Critical Fallibilism CF
A rational philosophy which explains how to evaluate ideas using decisive, critical arguments and accept only ideas with zero refutations (no known errors).
An open reference
CF Dictionary
Critical Fallibilism is a rational philosophy for evaluating ideas using decisive, critical arguments. This dictionary collects its terminology — including concepts it borrows from Critical Rationalism, Theory of Constraints, and Objectivism — and links them together so you can navigate the ideas as a network rather than a list.
The four traditions
Critical Rationalism CR
An epistemology by Karl Popper (refined by David Deutsch). Reality and objective truth exist; we are fallible; we learn by conjectures and refutations.
Theory of Constraints ToC
A methodology by Eli Goldratt about how to think, especially focusing on constraints (bottlenecks, limiting factors) and global optima.
Objectivism Oism
A philosophical system by Ayn Rand. CF draws especially on Oism's ideas about automatization, integration, and contextual knowledge.
Core CF Concepts 5
Binary Evaluation
CF's core method: evaluate IGCs in a digital, two-state way as refuted or non-refuted — never by degree.
Critical Fallibilism (CF)
A rational philosophy that explains how to evaluate ideas using decisive, critical arguments and accept only ideas with zero refutations.
Digital vs. Analog (epistemic)
Two kinds of evaluation: digital (discrete states) and analog (continuous). CF favours digital for [[error-correction|error correction]].
Partial Truth
Statements that contradict CF but are reasonable approximations. CF treats them as translatable, not false.
Yes or No Philosophy
CF's stance that epistemology is fundamentally binary — accept or reject, refute or don't refute.
Evaluating Ideas & Arguments 19
Argument
In CF, an argument is a chain of reasoning offered in favour of (positive) or against (negative) an IGC.
Context
In CF, a context is the situation in which an idea is being evaluated — the relevant facts, conditions, and other ideas.
Counter-Example
A single instance that contradicts a universal claim — the workhorse decisive criticism.
Criticism
A criticism is an argument that points out an error in an IGC. CF treats criticism as the core engine of progress.
Decisive Argument
An argument that, if accepted, makes its target IGC impossible to accept. CF's preferred kind of argument.
Decisive Criticism
A criticism that, if accepted, makes the targeted IGC impossible to accept — it refutes.
Error
In CF, an error is a reason an idea fails at a goal in a context — i.e., a refutation of an IGC.
Goal
In CF, a goal is what an idea is being evaluated against — what success or failure would mean for that idea.
Idea
In CF, an idea is anything that can succeed or fail at a goal — a hypothesis, a plan, a theory, a definition, an explanation.
IGC (Idea–Goal–Context)
The triple {Idea, Goal, Context} that CF uses as the unit of evaluation. An idea is refuted or non-refuted only for a specific IGC.
Indecisive Argument
An argument that doesn't decisively refute its target. In CF, indecisive arguments count for nothing — they don't get partial credit.
Negative Argument
An argument against an idea — a criticism. CF's preferred form of argument.
Non-Refuted
An IGC is non-refuted if no decisive criticism is known. CF treats this as the only positive evaluation status an idea can have.
Positive Argument
An argument in favour of an idea — providing a reason to accept it. CF rejects positive arguments as ineffective.
Refutation
In CF, a refutation is a decisive negative argument — a demonstration that an IGC fails.
Refuted
An IGC is refuted if there is a known decisive criticism of it — i.e., at least one error has been identified.
Translation of Arguments
CF's technique of rewriting positive arguments as negative ones (criticisms) to be compatible with CF principles.
Universal Premise
A premise of the form 'all X are Y'. Universal premises are the only reliable source of decisive positive arguments.
Universal-Premise Argument
An argument from a universal premise ('all X are Y') to a specific instance. The only reliable source of decisive positive arguments.
Error Correction & Learning 16
Bounded vs. Unbounded Progress
A CR distinction: progress is either bounded (limited to a known domain) or unbounded (open-ended across all of reality).
Brainstorming
A relaxed, divergent phase of idea generation, deliberately separated from critical review. CF uses it for conjectures.
Conjectures and Refutations
Popper's name for the evolutionary method: conjecture boldly, refute rigorously. CF adopts this as its engine.
