of knowledge growth can begin with a problem, but is both circular and deeply mysterious. The book calls this the hard problem of problems. If genuinely new problems can arise, then the diagram describes a structure whose most important transitions still defy explanation.
Coordination, evil, and justification
10 - Coordination, 11 - Evil, and 12 - Justification move from individual error correction into social arrangements.
The Critical Rationalist reader may recognize the inherited suspicion of authority here, but the manuscript's concern is narrower than a general political doctrine. It asks what happens when systems must coordinate action while remaining capable of discovering errors in the explanations governing that coordination.
10 - Coordination develops the coordination problem. 11 - Evil asks how harmful objective knowledge can persist, reproduce, and become difficult to correct without requiring an evil essence or an omniscient evil agent. 12 - Justification returns directly to the epistemological temptation the book has inherited: even after justification has been rejected in principle, justificatory structures can reappear in confidence, consensus, evidence, expertise, probability, or other representations that begin to function as warrants.
The useful question is not merely:
Is this justified?
but:
What criticism could expose an error here, and what would allow that criticism to survive?
Inexplicit knowledge and taking explanations seriously
13 - Inexplicit and 14 - Seriously deepen the problem inside the person.
Not all knowledge is explicit. You can possess skills, expectations, sensitivities, habits, and criticisms that affect what you notice or do before you can formulate them propositionally. 13 - Inexplicit asks how such knowledge participates in epistemic change without becoming mysterious or immune to criticism.
14 - Seriously then returns to one of the book's governing rules: an explanation is not being taken seriously merely because you agree with it. To take it seriously is to allow its consequences to travel into other explanations and create new problems there.
For a reader already familiar with Critical Rationalism, this is a shift in emphasis. The manuscript is less interested in whether you can state the epistemology than in what happens when you let it reorganize the rest of what you think you know.
Personhood, mystery, and the arrows
The middle of the book turns directly toward what remains unexplained.
Read 15 - Personhood, 16 - Mystery, 17 - Arrows, and 18 - Unpredictability as a sequence.
The inherited epistemology describes relations among problems, conjectures, criticism, and successor problems. But a description of those relations is not yet an explanation of their physical or developmental implementation.
What makes this problem become salient rather than another? Why does this conjecture appear? Why does a criticism that once seemed irrelevant later become decisive? What turns an inexplicit mismatch into an explicit problem? What is actually compressed by the arrows in
17 - Arrows makes that compression itself a problem. 18 - Unpredictability then separates the ancestry of a new explanation from its derivation. Once an explanation exists, its history may be reconstructed. That does not mean the explanation could have been derived in advance from the knowledge that preceded it.
Values, resources, and error-correcting systems
19 - Values, 20 - Resources, and 21 - Error ask what changes when epistemic openness occurs under finite material and institutional conditions.
People have limited time, attention, memory, energy, money, and access to objective knowledge. Systems also allocate resources unevenly and can make some errors easier to discover than others.
The question is not whether scarcity, values, or coordination disappear under epistemology. It is how they alter the error-correction surface: which problems can be pursued, which criticisms can be expressed, which alternatives can be tested, and which mistakes can persist.
A Critical Rationalist reader may find it useful to treat these chapters as a move from fallibility in principle to the architecture of corrigibility in practice.
Interfaces, understanding, and AI
22 - Interfaces, 23 - Understanding, and 24 - AI follow objective knowledge into technological form.
A tool can embody knowledge without being the person who created it. Interfaces can change which knowledge is reachable and which operations are easy. Artificial systems can preserve, transform, route, recombine, and return objective knowledge to you.
The manuscript therefore does not begin with the question, "Is AI intelligent?" It first asks what kind of epistemic object the system is, what transformations it performs, what knowledge is embodied in it, and how its outputs enter another person's problem situation.
This matters because powerful performance is not, by itself, an explanation of explanatory creativity. 23 - Understanding and 24 - AI keep open the distinction between performing transformations over objective knowledge and explaining the creation of genuinely new explanatory knowledge.
The question to carry forward is:
What changes when objective knowledge becomes interactive enough to participate continuously in your problem-solving environment?
Taboo, invention, weights, and motion
25 - Taboo, 26 - Invention, 27 - Weights, and 28 - Motion investigate consequences of treating knowledge creation as open-ended rather than as the execution of a fixed inventory of possibilities.
These chapters examine domains in which inherited abstractions can become especially difficult to criticize, where novelty may be hidden by retrospective explanation, and where present measurements can acquire more causal or normative weight than their explanatory content warrants.
The continuity with the earlier epistemology is the recurring distinction between:
- what can be represented now;
- what can be criticized now;
- and what new explanatory structure may become possible later.
Authorities: turn the criticism back on the book
Read 29 - Authorities late, not early.
The chapter is most useful after the manuscript has spent many chapters constructing its own vocabulary. The relevant Critical Rationalist question is therefore recursive:
What would make the explanations in this book removable?
The issue is not whether authority, expertise, institutions, or inherited knowledge can ever be useful. The issue is whether any of them becomes a terminal epistemic court whose judgment substitutes for criticism.
That question applies to the manuscript too.
Inepistemic: the destination is not doctrine
Finish with 30 - Inepistemic.
The book does not end by asking you to possess an epistemological system. Its final movement is toward a concept of epistemic analogous to literacy or numeracy; perhaps putting epistemology in motion.
The desired change is therefore not from ignorance to authority, or from uncertainty to justified. It is from epistemology as something you can state to epistemology as something that can participate in how you live and build.
One warning about terminology
Readers familiar with Popper should not assume that every familiar term is being used in exactly the same way here.
In particular, pay attention to objective knowledge, objective truth, authority, and manifest truth. The manuscript often takes an inherited epistemological distinction and follows it into a different explanatory setting. When a familiar term begins doing new work, treat that change as a problem to inspect rather than silently importing its earlier meaning.
This is especially important because the method of the book is recursive, and that recursion will surface my errors and misunderstandings.