Justified True Belief

Mapping the Landscape of Good Reasons for the Truth of Christianity

Christian Evidences

Evidence & Arguments for Christian Theism

Divinity of Christ

Biblical Evidence for High-Christology: Jesus is God

TBD

Old Testament Criticism

Evidence for the Reliability of the OT Bible

Common Objections

Objection Analyses to Christian Theism

World Religions

Critical Analyses of Non-Christian Religions

Philosophy

Phileō Sophia - to Love Wisdom.

Epistemology

Study of Knowledge

Coherentism

(●) Coherentism holds that beliefs are justified by how well they fit together, not by resting on foundations. + A detective sits at her kitchen table late at night with the case file spread out in front of her. No single piece of evidence settles it. The receipt could have been planted, the witness could be mistaken, and the phone record could have an innocent explanation. But laid side by side, each piece explains the others, and the story they tell together is hard to doubt. That is the picture behind coherentism. - Epistemology, from the Greek word for knowledge, is the part of philosophy that studies knowledge and justification. - Justification is the good standing a belief needs, besides being true, to count as knowledge. A justified belief is one it is reasonable to hold. If you believe it will rain because you read the forecast, your belief is justified; if you believe it because you flipped a coin, it is not, even if it rains. - Coherentism says that a belief is justified by how well it fits into the whole system of a person's beliefs. Nothing is believed on its own, and there is no bottom layer. Every belief is held up by its connections to the rest. The view began as an answer to an old puzzle. Every reason seems to need a reason of its own, like a child who meets each explanation with another "But why?" Why do you think the shop is closed? Because the sign on the door says so. But why trust the sign? Philosophers call this the regress of reasons, and the question is where, if anywhere, it stops. Foundationalism, the older answer, stops it at basic beliefs that need no further reason: beliefs with some standing of their own, such as your awareness that you are in pain right now. Coherentism denies there are any such beliefs. On its view the support runs sideways, across the whole web, rather than up from the bottom. So what makes a set of beliefs fit together? Coherentists usually name several things. - Consistency. No two of the beliefs contradict each other. - Inferential connection. The beliefs support one another by reasoning: this one makes that one likely. - Explanation. The beliefs explain each other. A belief that it rained explains the wet street, the puddles, and the umbrella in the hall. - Few loose ends. The more beliefs stand apart, connected to nothing else, the less coherent the system. Think of a crossword puzzle half filled in. An answer fits because it crosses other answers that fit. Each entry is checked by its neighbors, and the more entries interlock, the more confident you are in all of them.

(●) Neurath's boat, Quine's web, and the coherentists of the twentieth century. + Coherentism came into its own in the twentieth century. Its defenders gave it its best-known pictures, and one philosopher who was not a coherentist at all gave it its most influential argument. - Neurath's boat. Otto Neurath, an Austrian philosopher of science, wrote in 1932 that we are like sailors who must rebuild their ship on the open sea, plank by plank, without ever bringing it into dry dock to start again from the keel. There is no place to stand outside our beliefs. - Quine's web. W. V. Quine (1951) pictured our beliefs as a fabric that meets experience only along its edges, and later, with J. S. Ullian, he called it the web of belief (1970). When experience conflicts with the web, something has to give, and we revise wherever the change disturbs the rest least. No belief, he argued, is immune from revision. - Sellars's challenge. Wilfrid Sellars (1956) attacked what he called the myth of the given: the idea that experience hands us something that can justify a belief without itself needing justification. Sellars had no love for the coherentist picture either, and likened it to a serpent of knowledge with its tail in its mouth. Even so, his attack persuaded many readers that if only beliefs can be reasons, foundations of the classic kind are impossible, and coherence is what remains. - Davidson's slogan. Donald Davidson (1986) put the point most sharply: "nothing can count as a reason for holding a belief except another belief." - BonJour's theory. Laurence BonJour (1985) gave the most careful defense. He tied coherence to observation by requiring that a justified system include beliefs that arise spontaneously from experience, and he tried to show that coherence kept up over time makes truth likely. - Lehrer's theory. Keith Lehrer (1990) held that a belief is justified when it can meet every objection from within the believer's system of acceptances, his word for the claims a person accepts for the sake of getting at the truth. One of those acceptances is the person's acceptance that they are trustworthy in what they accept. One distinction keeps the picture clear. Coherence as a test of justification is different from coherence as a theory of truth. Some idealists, philosophers who hold that reality is at bottom mind-like rather than material, held that truth itself is coherence. Brand Blanshard, an American who carried on the British idealist tradition, held it, and so, on the usual reading, did F. H. Bradley before him. Most coherentists about justification reject that. They hold with common sense that a belief is true when it matches the way things are: "the milk is in the fridge" is true when the milk is in the fridge.