Error Correction
The process of finding and fixing mistakes. CF treats error correction as the central mechanism of learning and progress.
Error Correction Budget
Your finite capacity to fix mistakes over a period. CF treats it as a real, budgetable resource like time or money.
Error Rate
The rate at which you create new errors. CF treats it as a manageable variable relative to error-correction capacity.
Evolution (epistemic)
The process by which knowledge grows through variation and selection. CF, with Popper, treats ideas as evolving.
Learning
The acquisition of non-refuted knowledge and skills. CF treats learning as evolution-driven and practice-dependent.
Meta Level
A higher level of reasoning about your reasoning. CF uses meta levels for stuck problems and to avoid infinite regress in debate.
Overreach
Doing work that's too hard for your error-correction rate. CF's term for chronic, self-defeating difficulty.
Paths Forward
A CF protocol for being open to criticism so that, if you're wrong, there's a clear path for your error to be corrected.
Piecemeal Improvement
CR's strategy: improve knowledge in small steps, one criticism at a time. CF keeps this as its operating mode.
Postmortem
A review of a failed project to find its errors and prevent their recurrence. CF treats postmortems as central to learning.
Problem Solving
The general process of finding and fixing errors in IGCs. CF frames all problem solving as error correction.
Progress
Improvement in ideas, methods, or knowledge over time. CF treats unbounded progress as achievable through error correction.
Technical Debt (analog)
The cost of accumulated, unfixed errors in software — and the CF analogy for life.
Knowledge & Epistemology 28
Authoritarianism (epistemic)
The epistemic view that some source — tradition, authority, or revelation — provides certain knowledge. CF rejects this.
Bayesianism
The view that rational belief is updated credences via Bayes' theorem. CF rejects it as analogous to credence and qualitative-error blindness.
Bias
A systematic distortion in how you evaluate ideas. CF treats bias as universal but partially correctable.
Certainty
The state of being sure an idea is correct. CF rejects certainty as unattainable and as a way to block criticism.
Confidence
In CF, rational confidence is grounded in having no known decisive criticisms and in objective reasoning — not in subjective strength.
Content-Based Judgment
Judging an idea by its arguments and content, not by the source's credentials, popularity, or status.
Credence
A numerical (0–1) degree of belief in a claim. CF rejects credences entirely.
Effective Thinking
Thinking that reliably finds and fixes errors. CF treats effective thinking as a learnable skill, not an innate talent.
Epistemology
The philosophy of knowledge — the study of how we evaluate ideas and learn. CF is an epistemology.
Explanation
An idea that says why something is the case. CF, with CR, treats explanations as central to scientific knowledge.
Fallibilism
The doctrine that humans are unavoidably capable of making mistakes — and that mistakes are common, not rare.
Fallible Knowledge
Knowledge that may be wrong but is good enough to act on. CF's substitute for justified knowledge.
Falsifiable
An idea is falsifiable if some possible observation would refute it. Popper's criterion of science.
Hypothesis
A testable guess. CF treats hypotheses as conjectures to be refuted, not confirmed.
Induction
Reasoning from particular observations to general claims. CF, following Popper, rejects induction.
Infallibilism
The view that some knowledge can be guaranteed true. CF rejects this in all forms, including probabilistic.
Justificationism
The view that knowledge requires justification by further claims. CF, with Popper, rejects this.
Knowledge
In CF, knowledge is non-refuted IGC — conjectural, contextual, and held open to criticism.
Objective Reasoning
Reasoning using facts, evidence, and logic that could convince any reasonable person — not opinions or intuitions.
Positivism
The view that only verifiable (or falsifiable) statements are meaningful. CF treats positivism as an error.
Rationality
The skill of finding and correcting mistakes. CF treats rationality as a learnable craft, not a fixed faculty.
Reason
In CF, reason is the evolutionary process of finding and correcting mistakes — a method, not a guarantee.
Regress Argument
The argument that any justification chain either regresses infinitely, loops, or stops at dogma. CF uses it to refute infallibilism.
Scientific Testing
A method of trying to refute a hypothesis with an experiment or observation. CF treats testing as the engine of science.
Skepticism
The view that we cannot have knowledge. CF rejects skepticism as incompatible with error-correction and action.
