Epistemology
Epistemology is the branch of philosophy that asks what knowledge is, where it comes from, and how far it can reach. The word feels ancient, yet it was coined only in the 19th century to name this field as a distinct branch of philosophy. Its roots run deeper. It joins two ancient Greek terms, episteme, meaning knowledge or understanding, and logos, meaning study of or reason. Read literally, the word means the study of knowledge.
Consider a simple sentence: Ravi knows that kangaroos hop. Buried inside that claim is a tangle of questions philosophers have argued over for centuries. What does it take to truly know something rather than merely believe it? Can a belief that happens to be true still fall short of knowledge? And is knowing a fact actually better than just having a correct opinion about it?
This documentary follows those questions through their many homes. It moves through the pieces epistemologists pull apart, the rival schools that fight over the origins of knowing, and a history that stretches from Plato to thinkers in ancient India and China. Along the way it meets fake barns, a man who doubted everything, and a slate that starts out blank.
Knowledge, in its broadest sense, is an awareness, familiarity, understanding, or skill. Each form involves a cognitive success, a moment where a person makes epistemic contact with reality. Epistemologists usually treat it as an aspect of individuals, a mental state that helps a person interpret and interact with the world. The same word can also point to information stored in documents and computers, or to what a group of people share.
Not all knowing is the same. Propositional knowledge is knowledge of facts, the kind that fits into a that-clause, as in knowing that kangaroos hop. Because it is theoretical, it is usually held that only creatures with highly developed minds, such as humans, possess it. It contrasts with knowledge-how, a practical skill like knowing how to read or how to prepare lasagna. It also contrasts with knowledge by acquaintance, the immediate familiarity that comes from direct experience, like familiarity with the city of Perth, knowing the taste of tsampa, or knowing Marta Vieira da Silva personally.
Another divide runs between a posteriori and a priori knowledge. A posteriori knowledge rests on sensory experience, like seeing that the sun is shining or smelling that a piece of meat has gone bad. A priori knowledge depends on no such evidence and belongs to fields like mathematics and logic. A related contrast separates analytic truths, true by the meanings of their words, from synthetic truths, true because of additional facts. The sentence all bachelors are unmarried is analytically true, since bachelor already includes the meaning unmarried. Snow is white is synthetic, since its truth depends on the actual color of snow. Willard Van Orman Quine went further and rejected the distinction entirely, denying that there are any analytic truths.
For a long time the traditional analysis of knowledge held that knowledge has three components. It is a belief that is justified and true. In the second half of the 20th century, a series of thought experiments set out to show that some justified true beliefs are not knowledge at all.
Picture a traveler driving through a region full of fake barns, mere facades, unaware that they exist. By coincidence they stop in front of the only real barn and form a justified true belief that it is a real barn. Many epistemologists agree this is not knowledge, because the justification is not directly relevant to the truth. The success comes from epistemic luck, a cognitive win born of fortunate circumstances rather than competence.
The scramble to repair the definition produced many proposals. One view requires that the known fact cause the belief in the right way. Another says the belief must come from a reliable belief formation process. Others demand that the person would not hold the belief if it were false, that it is not inferred from a falsehood, or that the justification cannot be undermined. No consensus emerged. Timothy Williamson rejected the whole project, arguing that propositional knowledge is a unique state that cannot be dissected into simpler components.
Imagine a person who wants to travel to Larissa. A true opinion about the directions can guide them there just as effectively as knowledge. This puzzle sits at the heart of a long debate about the value of knowledge, the worth it holds by expanding understanding and guiding action.
Knowledge often carries instrumental value, the way knowledge of a disease helps a doctor cure a patient. Yet its usefulness depends on circumstances. Memorizing random phone numbers from an outdated phone book yields knowledge with little to no use. Being able to judge that worth shapes real decisions, like which subjects to teach at school and how to allocate funds to research projects.
So why prize knowledge over a mere true opinion when both accurately represent reality? Plato proposed that knowledge is better because it is more stable. Another answer points to practical reasoning, noting that people put more trust in knowledge than in true opinion when deciding what to do. A third reply holds that knowledge has intrinsic value on top of its usefulness, valuable always, while true opinion is valuable only when it happens to help.