(●) Classical foundationalism answers that coherence multiplies support but cannot create it. + Timothy McGrew defends classical foundationalism, the view that all knowledge rests on foundations that are certain. There are two kinds: a person's awareness of their own present experience, and truths seen to be true just by understanding them, such as two plus two making four. Everything else is reached by reasoning from those, and the reasoning yields probability. His argument that the foundations must be certain follows C. I. Lewis (1946): "if anything is to be probable, then something must be certain." - The thought behind Lewis's line is that a belief is always probable on something, some evidence. If that evidence were itself only probable, and its evidence too, all the way down, nothing would ever get started. It would be like a stack of IOUs with no money behind any of them. - Lewis also gave the classic argument about coherence, which he called congruence. He pictured several witnesses, none of them very reliable, who tell the same detailed story without having compared notes. Their agreement can make the story highly probable, but only if each of them has some credibility of their own, some real chance of having seen what happened. - Among people with no access to the facts, it is a different matter. There, detailed agreement is either a fluke, or collusion (secretly comparing notes), or a shared rumor, and none of those makes the story any more likely to be true. - On the classical view, then, coherence is real evidence, and a foundationalist uses it constantly. But it works by multiplying support that is already there. It cannot start from nothing. The argument against coherentism is deductive. Its premises work like a locked chain: if each link holds, the conclusion cannot slip out. Each premise need only be more plausible than its denial, not certain beyond all doubt.

(P1) If coherentism is true, coherence alone can make beliefs probable. + Coherentism makes coherence the only source of justification. And justification is a matter of how probable a belief is made by whatever supports it. (1) Coherentism denies any source but coherence. - That is exactly what separates it from foundationalism. On coherentism no belief has any standing of its own; whatever justification a belief has, it gets from fitting with the others. - Think of the difference between a pyramid and a raft. In a pyramid each layer rests on the one beneath, and the bottom layer rests on the ground. In a raft the planks are lashed to one another, and nothing rests on anything below. Coherentism is the raft. - The classic coherentists drew the line there themselves. If some beliefs had a little standing of their own, from experience for instance, those beliefs would be foundations, however modest. Views that grant that much are weighed at (D5) and (D8). (2) Justification must make a belief likely to be true. - A justified belief is a good bet on the truth, given what the believer has to go on. What would be the use of a good standing that had nothing to do with whether the belief is true? It would not be worth having. - Leading coherentists agree. The most careful defense of the view devotes a long argument to showing that coherence makes truth likely. - Think of a weather forecast. A forecast is good when the evidence behind it makes rain probable, not when the announcer merely sounds confident. (3) So coherentism needs coherence to do this work alone. - If coherence is the only source of justification, and justification makes beliefs probable, then coherence alone must be able to make beliefs probable. - Think of a family with a single income. If the bills get paid at all, that one paycheck is paying all of them. Therefore, if coherentism is true, coherence by itself, starting from beliefs with no standing of their own, can make those beliefs probable.

(P2) Coherence alone cannot make beliefs probable. + Coherence can raise probability, but only when each belief already has some support of its own. (1) Agreement adds weight only to witnesses who have some. - Two witnesses who did not compare notes and tell the same story make it far more probable. The law of Moses relied on exactly that: "Only on the evidence of two witnesses or of three witnesses shall a charge be established" (Deuteronomy 19:15). - But the rule assumes that each one is a real witness, someone who saw something. Two people who saw nothing add nothing. If their stories happen to match, that is a fluke, and if they match in every detail, someone has been talking to someone (D12). - Probability theory bears this out. A report is evidence to the degree that it is more likely to be given if the story is true than if it is false, and that ratio is the report's evidential weight. A witness who really saw a car might be ten times likelier to say "red" if it was red than if it was not, so her report has a weight of ten. - When independent witnesses agree, their weights multiply, which is why agreement can be so powerful. A witness who saw nothing, though, is just as likely to say it whether the story is true or false, so he brings a weight of exactly one. Multiplying by one, however many times, leaves the odds where they were. (2) Coherence can be built without truth. - A good novel is perfectly coherent. Its characters are consistent, its clues add up, and every event explains the next. None of it happened. - Think of a group of conspirators who have rehearsed their alibi for a week. Their stories fit together without a seam, precisely because they were built to fit. - Fit is evidence of truth only when truth explains the fit better than its rivals do. Where an author or a rehearsal explains it just as well, the fit tells you nothing. Philosophers call this the isolation objection: coherence alone does not connect a set of beliefs to the world. (3) Equally coherent systems can contradict each other. - For any coherent system of beliefs, a different system could be built that is just as coherent and says the opposite on many points. Coherence alone gives no way to choose between them. - Think of two maps of an island, each perfectly drawn, with every road meeting every other where it should. One is the real island and one is invented, and neither map can tell you which is which. - Coherentists reply that a person's actual system keeps taking in beliefs from experience, which an invented one does not. That is their best answer, and it is weighed at (D3). (4) Formal results point the same way. - Beginning in the 1990s, philosophers tested coherence with the mathematics of probability. An early result showed that adding a belief can make a set fit together better while making it less likely to be true. - Think of a detective whose evidence says the suspect did it, while a witness swears he was across town. Add the belief "he has an identical twin," and suddenly everything fits. But the fuller story, with one more thing in it that could be wrong, is less probable than before. - Later proofs went further. Hold fixed the other things that matter, such as how reliable each source is on its own and how likely each belief was to begin with. Even then, no way of measuring how well beliefs fit together reliably tracks how likely they are to be true. - These results rest on technical assumptions, and philosophers still argue over how much they show. They also reach further than (P2) needs, since they concern beliefs that already have some credibility, while (P2) is about beliefs that have none. But they point the same way as the witnesses and the alibi: more coherence does not, by itself, mean more probable. Therefore, coherence alone, among beliefs with no standing of their own, cannot make them probable.