Truth
In CF, truth is approached but never guaranteed — error correction moves us closer without arriving.
Uncertainty
The state of not knowing whether an idea is right. CF handles uncertainty with binary IGC evaluations, not credences.
Unfalsifiable
An idea is unfalsifiable if no possible observation could refute it. CF treats such claims with suspicion.
Decision Making 11
Breakpoint
A point on a quantitative spectrum where an incremental change makes a qualitative difference. CF's key qualitative concept.
Category (of a factor)
A discrete group on a continuous spectrum, defined by breakpoints. CF uses categories to convert analog factors to digital ones.
Dimension
The unit of a factor. CF treats different-dimension factors as incommensurable — you can't directly add them.
Factor
A trait, characteristic, or sub-goal used to evaluate options. CF treats factors as a sub-goal of the overall goal.
Margin of Error
The imprecision around a measurement or breakpoint. CF uses it to handle borderline cases and to design buffer.
Most Factors Aren't Borderline
CF's empirical observation: most factors are obviously sufficient or obviously insufficient. Only borderline factors deserve attention.
Multi-Factor Decision
A decision based on multiple factors. CF shows the standard weighted-sum approach doesn't work and proposes pass/fail multiplication.
Pass / Fail
The binary evaluation of a factor at a breakpoint. CF's standard factor-level judgement.
Qualitative Difference
A difference of kind, not of degree. CF treats qualitative differences as the things that matter in evaluation.
Quantitative Difference
A difference of degree within the same dimension. CF treats most quantitative differences as unimportant at the breakpoint level.
Weighting
Assigning importance multipliers to factors. CF rejects arbitrary weighting and shows it's mathematically broken across dimensions.
Theory of Constraints Terms 15
Balanced Plant (anti-pattern)
Designing every step for 100% utilisation. ToC's classic anti-pattern — it makes the system slower, not faster.
Bottleneck
The slow step that limits the throughput of the whole system. The single constraint to focus on.
Buffer (ToC)
Extra capacity or inventory placed in front of a step (especially a bottleneck) to absorb variance.
Constraint
Anything that limits a system's throughput. ToC's general term for the bottleneck, often used interchangeably.
Excess Capacity
When a factor has more capacity than the system needs. ToC says don't optimise it — it's not the constraint.
Five Focusing Steps
ToC's iterative process for finding and managing the constraint: identify, exploit, subordinate, elevate, repeat.
Global Optimum
The best outcome for the system as a whole. ToC says aim here, not at local optima.
Inherent Simplicity
ToC's principle that every problem has a simple core solution, even when surrounded by apparent complexity.
Local Optimum
A narrow improvement that doesn't help the whole system. The trap ToC warns against.
Resolving Conflict (vs. Compromise)
ToC's preference for finding win/win solutions that resolve a conflict, rather than compromise between bad options.
Silver Bullet
A simple, high-leverage solution to a problem. ToC's term for a fix that resolves the core issue.
Theory of Constraints (ToC)
A methodology by Eliyahu Goldratt about how to think: focus on constraints (bottlenecks, limiting factors) for achieving goals.
Throughput
The rate at which a system generates its goal. ToC's measure of success.
Variance (ToC)
Statistical fluctuations in real-world performance. ToC designs for it; CF treats it as a normal feature of systems.
Win/Win Solution
Resolving a conflict by removing its underlying false assumption. ToC's preferred alternative to compromise.
Critical Rationalism Terms 12
Beginning of Infinity
David Deutsch's extension of CR: knowledge creation is the beginning of infinity — problems are inevitable, but all are soluble in principle.
Critical Rationalism (CR)
An epistemology by Karl Popper. Reality and objective truth exist; humans are fallible; knowledge grows by conjectures and refutations.
Falsificationism
Popper's epistemology: scientific theories must be falsifiable, and science progresses by attempting refutation.
Instrumentalism (CR critique)
The view that scientific theories are just useful instruments, not attempts to describe reality. CR rejects this.
Jump to Universality
The error of treating 'all observed X are Y' as 'all X are Y'. CR rejects this as the failure mode of induction.
Non-Empirical Criticism
Criticism that doesn't depend on observation — logical, mathematical, or conceptual. CR and CF both rely on it heavily.
