(●)Richard Swinburne holds that things are probably as they seem, unless there is reason to think otherwise.
+You wake in the night and hear rain on the roof. You do not reason your way to the rain. You do not first check that your ears are working, or try to recall how often your hearing has been right before. You hear it, you believe it is raining, and you are right.
Richard Swinburne, the Oxford philosopher and one of the most important Christian philosophers of the last half century, built a theory of knowledge on that ordinary moment. He calls its central rule the principle of credulity.
- As Swinburne states it, the principle is that, in the absence of special considerations, if it seems to a person that something is present, then probably it is present. Put more simply: things are probably as they seem.
- Philosophers call that kind of experience a seeming: the way something presents itself to you as being so. The drumming that seems to be rain is one. So is the figure across the street that seems to be your neighbor, or the memory that seems to be of locking the back door.
- Credulity here does not mean gullibility. It means a willingness to believe, as a starting point, what experience presents. The name is older than Swinburne: Thomas Reid used it in 1764 for our natural readiness to believe what other people tell us.
- Swinburne calls it a basic principle of rationality, one he does not derive from anything else. Anyone who refuses it, he argues, could never get started knowing anything at all.
The "special considerations" are the cases where that starting presumption gives way. Swinburne lists four kinds:
- the person or the circumstances are unreliable, as with someone who has taken a hallucinogenic drug;
- similar experiences in similar conditions have proved false before;
- background evidence makes it very probable that the thing was not there;
- the thing, even if it was there, probably did not cause the experience.
If none of these applies, the seeming stands.
(●)Swinburne uses the principle to ground perception, testimony and religious experience.
+The principle does three jobs in Swinburne's work.
- Perception and memory. In his 2001 book on justification, the principle is what makes some beliefs basic, which means reasonable without being inferred from (reasoned out of) other beliefs. On his account, a basic belief that a person holds, or is inclined to hold, starts out with a probability that matches how strongly he holds it or is inclined to. A faint patter that might be rain starts lower than a downpour drumming on the roof. Timothy McGrew's review of the book (2002) identifies the principle as the engine of Swinburne's foundationalism, his account of which beliefs hold up all the rest.
- Testimony. Swinburne pairs it with a principle of testimony: in the absence of special considerations, what other people report about their experiences is probably so. That is how a person can know about places he has never been and events he never saw.
- Religious experience. In the second edition (2004) of his book on the existence of God, the principle carries his argument from religious experience. If it seems to someone that God is present, then probably God is present, unless there are special considerations against it. Since millions of people report such experiences, Swinburne argues, they add real weight to the cumulative case for God.
On two large questions about knowledge, Swinburne stands on the same side as Timothy McGrew.
- Both are internalists. They hold that what justifies a belief lies in the believer's own mind, in his experiences and his reasons. McGrew holds this without qualification. Swinburne allows that the word "justified" has externalist senses too, though his sympathies are broadly internalist.
- Their opponents are the externalists, such as Alvin Plantinga. On the externalist view, what makes a belief good enough for knowledge (Plantinga's word for it is warrant) can include facts the believer cannot see from the inside, such as whether the faculty that produced the belief is in fact working properly.
- Both are also evidentialists. They hold that a belief is justified by how well the evidence supports it, and both measure that support by probability. A juror's confidence should rise and fall with the strength of the case, and so should ours.
- So the quarrel between Swinburne and McGrew is narrower than their quarrel with Plantinga. It is about where knowledge begins: where the foundations lie, and what a seeming can do on its own.
(●)Classical foundationalism answers that a seeming is evidence, but what it shows must be inferred.
+Foundationalism pictures knowledge as a building. Most of what we know rests on other things we know, but somewhere there has to be a foundation: beliefs that hold up the rest without resting on anything themselves. McGrew's version, classical foundationalism, is set out in his 1995 book on the foundations of knowledge and in a 1999 essay defending the view. It adds one demand, that the foundations be certain. The reason is that whatever is only probable must be probable on something, and that chain has to end somewhere sure (P2).
- What is certain is the experience itself, taken in by a belief like "this is how it sounds." Such a belief contains the experience it is about. It does not reach past the experience to something else; what it describes is right there inside it, and so it cannot be false. Think of a toothache: you can be wrong about whether you have a cavity, but not about whether it hurts.
- Beliefs about the world, such as "it is raining," are not basic. They are reached from the experience by probable reasoning, chiefly by inference to the best explanation, the detective's method of asking which account fits all the clues best. Rain on the roof best explains this sound, at this hour, in this season. That is also why such beliefs can be wrong while the experience cannot. They are conclusions, and conclusions can miss.
- That reasoning rests on principles of probability that can be known by reflection, the way you know that two plus two is four without counting anything. Timothy and Lydia McGrew defend those principles in their 2007 book on internalism.