(C) Therefore, coherentism is false. + Coherentism is false. If it were true, coherence alone could make beliefs probable (P1). It cannot (P2). So coherentism is not true. That pattern of reasoning is called modus tollens, and it is valid, meaning that true premises guarantee a true conclusion. It runs: if A, then B; not B; therefore, not A. If it had rained, the street would be wet. The street is dry. So it did not rain. (1) Coherence still matters. - Coherence is not worthless. What it cannot do is work alone. Where beliefs already have some support, their agreement can raise it enormously, and every good detective, historian and scientist relies on that. (2) Something must supply the first support. - If coherence multiplies support but cannot create it, the support has to come from somewhere. The witnesses show only that some first support is needed. - Classical foundationalism adds, by a further argument, where the deepest of it comes from: experience and truths seen to be true, which are known with certainty. Everything else grows from there through reasoning, coherence included. - Think of the witnesses again. Their agreement is worth a great deal, but only because each of them first saw something. (3) Coherentism raised the right questions. - Coherentists were right about several things: no belief should be held in isolation, experience is never the whole story, and our beliefs must be revised together. Those lessons survive the failure of the view. See also: Philosophy / Epistemology: Classical Foundationalism Philosophy / Epistemology: Skepticism Philosophy / Logic: Cumulative Case Arguments