Open Society
A society that allows criticism of its institutions and traditions. CR argues open societies are the only ones that progress.
Solipsism (CR critique)
The view that only one's own mind exists. CR shows that solipsism is self-refuting — you can't explain your own knowledge that way.
Static and Dynamic Memes
CR distinguishes memes that carry instructions (dynamic) from memes that just sit (static). Dynamic memes evolve.
Strong Empiricism (CR critique)
The view that all knowledge comes from sense experience. CR shows this fails to explain its own claims.
Three Worlds (Popper)
Popper's ontology: World 1 (physical), World 2 (mental), World 3 (abstract objects like ideas).
Tradition (CR)
The cumulative body of inherited, criticisable knowledge. CR treats tradition as essential to piecemeal improvement.
Objectivism Terms 6
Abstraction (Oism)
Oism's view of abstraction: abstractions are real, integrated conceptual units formed by omitting measurements.
Automatization
The process of making a skill automatic through practice, freeing conscious attention. Oism's key concept that CF adopts.
Contextual Knowledge
Knowledge that is correct in its context. A new context doesn't make a prior contextual evaluation incorrect for its original context.
Integration
Combining multiple simpler ideas into a single higher-level conceptual unit. Oism's core cognitive operation; CF builds on it.
Objectivism (Oism)
The philosophical system by Ayn Rand. CF draws especially on Oism's automatization, integration, and contextual knowledge.
Unit-Economy of Consciousness
Oism's claim that consciousness has a limited budget of attention units (roughly seven) — driving the need for integration and automatization.
Discourse & Debate 10
Appeal to Authority
Arguing that an idea is true because an authority says so. CF rejects this as an [[igc|IGC]] error.
Argument Strength and Intuitions
CF rejects the ordinary notion of 'strong' and 'weak' arguments. Either an argument refutes an IGC or it doesn't.
Criticism Policy
A personal or institutional rule for handling criticism. CF treats this as central to error-correction infrastructure.
Debate Policy
A personal rule for when and how to engage in debate. CF's idea of making one's openness to criticism explicit.
Debate Tree
An idea tree for a debate — possibly between two people, possibly about a whole field's open questions.
Discussion Forum
A public, moderated space for criticism and debate. CF treats forums as infrastructure for Paths Forward.
Discussion Tree
An idea tree applied to a specific conversation. CF uses it to track what's been said and what still needs answering.
Goal Tree
A tree diagram that decomposes a goal into sub-goals. CF / ToC uses it to clarify what success looks like.
Idea Tree
A tree diagram for organising ideas, arguments, and discussions. CF's central organisational tool.
Peer Review (CF critique)
The traditional scientific publishing gatekeeping system. CF critiques it as private, biased, and inferior to open criticism.
Practice & Mastery 6
Deliberate Practice
Practice aimed at improving specific weak areas, with feedback. CF adopts this as the standard for serious learning.
Habit
An automatized behavior. Habits can be useful (good [[practice|practice]]) or harmful (unexamined patterns).
Many Small Skills
CF's learning strategy: master many small sub-skills before tackling big ones. Avoids being stuck.
Mastery
Deep, [[automatization|automatized]] skill. CF treats mastery as the goal of long-term practice.
Practice
Repeated activity with the goal of improving. CF treats practice as the central mechanism for automatizing skills.
Similar Successes
CF's strategy for unbounded progress: do activities similar to past successes, generalising successful patterns.
Conscious & Subconscious 7
Automatized Knowledge
Knowledge that runs automatically — without conscious attention. Can be a source of bias, or a resistance to bias.
Conscious Ideas
Ideas currently held in conscious attention. Limited in number (roughly seven) and expensive to maintain.
Emotion (CF view)
Subconscious processes that signal and motivate. CF treats emotions as IGCs to be improved, not suppressed.
Intuition
A fast, subconscious judgment. CF treats intuitions as IGCs to be evaluated, not authorities.
Perfectionism (CF critique)
The drive to do things perfectly. CF treats perfectionism as a form of overreach that increases errors.
Procrastination (CF view)
Avoiding tasks by doing lower-value tasks instead. CF treats it as a symptom of overreach or unclear goals.
Subconscious
The mind's automatic processing layer. CF, with Oism, treats it as a powerful computer that runs automatized skills.