A direct answer
In philosophy, knowledge is often understood as a true belief whose truth is connected to the believer's grounds in an appropriate, non-accidental way. Truth and belief appear necessary for propositional knowledge, but philosophers disagree about what justification, reliability, competence, safety, or other condition turns true belief into knowledge.
Questions hidden inside
- What distinguishes knowledge from a belief that happens to be true?
- Must a person be able to explain their evidence in order to know?
- How much error or uncertainty can knowledge survive?
- When should testimony, instruments, institutions, or artificial systems be trusted as sources of knowledge?
Knowing is not one uniform relation
Philosophical discussions often focus on propositional knowledge: knowing that something is the case. Ordinary language also distinguishes knowing how to perform an activity and knowing a person, place, or experience through familiarity or acquaintance.
These forms can overlap without collapsing into one another. Someone may know every written instruction for riding a bicycle without yet knowing how to ride, or know many facts about a city without knowing the city through experience. A theory should therefore state which kind of knowledge it is trying to explain.
| Kind of knowing | Typical expression | Central issue |
|---|---|---|
| Propositional | Knowing that the library closes at six | Whether a belief is true and appropriately supported |
| Practical | Knowing how to repair a bicycle chain | Whether competence can be reduced to facts or instructions |
| Acquaintance | Knowing a place or recognizing a familiar voice | What direct experience contributes beyond description |
Truth, belief, and justification
A familiar proposal analyzes propositional knowledge as justified true belief. The truth condition excludes knowing what is false. The belief condition requires the person to accept the proposition rather than merely repeat it. Justification is meant to exclude guesses that reach the truth without adequate support.
Each condition raises further questions. Truth may exceed what a person can presently verify, belief can be partial or hesitant, and justification may depend on evidence, memory, perception, testimony, or reliable methods. The proposal is therefore a starting framework rather than a universally accepted definition.
Confidence is not the same as knowledge: a person can be confidently wrong, while genuine knowledge can coexist with appropriate caution.
A true belief can still be lucky
Gettier cases describe beliefs that are true, sincerely held, and apparently justified but true only because an unexpected fact compensates for something misleading in the evidence. Many readers judge that the person reaches the correct answer without knowing it.
The problem suggests that knowledge requires the truth of a belief to connect with the person's grounds in the right way. It does not produce one agreed extra condition. Instead, it motivates competing accounts of reliability, safety, intellectual competence, causal connection, and the absence of defeating information.
| Approach | Question it asks | Continuing difficulty |
|---|---|---|
| Reliability | Does the method generally produce true beliefs in relevant conditions? | Which methods and conditions are relevant? |
| Safety | Could the person easily have formed a false belief in a nearby situation? | How should nearby possibilities be selected? |
| Intellectual competence | Is the truth attributable to the person's epistemic abilities? | How much success must arise from the person rather than the environment? |
| No undefeated defeater | Is there overlooked information that would undermine the grounds? | Must an unknown defeater always prevent knowledge? |
Evidence can support without guaranteeing
Perception, memory, reasoning, testimony, records, and instruments can all provide grounds for belief. None is infallible. Good inquiry asks whether a source is functioning appropriately, whether independent evidence agrees, whether important alternatives have been considered, and whether new information defeats the original support.
The amount of checking reasonably required can depend on the stakes, the available time, the person's role, and the risk of error. Everyday knowledge would become impossible if every belief required a complete defense against every imaginable doubt, but careless confidence can also impose serious costs on others.
- Separate the claim from the confidence with which it is expressed.
- Ask what evidence connects the belief to the fact that makes it true.
- Check whether the source or method is reliable for this kind of claim.
- Look for relevant counterevidence, conflicts, and plausible defeaters.
- Increase verification when error would create greater harm or responsibility.
Knowledge need not mean absolute certainty
Skeptical scenarios ask whether ordinary evidence could be reproduced while the world is radically different from how it appears. Descartes's dream and deceiver arguments and the later Brain in a Vat place pressure on perception, inference, and claims about the external world.
Responses differ. Some seek foundations that survive methodic doubt. Others argue that knowledge can be fallible, that ordinary standards do not require eliminating unsupported remote possibilities, or that meaningful thought already depends on causal and social relations with a world. No single response has ended the debate.
Continue with Ada
Professor Ada Rowan
Test an idea or objection
Compare perspectives, test an assumption, or develop an objection using this reviewed page as context.
Related paths
Sources
- The Analysis of Knowledge (2026)
Current scholarly survey of attempts to analyze propositional knowledge, the justified-true-belief account, Gettier counterexamples, epistemic luck, no-false-lemma proposals, safety, sensitivity, reliabilism, virtue approaches, and knowledge-first alternatives.
View source - Is Justified True Belief Knowledge? — Edmund L. Gettier (1963)
Gettier’s short paper presents counterexamples to analyses on which justified true belief is sufficient for knowledge. It is the primary source for the modern Gettier problem and supports explaining the logical structure of the cases without treating one repair as settled.
View source - Content Externalism and Skepticism (2025)
Scholarly treatment of how the content of thought may depend partly on social or physical environments and how that claim interacts with skeptical scenarios, including Putnam’s brain-in-a-vat argument. It supports the thought experiment without confusing subjective indistinguishability with identical thought content.
View source - Reason, Truth and History — Hilary Putnam (1981)
Putnam develops the brain-in-a-vat case within a broader argument about reference, truth, realism, and the dependence of thought content on relations to a world. It is a primary source for a semantic response to skepticism, not a simple demonstration that perception is always reliable.
View source - Qualia: The Knowledge Argument (2024)
Detailed survey of the knowledge argument, Mary’s Room, formulations of physicalism, ability and acquaintance replies, phenomenal concepts, and questions about what new experience teaches. It supplies the objections needed to avoid treating the case as a simple proof.
View source
Last modified 2026-07-31. Reviewed by Thinking Paths editorial team.