On coherentism: Blanshard, Brand. The Nature of Thought. 2 vols. London: George Allen & Unwin, 1939. BonJour, Laurence. The Structure of Empirical Knowledge. Cambridge, MA: Harvard University Press, 1985. Bradley, F. H. Essays on Truth and Reality. Oxford: Clarendon Press, 1914. Davidson, Donald. "A Coherence Theory of Truth and Knowledge." In Truth and Interpretation: Perspectives on the Philosophy of Donald Davidson, edited by Ernest LePore, 307-319. Oxford: Blackwell, 1986. Haack, Susan. Evidence and Inquiry: Towards Reconstruction in Epistemology. Oxford: Blackwell, 1993. Harman, Gilbert. Change in View: Principles of Reasoning. Cambridge, MA: MIT Press, 1986. Lehrer, Keith. Self-Trust: A Study of Reason, Knowledge, and Autonomy. Oxford: Clarendon Press, 1997. Lehrer, Keith. Theory of Knowledge. Boulder, CO: Westview Press, 1990. Neurath, Otto. "Protokollsätze." Erkenntnis 3 (1932): 204-214. Quine, W. V. "Two Dogmas of Empiricism." Philosophical Review 60 (1951): 20-43. Quine, W. V., and J. S. Ullian. The Web of Belief. New York: Random House, 1970. Sellars, Wilfrid. "Empiricism and the Philosophy of Mind." In Minnesota Studies in the Philosophy of Science, vol. 1, edited by Herbert Feigl and Michael Scriven, 253-329. Minneapolis: University of Minnesota Press, 1956. On coherence and probability: Bovens, Luc, and Stephan Hartmann. Bayesian Epistemology. Oxford: Clarendon Press, 2003. Huemer, Michael. "Probability and Coherence Justification." Southern Journal of Philosophy 35 (1997): 463-472. Klein, Peter, and Ted A. Warfield. "What Price Coherence?" Analysis 54 (1994): 129-132. Lewis, C. I. An Analysis of Knowledge and Valuation. La Salle, IL: Open Court, 1946. Olsson, Erik J. Against Coherence: Truth, Probability, and Justification. Oxford: Clarendon Press, 2005. Shogenji, Tomoji. "Is Coherence Truth Conducive?" Analysis 59 (1999): 338-345. On foundationalism and the case against coherentism: BonJour, Laurence. "The Dialectic of Foundationalism and Coherentism." In The Blackwell Guide to Epistemology, edited by John Greco and Ernest Sosa, 117-142. Oxford: Blackwell, 1999. BonJour, Laurence. In Defense of Pure Reason: A Rationalist Account of A Priori Justification. Cambridge: Cambridge University Press, 1998. BonJour, Laurence, and Ernest Sosa. Epistemic Justification: Internalism vs. Externalism, Foundations vs. Virtues. Malden, MA: Blackwell, 2003. Fumerton, Richard. Metaepistemology and Skepticism. Lanham, MD: Rowman & Littlefield, 1995. McGrew, Timothy J. The Foundations of Knowledge. Lanham, MD: Rowman & Littlefield, 1995. McGrew, Timothy, and Lydia McGrew. Internalism and Epistemology: The Architecture of Reason. London: Routledge, 2007. Critics of classical foundationalism: Bergmann, Michael. Justification without Awareness: A Defense of Epistemic Externalism. Oxford: Oxford University Press, 2006. Christian apologists engaged: Carnell, Edward John. An Introduction to Christian Apologetics. Grand Rapids, MI: Eerdmans, 1948. Clark, Gordon H. A Christian View of Men and Things. Grand Rapids, MI: Eerdmans, 1952. Clark, Gordon H. Three Types of Religious Philosophy. Nutley, NJ: Craig Press, 1973. Biblical citations: Deuteronomy 19:15; Acts 2:32. Scripture quotations are from the English Standard Version.
+ Neurath's boat and Quine's web: no belief is immune from revision, not even a report of what someone observed. We repair our beliefs plank by plank while staying afloat, and that is just what coherentism describes.
1. The picture is true of how we revise. Nobody rebuilds their beliefs from scratch. We learn something new and adjust whatever it touches, keeping most of what we had. Neurath went further. Even the reports scientists treat as their data, which he called protocol sentences (a report like "Otto saw a table in the room at a quarter past three"), can be thrown overboard if they clash with too much else. Quine agreed that any statement can be kept or given up, given enough adjustment elsewhere. - About beliefs concerning the world, they were right, and a foundationalist should say so. - Quine extended the point even to the laws of logic, and there the classical view parts company with him. A truth seen to be true by understanding it is not revised by rearranging the web. - In fact, the web cannot be rearranged without such truths. When experience clashes with your beliefs and you decide what to give up, you are already assuming that a contradiction must go. A rule used to settle every revision cannot itself be justified by a revision. - Think of a referee and a rulebook. The rulebook settles every disputed call, and no disputed call can settle the rulebook. 2. The revisable reports are claims about the world. Neurath's protocol sentences, and the reports along the edge of Quine's fabric, describe public things: a table, a meter reading, a flash on a screen. Classical foundationalism agrees that such claims can be wrong, which is exactly why it does not make them foundations. The foundations are a person's awareness of their own present experience, and no rearranging of the web touches that. - Think of a driver who sees water shimmering on a desert road. She revises "there is water ahead" the moment the car reaches the spot. She never has to revise "it looked like water," and that was what she had to go on all along. 3. How we revise is different from what justifies. A family can renovate a house room by room while living in it. No room is built from nothing, and yet the house stands on its foundation the whole time. In the same way, a person revising beliefs one at a time can be relying throughout on experience and on truths seen by reflection, which are not just more beliefs in the web. - The two pictures leave room for this themselves. Neurath's sailors are afloat on a real sea, and Quine's fabric meets experience along its edges. What pure coherentism must explain is how that meeting gives anyone a reason for belief when experience has no standing of its own (P2).
+ Wilfrid Sellars and Donald Davidson: only a belief can be a reason for a belief. Experience is not a belief, so it cannot justify anything, and foundations of the classic kind are a myth.