To believe that snow is white is to affirm the proposition snow is white. On the common view, beliefs are subjective attitudes that affirm or deny a proposition, representations stored in memory and retrieved when a person thinks or acts. A rival view treats beliefs as behavioral patterns instead. On that account, believing there is mineral water in the fridge is just a cluster of dispositions, like answering questions about it affirmatively and going to the fridge when thirsty. Some theorists deny that beliefs exist at all, calling the concept a folk-psychology oversimplification of more complex processes.
Truth is what beliefs aim at. The correspondence theory of truth says a belief is true if it stands in the right relation to the world by accurately describing it, which makes truth objective. The coherence theory says a belief is true if it belongs to a coherent system of beliefs, which makes truth relative. Pragmatist, semantic, pluralist, and deflationary theories offer further accounts.
Justification is the property of beliefs that meet certain norms about what a person should believe. It separates well-founded beliefs from superstition and lucky guesses, yet it does not guarantee truth, since strong but misleading evidence can justify a false belief. Epistemologists draw a line here between propositional justification, having sufficient reason for a belief, and doxastic justification, holding the belief because of that reason. A person may have good reason to think a neighborhood is dangerous, but if they believe it out of superstition, they have the first kind without the second.
Perception, introspection, memory, reason, and testimony are the often-discussed sources of justification, though philosophers do not all agree how far each one delivers. Perception uses the sensory organs to gather empirical information, splitting into visual, auditory, haptic, olfactory, and gustatory forms. It is not the passive reception of impressions but an active process that selects, organizes, and interprets signals. Introspection turns this inward, toward internal mental states. Seeing a bus at a bus station belongs to perception, while feeling tired belongs to introspection.
Reason serves rationalists as a source of justification for non-empirical facts, explaining how people grasp mathematical, logical, and conceptual truths. It also drives inferential knowledge, where one or more beliefs act as premises supporting another. Memory adds nothing new on its own. It retains and recalls information from other sources, like remembering a phone number perceived earlier. Testimony relies on what one person communicates to another, whether through conversation, a letter, a newspaper, or a blog.
Close to justification sits rationality, sometimes used interchangeably with it, though rationality reaches wider, covering both beliefs and practical matters like decisions and actions. Beyond these run further tools of evaluation. Epistemic norms serve as criteria for judging beliefs, split into deontic norms that prescribe what to believe and axiological norms that identify the goals of belief. These connect to epistemic virtues, character traits like open-mindedness and conscientiousness that help people form true beliefs, the foundation of virtue epistemology. Evidence supports beliefs and is often framed in terms of probability, while a defeater works against a belief. Witness testimony linking a suspect to a crime is evidence of guilt, and an alibi is a defeater. René Descartes' foundationalist epistemology grew from exactly this terrain, the search for a belief immune to doubt.
Philosophical skepticism challenges the very foundations on which knowledge claims rest. Some skeptics narrow their fire. Religious skeptics deny that anyone can know whether deities exist, moral skeptics question moral knowledge, and external world skeptics doubt knowledge of external facts. Global skepticism goes furthest, asserting there is no knowledge in any domain. In ancient philosophy academic skeptics embraced it, while Pyrrhonian skeptics recommended suspending belief to attain tranquility. The dream argument supports this stance, noting that a person dreaming usually cannot tell, so no one can ever be sure they are awake. Critics call global skepticism self-refuting, since denying that knowledge exists is itself a knowledge claim. Fallibilism offers a softer path, agreeing that absolute certainty is impossible but concluding that fallible knowledge still exists.
The debate between empiricism and rationalism turns on where knowledge begins. Empiricists hold that sense experience is the primary source of all knowledge, some describing the mind as a blank slate that builds ideas only from sense data. Rationalists counter that certain knowledge, such as mathematical and logical truth, is reached directly through reason. Some hold the mind carries inborn ideas, others posit a faculty of rational intuition.
Foundationalists and coherentists clash over the structure of knowledge. Foundationalism splits beliefs into basic ones, justified directly, and non-basic ones, justified by other beliefs, like inferring it rained last night from a wet street. Coherentism rejects that split, treating justification as holistic, a web in which beliefs support each other. Foundherentism blends the two, and infinitism argues instead that beliefs form infinite justification chains, each link supporting the next.
