Philosophy › Section G · Epistemology › Topic G.3

Justification

Justification is what separates a reasonable belief from a hunch. Explaining where it comes from raises some of epistemology's deepest disputes: how chains of reasons end, and whether justification lies inside the mind or in its connection to the world.

What justification does

"Knowledge and the Gettier problem" showed that knowledge needs more than true belief, and justification is the usual candidate for the extra element. A justified belief is one the believer has good grounds for: grounds that make it likely, though not certain, to be true. Justification comes in degrees, since evidence can be weak or strong. And it is fallible: a belief can be fully justified and still false, as when a reliable friend's directions turn out to be wrong. The theories below ask what grounds count and how they are structured.

The regress problem

If a belief is justified by a reason, that reason is a further belief, which seems to need its own justification, and so on. The ancient sceptic Agrippa, whose arguments Sextus Empiricus reports, said the chain has only three ways to go. It runs on forever, which seems impossible for a finite mind; it loops back on itself, which looks circular; or it stops at some belief accepted without reasons, which looks arbitrary. Each option seemed to him to leave the original belief unjustified. In India, Nāgārjuna, of the second century CE, pressed a similar regress against the Nyāya theory of the pramāṇas: what certifies the means of knowledge themselves?

Foundationalism

Foundationalists accept that the chain stops, but deny that it stops arbitrarily. Some basic beliefs are justified without being inferred from others, and everything else rests on them, as a house rests on its foundations. René Descartes was a classical foundationalist: the foundations must be indubitable, such as his awareness that he is thinking or that he seems to see a green patch. The objection is that so little is indubitable that hardly any ordinary knowledge can be built on it. Modest foundationalists reply that basic beliefs, such as ordinary perceptual ones, need only be justified, not certain. Critics ask what justifies a basic belief if not another belief.

Coherentism

Coherentists deny that any belief is basic. A belief is justified by fitting into a system of beliefs that support and explain one another, more like a web than a building. Coherence here means more than consistency, which the topic "How philosophers arrive at truth" showed is not enough for truth: beliefs must reinforce each other. Otto Neurath compared us to sailors who must rebuild their ship at sea, plank by plank, never able to start again in dry dock. Objections: the support looks circular, and a perfectly coherent system, like a detailed novel, could be cut off from reality.

Worked example: tracing a regress

Leila believes the river bridge is closed. 1. Ask for her reason: a neighbour told her. 2. Ask why trust the neighbour: he has been reliable before, and she remembers him saying it. 3. Ask why trust memory: here the theories divide. 4. Foundationalist reading: her memory experience and her past observations of the neighbour are basic, justified without further inference. 5. Coherentist reading: the belief is justified because it fits everything else she believes: the roadworks sign she passed, the heavy traffic on the other route. 6. Infinitist reading, defended by Peter Klein: justification grows as more reasons are available, with no final stopping point.

Internalism and externalism

Internalists hold that whatever justifies a belief must be accessible to the believer on reflection: their evidence, experiences or reasons. Externalists hold that justification can depend on facts the believer cannot access, such as whether the process producing the belief is in fact reliable. Consider a belief you formed by reading a trustworthy encyclopedia, though you have now forgotten where you learned it. An internalist may say you are no longer justified, unless something else you can recall supports it; an externalist says the reliable origin still justifies you. Internalism ties justification to responsible believing; externalism ties it to being connected to the truth.

Reliabilism

The best-known externalist theory is Alvin Goldman's process reliabilism (1979): a belief is justified if it was produced by a reliable process, one that tends to yield true beliefs. Vision in good light, memory and careful reasoning qualify; wishful thinking and guessing do not. The view explains why young children and animals can have justified beliefs they cannot defend. Two objections are prominent. The new evil demon problem, pressed by Keith Lehrer and Stewart Cohen in the 1980s: a person deceived by a demon, with experiences just like ours, seems as justified as we are, though their processes are unreliable. The generality problem, noted by Goldman himself: the same process can be described broadly or narrowly, with different reliability.

Virtue epistemology

Virtue epistemologists shift attention from beliefs to believers. Ernest Sosa treats reliable faculties such as perception and memory as intellectual virtues: knowledge is true belief that is right because of the believer's competence, as an archer's shot hits because of skill, not a lucky gust. Linda Zagzebski stresses character traits instead: open-mindedness, intellectual courage, carefulness and humility. Critics such as Jennifer Lackey (2007) object that when you learn a fact from a stranger's directions, the credit seems to belong mostly to the stranger, yet you still know.

Worked example: four theories, one case

Ravi believes the library opens at nine. He read it on the library's website last month but has forgotten the source. 1. Internalist: he can recall no reason, so his justification is in doubt, unless he can reflect that such beliefs of his are usually reliable. 2. Reliabilist: reading an official website is a reliable process, so he is justified. 3. Virtue theorist: the belief comes from his competent reading and memory, so it is creditable. 4. Now imagine Ravi's double in a demon world, with identical experiences but false beliefs. 5. Internalists say the double is equally justified; strict reliabilists must say he is not. 6. The case shows where the theories divide.

Fallible sources

Every source of justification can fail. Memory can mislead: seeming to remember feels the same as remembering. Inference is only as good as its premises, and inductive inference, as "Deductive, inductive, and abductive inference" explained, gives probability, not certainty. Testimony depends on the honesty and competence of others. Perception is open to illusion: in the Müller-Lyer illusion, two equal lines look unequal. Nyāya philosophers answered regress worries practically: a cognition is shown to be veridical by checking its source and by success in action, as when water you seem to see quenches your thirst. Fallibility is why sceptics, discussed next, demand more.

Common mistakes to avoid

First, equating justification with truth: a justified belief can be false, and a true one unjustified. Second, treating coherence as mere consistency; coherentists require mutual support. Third, assuming foundationalism requires certainty; modest foundationalists allow fallible basic beliefs. Fourth, confusing internalism with the claim that you must be thinking of your reasons right now; it requires only that they be accessible on reflection. Fifth, thinking reliabilism needs the believer to know the process is reliable; on Goldman's view, the process need only be reliable in fact.

Key takeaways

Justification makes belief likely to be true, comes in degrees and is fallible. Agrippa's regress offers three options: endless chains, circles or stopping points. Foundationalism stops at basic beliefs, classically certain, modestly fallible; coherentism justifies by mutual support, facing circularity and isolation objections; infinitism accepts endless chains. Internalists require reflectively accessible justifiers; externalists do not. Goldman's reliabilism faces the new evil demon and generality problems. Virtue epistemology grounds knowledge in competences or character. Nāgārjuna's regress and Nyāya's practical reply show the same debate in classical India.

Flashcards

Flashcard

Justification

Good grounds making a belief likely to be true

Comes in degrees and is fallible.

Test your understanding

Question 1 of 5

Which theory holds that a belief is justified by fitting into a system of mutually supporting beliefs, with no beliefs that are basic?

Question 2 of 5

A person deceived by a demon has experiences exactly like yours and reasons just as carefully. Many think she is as justified as you are. Which theory does this case most directly challenge?

Question 3 of 5

You formed a belief from a reliable encyclopedia but have forgotten the source and can recall no reason for it. What would a typical externalist say?

Question 4 of 5

According to Agrippa's trilemma, which three options does a chain of reasons face?

Question 5 of 5

What distinguishes modest foundationalism from classical foundationalism?

← Previous topicNext topic →