1. This is the most influential argument for coherentism. Sellars's attack, in the form later philosophers sharpened it into, is a dilemma: a choice between two options that both seem to lose. - Either what experience gives us is like a belief, with a content (something it says) that can be true or false. Then it needs justifying too. - Or it is not like a belief. Then it cannot be a reason for anything. Davidson drew the coherentist conclusion: "nothing can count as a reason for holding a belief except another belief." 2. A person can be directly aware of an experience. Think of a headache. You do not infer it from evidence, the way you might infer a storm from dark clouds. The pain itself is present to you, and you are aware of it directly. Philosophers call this acquaintance. It is not a belief, and it is not nothing either. 3. A belief can contain the experience it is about. The simplest foundational belief points straight at an experience: "this is like this." - Picture yourself looking at a ripe tomato. You attend to the patch of red you are seeing and think, "this is like this." The experience itself is part of what the belief says. - So the belief is not floating free of the experience. And it cannot be false, because without the experience there would be no belief at all: take away the red patch, and there is nothing for "this" to point to. - That is how the classical view meets Sellars's dilemma. It grants the second side: experience on its own, apart from any belief, is not a reason. But the foundation is a belief, so it has content and can be a reason. And the first side fails, because that content is fixed by what is present to the mind, so the belief is not left waiting on some further belief to justify it. 4. Causes are not reasons. Davidson did not deny that experience matters. He held that sensations cause beliefs, and he granted that a cause is not a justification: knowing what produced a belief does not show that the belief is reasonable. So on his view the web touches the world only through causes, and he needed a separate argument to show that a coherent web is mostly true (D7). - Think of a smoke alarm. A fire sets it off, and so does burnt toast. The noise alone gives you no reason to tell the two apart. - A theory on which experience causes beliefs but never supports them has to find some other way to reconnect the web with the world. That is the isolation objection again (P2). See also: Philosophy / Epistemology: Classical Foundationalism
+ Laurence BonJour's observation requirement ties coherence to the world. A justified system must keep taking in beliefs that arise unbidden from experience, and if it stays coherent over time, the best explanation is that it is true.
1. The requirement was a real improvement. BonJour saw the isolation objection clearly and tried to answer it. He required that a justified system contain observation beliefs, beliefs that come to us unbidden, like "there is a red book on the desk," and that the system include beliefs about how reliable such observations are. Then he argued that if a system keeps taking in such beliefs and still hangs together over the long run, the best explanation is that it is true. This is the most serious answer coherentism has given. 2. On coherentism, those beliefs get their weight only from the web. An observation belief, on BonJour's coherentist view, is justified not by the experience but by the system's belief that observations of that kind are usually reliable. That belief, in turn, is justified only by coherence. So the observations count only because the web says they count, and the web is what was in question. - Think of a newspaper that prints letters from readers and vouches for them by citing its own editorial policy. The letters connect the paper to its readers only if the policy is trustworthy, and the paper cannot establish that by quoting itself. 3. The argument needs facts the web cannot supply. Reasoning from long-run order to its best explanation is good reasoning, and a classical foundationalist uses it to reach a real world (D6). What decides the matter is where the reasoning starts. BonJour's argument from long-run coherence starts from facts about the believer herself: which beliefs she holds, and which of them arose unbidden. - How does she know those facts? Not from coherence, BonJour saw, because judging whether her beliefs cohere already requires knowing what they are. That would be going in a circle. - So he presumed them, and called this the doxastic presumption (from the Greek doxa, belief): the assumption that a person's picture of her own beliefs is roughly right. He granted that it left justification conditional, good only if that picture is right. - Think of an auditor who declares a company's books consistent. Her verdict is worth something only if the ledger in front of her is the company's real ledger, and the ledger cannot tell her that. - Here is the difference between the two arguments. The foundationalist reasons from experiences she is directly aware of, data with standing of their own that no presumption props up. BonJour's coherentist reasons from her picture of her own beliefs, which on his view has no standing except what coherence gives it. - That leaves the presumption in an awkward spot. Everything rests on it, and coherence does not justify it. So it is either a foundation in all but name or a support with nothing under it. 4. BonJour himself gave up coherentism. In later work BonJour concluded that the view could not be made to work, chiefly because he could find no way within it for experience to count as a reason. He turned to a foundationalism in which beliefs about experience are justified by direct awareness of the experience, though he did not require them to be infallible. Of course, a change of mind is not an argument, and other coherentists hold their ground. But his reasons were the isolation objection (P2) and the question of how experience can count as a reason (D2), and those have not gone away. See also: Philosophy / Epistemology: Skepticism
+ Keith Lehrer: every theory of knowledge ends in self-trust, foundationalism included. Coherentism is honest about it, and the person's acceptance that they are trustworthy is the keystone that holds the system together.