Internalism and externalism part ways over the sources of justification. Internalists say justification depends only on factors within the individual, like perceptual experience, memories, and other beliefs, factors accessible through reflection. Evidentialism is the leading internalist view, tying justification to evidence the mind possesses. Externalists insist that at least some relevant factors lie outside the individual, like the quality of a person's eyesight or their ability to differentiate coffee from other beverages. Reliabilism takes this route, requiring a reliable connection between belief and truth, a process that mostly produces true beliefs.
Plato, who lived from 427 to 347 BCE, studied how knowledge differs from true opinion by being based on good reasons, and proposed that learning is a form of recollection, the soul remembering what it had forgotten. His student Aristotle, from 384 to 322 BCE, turned toward scientific knowledge and inference from general principles. The Hellenistic schools followed, with Epicureans treating sensations as always accurate, Stoics trusting only lucid and specific sensations, and skeptics urging suspension of judgment. Neoplatonism, emerging in the 3rd century CE, held that knowledge is infallible and limited to immaterial forms.
The Upanishads, composed in ancient India between 700 and 300 BCE, examined how people acquire knowledge through introspection, comparison, and deduction. In the 6th century BCE the school of Ajnana developed a radical skepticism, while Nyaya, emerging around 200 CE, asserted that knowledge is possible and sorted valid sources from invalid ones. Dharmakirti, in the 6th or 7th century CE, analyzed knowing as a series of causally related events. In China, Mozi, who lived from 470 to 391 BCE, used historical records, sensory evidence, and practical outcomes to validate beliefs. Mencius explored analogical reasoning, and Xunzi sought to combine empirical observation with rational inquiry.
The medieval period made reason and faith its central question. In Arabic-Persian philosophy, al-Farabi and Averroes debated whether philosophy or theology better reaches truth, while al-Ghazali attacked earlier Islamic philosophers for relying on unproven assumptions. Thomas Aquinas, from 1225 to 1274, held that nothing is in the intellect unless it first appeared in the senses. William of Ockham proposed an early direct realism, and in 14th-century India Gangesa developed a reliabilist theory.
Modern philosophy was shaped by René Descartes, from 1596 to 1650, who sought knowledge that cannot be doubted and found it in the assertion I think, therefore I am. He stood with Baruch Spinoza and Gottfried Wilhelm Leibniz among the rationalists. John Locke answered with empiricism and the blank slate, David Hume pressed the limits of certainty, and Immanuel Kant sought a middle ground in principles that structure all experience. In the 20th century, Edmund Gettier conceived counterexamples against knowledge as justified true belief, prompting alternatives like the reliabilism of Alvin Goldman and the virtue epistemology of Ernest Sosa and Linda Zagzebski. The argument that began with a single sentence about kangaroos remains unfinished, still open, as Charles Peirce insisted all knowledge must be, to revision in light of new evidence.
Up Next
Common questions
What is epistemology in philosophy?
Epistemology is the branch of philosophy that examines the nature, origin, and limits of knowledge. Also called the theory of knowledge, it studies concepts like belief, truth, evidence, justification, and reason, and stands alongside ethics, logic, and metaphysics as a main branch of philosophy.
Where does the word epistemology come from?
The word epistemology comes from the ancient Greek terms episteme, meaning knowledge or understanding, and logos, meaning study of or reason, so it literally means the study of knowledge. Despite these ancient roots, the term itself was coined only in the 19th century to name the field as a distinct branch of philosophy.
What are the three components of knowledge in the traditional analysis?
According to the traditional analysis, knowledge has three components: it is a belief that is justified and true. In the second half of the 20th century this view was challenged by thought experiments, including Edmund Gettier's counterexamples, showing that some justified true beliefs do not amount to knowledge.
What is the difference between empiricism and rationalism in epistemology?
Empiricism holds that sense experience is the primary source of all knowledge, sometimes describing the mind as a blank slate. Rationalism argues that some knowledge, such as mathematical and logical truths, is accessed directly through reason without sense experience.
What are the main sources of justification in epistemology?
The often-discussed sources of justification are perception, introspection, memory, reason, and testimony. Perception uses the sensory organs, introspection focuses on internal mental states, memory recalls information from other sources, reason supports non-empirical and inferential knowledge, and testimony relies on what one person communicates to another.
How does skepticism differ from fallibilism in epistemology?
Philosophical skepticism questions the human ability to attain knowledge, and global skepticism asserts there is no knowledge in any domain. Fallibilism agrees that absolute certainty is impossible but rejects the idea that knowledge requires it, concluding that fallible knowledge still exists.
All sources
153 references cited across the entry
- 1harvnbMerriam-Webster (2024)Merriam-Webster — 2024
- 6harvnbCrumley II (2009) p. 16Crumley II — 2009
- 7harvnbO′Donohue, Kitchener (1996) p. [https://books.google.com/books?id=T7uYSFSxxVkC&pg=PA2 2]O′Donohue, Kitchener — 1996
- 10harvnbWolenski (2004) p. 3Wolenski — 2004
- 11harvnbOxford University Press (2024)Oxford University Press — 2024
- 12harvnbAlston (2006) p. 1–2Alston — 2006
- 13harvnbSturm (2011) p. 308–309Sturm — 2011
- 21harvnbHetherington, "''Knowledge''" p. § 1b. Knowledge-ThatHetherington, "''Knowledge''"
- 23harvnbPritchard (2013) p. [https://books.google.com/books?id=sfUhAQAAQBAJ&pg=PA4 4]Pritchard — 2013
- 27harvnbJuhl, Loomis (2009) p. [https://books.google.com/books?id=8kiPAgAAQBAJ&pg=PA4 4]Juhl, Loomis — 2009
- 32harvnbCrumley II (2009) p. 67–68Crumley II — 2009
- 33harvnbIchikawa, Steup (2018) p. § 6.1 Reliabilist Theories of KnowledgeIchikawa, Steup — 2018
- 34harvnbIchikawa, Steup (2018) p. § 5.1 SensitivityIchikawa, Steup — 2018
- 35harvnbCrumley II (2009) p. 69Crumley II — 2009
- 36harvnbIchikawa, Steup (2018) p. § 7. Is Knowledge Analyzable?Ichikawa, Steup — 2018
- 38harvnbMcCormick (2014) p. [https://books.google.com/books?id=3BAhBQAAQBAJ&pg=PA42 42]McCormick — 2014
- 39harvnbPritchard (2013) p. [https://books.google.com/books?id=sfUhAQAAQBAJ&pg=PA11 11–12]Pritchard — 2013
- 42harvnbPritchard, Turri, Carter (2022) p. § 6. Other Accounts of the Value of KnowledgePritchard, Turri, Carter — 2022
- 44harvnbIchikawa, Steup (2018) p. § 1.3 The Justification ConditionIchikawa, Steup — 2018
- 46harvnbBlaauw, Pritchard (2005) p. 92–93Blaauw, Pritchard — 2005
- 47harvnbSilva, Oliveira (2022) p. [https://books.google.com/books?id=99FkEAAAQBAJ&pg=PT10 1–4]Silva, Oliveira — 2022
- 51harvnbSteup, Neta (2024) p. § 5.2 IntrospectionSteup, Neta — 2024
- 53harvnbMcGrew (2011) p. 59McGrew — 2011
- 55harvnbBrown, Gerken (2012) p. 1–2Brown, Gerken — 2012
- 61harvnbCohen (1998) p. § 1. The Philosophical Problem of Scepticism, § 2. Responses to ScepticismCohen — 1998
- 63harvnbWolenski (2004) p. 17–18, 22–23Wolenski — 2004
- 64harvnbLacey, 2005a p. 783Lacey, 2005a
- 65harvnbTieszen (2005) p. [https://books.google.com/books?id=2fgQ_fuCcKAC&pg=PA175 175]Tieszen — 2005