1. Lehrer's view is a sophisticated one. On Lehrer's account, a belief is justified for a person when it can beat or neutralize every objection to it, judged by what the person accepts. To beat an objection is to accept the belief as more reasonable than the objection; to neutralize it is to accept something else that takes away its force. At the center is a principle of self-trust, which he likened to a keystone: the person accepts that they are trustworthy in what they accept. - Think of someone who believes she sees a hawk, and a friend objects that her eyes could be fooling her. On Lehrer's view she answers from what she accepts: the light is good, her eyes are sound, she knows what hawks look like, and she is trustworthy about such things. - Lehrer grants that the principle of self-trust helps to justify itself, which is a loop, and he argues that the loop is no vice. Everyone must trust their own faculties somewhere, he says, and it is better to admit it. 2. A loop of self-trust certifies any system equally. Suppose the person's trustworthiness is accepted because it coheres with their other acceptances. Then the system is vouching for its own trustworthiness, and the same loop is open to anyone whose acceptances hang together, however badly they fit the world (P2). If instead the trustworthiness is accepted for some other reason, that reason is a foundation. - Think of a club whose membership rules can be changed only by a vote of members, and whose members are chosen only by those rules. Everything is in order inside the club, and nothing inside it can show the rules are wise. 3. The classical foundations do not rest on self-trust. The reply that everyone trusts themselves somewhere assumes that every foundation needs a further premise, "I am reliable." The classical foundations need no such premise. "This is like this," thought while attending to an ache, is true because the ache is part of what the belief says, not because the believer has a good track record. "Two plus two is four" is seen to be true rather than accepted on one's own say-so, and on the classical view so are the principles that say what evidence supports what. Using one's mind is not the same as assuming, as a premise, that one's mind is reliable. - Think of checking a sum on a receipt. You do not first certify yourself as good at sums. You look at seven plus five and see that it is twelve. 4. Lehrer's test for knowledge lies outside the believer's view. Lehrer saw the danger himself: a person whose acceptances hang together can still be untrustworthy. So for knowledge he added a further test. The justification must survive every correction of the person's false acceptances, and he called the corrected systems the ultrasystem. - That does tie knowledge to truth. But the believer cannot check it from the inside. - Lehrer's account of personal justification is internalist: what justifies a belief must be available to the believer, not just a fact about the world they may know nothing of. So, from the inside, the person still has only coherence to go on. - Think of a man whose memory is quietly failing. He can accept, calmly and consistently, that his memory is reliable, and everything else he accepts can hang together around that one belief. Nothing inside his system tells him anything is wrong. See also: Philosophy / Epistemology: Externalism
+ Moderate coherentism: beliefs a person already holds have some standing just by being held, as Gilbert Harman argued, and coherence builds on that. No belief is privileged or beyond revision, so this is still coherentism.
1. This is a serious position. Harman defended what he called a coherence theory of reasoning against a foundations theory. At its heart is a principle of conservatism: keep what you already believe unless you have a special reason to give it up. Other coherentists grant some initial credibility to memory and perception. And the mathematics favors the position: the probability results agree that coherence can raise probability a great deal when beliefs start with some credibility. All of this should be granted. 2. Credibility that coherence did not supply is a foundation. Whatever gives a belief standing before it is fitted into the web is a second source of justification, besides coherence. In the standard classification, a view on which some beliefs have a little credibility of their own, which coherence then raises, is called weak foundationalism, and the classic coherentists drew the line in the same place (P1). - Revisability does not change this. Modest foundationalists, who do not require foundations to be certain, hold that their foundations can be overturned too. - Think of a jury that convicts on the agreement of five witnesses. If it says each witness was credible to begin with, it has granted that the case rested on something besides their agreement. 3. Believing a thing does not make it likely. As a rule of economy, conservatism is sensible. Nobody can recheck everything, and a person who threw out every belief she could not defend on the spot would soon have none left. As an account of what makes a belief likely to be true, though, it fails. Holding a belief is evidence for it only if the way the believer came to hold it is connected to the truth, and that connection is one more thing to have a reason for. - Think of a man who wakes up convinced that his neighbor is a spy. Conservatism lets the conviction stand until something counts against it. Its being his conviction does not make it any more likely to be true. - Where the starting credibility really comes from memory or the senses, it is evidence, and the view is foundationalism under another name. 4. The dispute is then about how firm the foundations are. Once initial credibility is granted, the live question is what supplies it and how strong it must be. Classical foundationalism answers that the deepest support comes from experience and truths seen to be true, which are certain. Everything else, including the credibility of ordinary perceptual beliefs (beliefs from the senses, like "there is a cup on the table"), is reached from them by reasoning. See also: Philosophy / Epistemology: Classical Foundationalism
+ The foundationalist's own route to the world uses coherence. A single experience proves nothing about the world, yet the agreement of many is supposed to make a real world probable, so coherence creates support after all.