- 67harvnbBradley (2015) p. [https://books.google.com/books?id=qKXDCQAAQBAJ&pg=PA170 170–171]Bradley — 2015
- 69harvnbBradley (2015) p. [https://books.google.com/books?id=qKXDCQAAQBAJ&pg=PA170 170]Bradley — 2015
- 70harvnbKlein (2011) p. [https://books.google.com/books?id=1ETRCwAAQBAJ&pg=PA484 484–485]Klein — 2011
- 71harvnbBlaauw, Pritchard (2005) p. 64Blaauw, Pritchard — 2005
- 72harvnbStairs (2017) p. 155Stairs — 2017
- 76harvnbMittag p. § 2b. EvidenceMittag
- 77harvnbCrumley II (2009) p. 99, 298Crumley II — 2009
- 78harvnbCrumley II (2009) p. 160Crumley II — 2009
- 79harvnbDouven, Schupbach (2014) p. Lead sectionDouven, Schupbach — 2014
- 82harvnbBeebe (2017) p. Lead sectionBeebe — 2017
- 83harvnbLackey (2021) p. 3, 8–9, 13Lackey — 2021
- 84harvnbGerken (2018) p. Lead sectionGerken — 2018
- 87harvnbSharpe (2018) p. 318–320, 328Sharpe — 2018
- 93harvnbPhillips (1998) p. § 2. Mīmāṃsā Self-certificationalismPhillips — 1998
- 96harvnbSturm (2011) p. 303–304, 308–309Sturm — 2011
- 97harvnbSturm (2011) p. 304Sturm — 2011
- 100harvnbSiegel, Silins, Matthen (2014) p. 781Siegel, Silins, Matthen — 2014
- 101harvnbConee (1998) p. Lead sectionConee — 1998
- 102harvnbPritchard (2004) p. [https://www.jstor.org/stable/3050633 326]Pritchard — 2004
- 103harvnbWarren (2020) p. § 6. The Epistemology of LogicWarren — 2020
- 104harvnbChignell (2018) p. Lead sectionChignell — 2018
- 106harvnbBarber (2003) p. 1–3, 10–11, 15Barber — 2003
- 107harvnbVaidya, Wallner (2021) p. 1909–1910Vaidya, Wallner — 2021
- 108harvnbCroce (2023) p. Lead sectionCroce — 2023
- 109harvnbMaguire (2015) p. [https://link.springer.com/chapter/10.1007/978-94-017-9309-4_4 33–34]Maguire — 2015
- 110harvnbAlston (2006) p. 2Alston — 2006
- 112harvnbWheeler, Pereira (2004) p. 469–470, 472, 491Wheeler, Pereira — 2004
- 114harvnbPomerol (2012) p. 190Pomerol — 2012
- 115harvnbStairs (2017) p. [https://books.google.com/books?id=Km1QDwAAQBAJ&pg=PA156 156]Stairs — 2017
- 116harvnbHansen (2023) p. § 3.5 The Epistemic Approach to FallaciesHansen — 2023
- 120harvnbSteele, Stefánsson (2020) p. Lead sectionSteele, Stefánsson — 2020
- 122harvnbHarasim (2017) p. 11Harasim — 2017
- 123harvnbHarasim (2017) p. 11–12Harasim — 2017
- 125harvnbEmaliana (2017) p. 59–61Emaliana — 2017
- 129harvnbPappas (1998) p. § Ancient PhilosophyPappas — 1998
- 131harvnbAdkins, Adkins (2014) p. [https://books.google.com/books?id=zGY1Sqjwf8kC&pg=PA393 393]Adkins, Adkins — 2014
- 132harvnbGerson (2014) p. [https://books.google.com/books?id=yhcWBAAAQBAJ&pg=PA266 266–267, 277–278]Gerson — 2014
- 133harvnbDunne (2006) p. 753Dunne — 2006
- 134harvnbBlack p. Lead sectionBlack
- 137harvnbBonevac (2023) p. [https://books.google.com/books?id=neDwEAAAQBAJ&pg=PR18 xviii]Bonevac — 2023
- 142harvnbWolenski (2004) p. 11Wolenski — 2004
- 144harvnbWolenski (2004) p. 14–15Wolenski — 2004
- 148harvnbPappas (1998) p. § Modern Philosophy: From Hume to PeircePappas — 1998
- 150harvnbHamlyn (2006) p. 317–318Hamlyn — 2006
- 152harvnbCrumley II (2009) p. 183–184, 188–189Crumley II — 2009