1. The agreement of the senses really does matter. The case for a real world does lean heavily on the agreement of the senses and the order of experience. You see the campfire, feel its heat on your hands, and hear it crackle, and each confirms the others. Coherence plays a central role in that reasoning, and a foundationalist has no reason to be embarrassed by it. 2. The experiences are data in hand. Notice what is agreeing in that reasoning. It is a set of experiences, and a person's awareness of their own experiences is certain. The sight, the heat and the crackle are not claims about the world still waiting for credentials. They are facts the person already has, and the only question is what explains them. - Think of a doctor who sees that a patient's fever, rash and joint pain fit one disease. The fit is powerful evidence. But the doctor is fitting together symptoms she really observed, not stories she made up. 3. Only a real world predicts the agreement. A world of stable objects leads us to expect the sight, the heat and the crackle to agree. Chance leads us to expect nothing of the kind, and a deceiver can match the agreement only by copying a whole world into its own story. So the agreement raises the probability that experience is connected to a world, in the same way that detailed agreement among witnesses can show that they saw something (D12). The support starts from data the person has, not from beliefs that have none. 4. Coherence as evidence and coherence as the only source are different. Every foundationalist uses coherence as evidence. Coherentism says something much stronger: that coherence is the only source of justification, so that even the beliefs reporting experience have no standing of their own (P1). The first claim is true, and the second is what (P2) denies. See also: Philosophy / Epistemology: Skepticism
+ Donald Davidson: belief is by nature mostly true. To interpret anyone's beliefs at all, we must take most of them to be true, so any coherent set of beliefs is mostly true.
1. Davidson's argument is original. Davidson argued that to understand what someone believes, an interpreter must assume that most of their beliefs are true, since otherwise there would be no way to fix what their words mean. He even imagined an all-knowing interpreter, who would find most of anyone's beliefs true. If he is right, coherence among beliefs comes with a built-in tie to truth. 2. The argument is widely disputed. Many philosophers doubt that the interpreter's need to assume truth shows that beliefs actually are true. A need of the interpreter is a fact about interpreting, not about the world. - Think of a translator who must assume a letter makes sense in order to translate it at all. That does not make the letter true. 3. "Mostly true" does not say which beliefs are true. Even if most of a person's beliefs are true, coherence gives no way to sort the true ones from the false ones. Two coherent systems that differ on many points (P2) would both be "mostly true" in Davidson's sense, and the question of which details to trust would still be open. - Think of a road atlas that is right about almost every road. That is reassuring, but it does not tell you whether the bridge you are about to cross is one of the few mistakes.
+ Susan Haack's foundherentism takes the best of both. Experience counts as evidence, and beliefs support each other like the entries in a crossword puzzle, so neither foundations nor coherence alone is the answer.
1. Haack's crossword is a helpful picture. Haack pictured justification as a crossword: an entry is supported both by its clue and by the entries that cross it. She held that experience gives evidence of its own, and that beliefs also support one another. She called the view foundherentism. 2. The clues are foundations. In a crossword, the clues come from outside the grid. They are what keep the whole puzzle from being any set of words that happen to cross. Haack's experiential evidence plays exactly that role, and so her view grants the main point of (P2): coherence cannot work without evidence from outside the web. 3. The remaining disagreement is about certainty. Haack denies that experiential evidence must be certain, and she rejects the label "foundationalist" for her view. Classical foundationalism holds that the deepest evidence is certain, and that probability needs something certain to rest on. That is a real disagreement. But on the question at issue, whether coherence alone can justify, she stands with the foundationalist. See also: Philosophy / Epistemology: Classical Foundationalism
+ Christian apologists have used coherence as their test of truth. Gordon Clark and Edward John Carnell argued that Christianity is true because it is the most consistent system, so a Christian should not reject coherence.
1. These apologists are serious thinkers and brothers in the faith. Gordon Clark held that logical consistency is the test of truth, and that the Christian system, taken from Scripture, passes it while its rivals fail. Edward John Carnell proposed systematic consistency: a system is true if it is logically consistent and fits the facts of experience. 2. Carnell's test already needs evidence from outside the system. Carnell's second requirement, fitting the facts, does the work that experience does on a foundationalist view. A system that fits the facts is supported by those facts, not by its internal consistency alone. Carnell's test sits closer to foundationalism than to pure coherentism. - Think of a map. We trust it not because its roads join up neatly but because someone checked them against the ground. 3. Clark's system starts from an axiom. Clark took Scripture as the starting point of Christian thought, the axiom (a first principle not proved from anything prior) from which its doctrines are drawn, and he judged rival systems by whether they could stay consistent. A starting point that everything else rests on is a foundation. The disagreement with Clark is over what the foundation is and how a person comes to hold it, not over whether there is one. - Think of geometry. Its theorems agree with one another because every one of them is drawn from the same few axioms. 4. Consistency is necessary but not enough. Every true system is consistent, but so is a well-made fiction. Christianity makes claims about history: that Jesus lived, died, and rose. Those claims are true or false because of what happened, and the evidence for them is what witnesses saw. The apostles spoke as witnesses: "This Jesus God raised up, and of that we all are witnesses" (Acts 2:32). - Clark held that the senses cannot give knowledge in the strict sense, so he would not accept this reply. But the apostles did not invite their hearers to test the resurrection by consistency alone. They offered what they had seen. See also: Philosophical Theology / Prolegomena: Meta-Apologetics: Presuppositional Philosophical Theology / Prolegomena: Meta-Apologetics: Evidentialist
+ The coherence theory of truth: if truth simply is coherence, as Blanshard held and Bradley is usually read as holding, the isolation objection disappears. A coherent system cannot be cut off from a truth that consists in coherence.
1. The move is consistent. If truth were coherence, then a perfectly coherent system would be true by definition, and there would be nothing for it to be isolated from. 2. Most coherentists reject it. BonJour, Lehrer and most coherentists about justification hold that truth is a matter of matching the way things are. They wanted coherence to be evidence of that kind of truth, which is why the isolation objection troubled them in the first place. 3. The theory makes too many things true. Suppose two systems are equally coherent and contradict each other (P2). The coherence theory must then count both true, or say that neither is, and neither answer fits what we mean when we ask whether something really happened. - Think of two witnesses whose accounts of a car accident are each perfectly coherent and contradict each other. Nobody thinks both are true. One of them is describing the accident, and the other is not.
+ The foundations classical foundationalism requires are too thin. Beliefs like "this is like this" carry too little to support anything, so coherence is the only workable source of justification.
1. The worry is a serious one. Michael Bergmann and others have argued that beliefs about present experience are too meager a basis for substantial conclusions about the world, the past, and other minds. Bergmann himself draws an externalist moral rather than a coherentist one. Externalism is the view that what justifies a belief can include facts the believer has no access to, such as whether her eyes are in fact reliable. Still, a coherentist can borrow the complaint. 2. Thin foundations can support a great deal by reasoning. A geometry textbook starts from a handful of axioms, starting statements accepted without proof, and from them proves hundreds of theorems. The foundations are few and simple, and yet what they support is vast. Experience is far richer than a handful of axioms, and reasoning to the best explanation of it reaches the world, the past, and other minds. 3. Coherentism does not escape the problem by moving it. If foundations are too thin to support the world, a web with no foundations at all has even less to work with. Coherence multiplies what is there (P2). Where nothing is there, it has nothing to multiply. See also: Philosophy / Epistemology: Classical Foundationalism
+ Independent witnesses who saw nothing would almost never agree by chance. So when stories agree in detail, the agreement itself creates credibility, even if no witness had any to begin with.
1. Chance agreement really is unlikely. Ask ten strangers who never met to describe a stolen car. If none of them saw it, they will almost never name the same make, color, and plate. Coherentists have pressed exactly this against Lewis: agreement that detailed cries out for an explanation, and the truth is the obvious one. The first half of that should be granted outright. 2. Agreement points to a cause. When detailed agreement is not a fluke, something produced it, and there are only a few candidates: - the witnesses saw the thing they describe; - they compared notes, or someone coached them; - they drew on one shared source, such as the same rumor or the same news report. Only the first makes the story more likely to be true. And the first is a witness with credibility of his own, a real connection to the facts. If there was a real chance, to begin with, that the witnesses saw something, their agreement can show that they did. Agreement can reveal that connection. It cannot conjure it. 3. Witnesses known to have seen nothing leave the story where it was. If we know that none of the witnesses had any access to the facts, the first explanation is ruled out. What remains, a fluke, collusion, or a shared source, is just as available whether the story is true or false. Their agreement then tells us something about the witnesses and nothing about the car. That is the precise sense in which agreement multiplies credibility and cannot create it (P2). - Think of three students who hand in identical answers to a question none of them studied. The teacher does not conclude that the answer is right. She concludes that they sat together. 4. Pure coherentism puts every belief in the unknown witness's place. On pure coherentism, no belief is credited with any access to the facts apart from the others (P1). Each belief is like a witness about whom nothing is known, not where he stood and not whether he could see. - A coherentist can reason, "my beliefs agree, and the best explanation is that they came from contact with the world," and that is good reasoning. - But look at what the reasoning starts from. It treats where the beliefs came from, and the experiences behind them, as facts the believer already has in hand. Facts held with standing of their own are foundations (D3, D6). - Think of the stolen car again. The agreement shows the witnesses saw it only because we already know they were standing on the street when it was taken.

Metaphysics: Free Will

Whether Our Choices Are Up to Us

Public Theology

Theological Analyses of Societal Issues

Theology of the Family

Where Faith and Family Intersect

Biographies

Notable Works & Great Quotes from Key Figures

Contemporary Voices

1960 AD – present
Playlist
Justified True Belief