On this view, the principle of credulity gets something right and something wrong. It is right that experience is evidence, and right that we are entitled to rely on it. Where it goes wrong is in saying that a seeming, all by itself, makes what it presents probable.
The disagreement matters most for religious experience. On Swinburne's view, the experience of God's presence is itself a reason to think God is present. On McGrew's view, the experience is evidence, and God's presence is a conclusion to be reasoned to, alongside everything else a person knows.
The argument is deductive. Its premises work like a locked chain: if each link holds, the conclusion cannot slip out. And no premise has to be certain. Each one need only be more plausible than its denial.
(P1)The principle of credulity says a seeming supports what it presents without an independent reason to trust it.
+The principle is a rule about where to start. A seeming counts the moment it arrives, and only a reason against it can take that weight away.
(1) The principle asks for nothing before the seeming counts.
- On Swinburne's statement of it, the seeming makes what it presents probable "in the absence of special considerations." To support a belief is to make it more probable than it would otherwise be.
- An independent reason to trust a seeming is one that comes from outside the seeming: a track record, a check against the other senses, an explanation of why seemings like it tend to be right. The principle asks for none of these. All the person needs is to lack a reason against.
- Think of a bank clerk who cashes every check unless something makes him suspicious. He does not look up the account first. He simply waits for a red flag.
(2) Swinburne says the principle cannot be derived from anything else.
- He presents it as a basic principle of rationality, not as a conclusion drawn from the track record of our senses. Any attempt to show that perception is reliable, he argues, would itself have to rely on perception, so the principle has to come first.
- Think of trying to check a tape measure using only that same tape measure. Every test would come out perfect, and it would prove nothing.
- The atheist philosopher Michael Martin argued the opposite: that the principle is not basic, and holds only where a track record backs it. Swinburne's principle, by contrast, asks for no track record at all.
(3) Only reasons against can defeat the seeming.
- All four of Swinburne's special considerations are reasons to doubt: an unreliable observer, a record of failure, strong evidence against, or a cause that is not the thing itself. None of them is a requirement that the person have a reason for trusting the seeming in the first place.
- The burden of proof, the job of making the case, runs one way. It works like the presumption of innocence in a courtroom, where the prosecution must prove guilt and the defendant never has to prove he is innocent. Under the principle, the seeming is the defendant.
(4) Background knowledge does not change the rule.
- Real people always bring background knowledge to a seeming. So in practice, as defenders of Swinburne rightly point out, no seeming is trusted with nothing behind it (D9).
- But on the principle, the seeming's weight does not come from that background. The weight is there first, and the background can only add to it or take some away. That matters most where the background has nothing to say, as with a new kind of experience, and above all an experience of God's presence.
- A clerk with a thick file on most of his customers still cashes a stranger's check by the same rule.
Therefore, on the principle of credulity, a seeming supports what it presents unless something counts against it, with no independent reason needed to trust it.
(P2)No seeming supports what it presents without an independent reason to trust it.
+A seeming, taken as an experience, is something a person can be certain of, and it is real evidence. What it supports beyond itself, though, depends on what else the person knows.
(1) A seeming cannot vouch for its own source.
- A jealous husband watches his wife laugh at a stranger's joke, and it seems to him, vividly, that she is flirting. Philosophers call this cognitive penetration: fear, desire or expectation shaping how things seem. From the inside, it can feel exactly like seeing clearly.
- He can be certain of his experience, and it is evidence of something: that she laughed, and that he is afraid. Whether it is evidence that she was flirting depends on the one thing the seeming cannot tell him, which is whether seemings like it tend to be right.
- Swinburne holds that reason alone favors trusting seemings (P1). Here is the trouble. A truth of reason, like two plus two making four, holds however minds are built. Whether seemings tend to be right depends on how minds actually were built, and a fact like that has to be learned.
(2) A probable belief must be probable on something.
- A probability is always a probability given something. Rain is likely given the drumming and what you know about weather. If those were only probable too, and what they rest on only probable, and so on all the way down, nothing would ever be settled. As C. I. Lewis put it, "if anything is to be probable, then something must be certain."
- Swinburne's basic beliefs about the world are probable, yet they are neither certain nor inferred from anything certain. So ask what they are probable on, and there are two options. If the belief is probable on the experience, then the experience is the foundation and the belief is inferred from it, which is the classical view. If it is probable on nothing, then nothing makes it probable.
- Laurence BonJour's case of Norman shows what the second option looks like. Norman is a reliable clairvoyant, someone who really can know distant things without using his senses. But he has no evidence for or against such powers, or for or against his having one. One day he simply finds himself believing that the President is in New York City. BonJour argued that Norman is not justified, because nothing is on the belief's side.
- Yet Swinburne's 2001 principle, read as stated, would give Norman's belief a probability matching its strength, since the belief is basic and nothing Norman knows counts against it. Defenders reply that if it truly seems so to Norman, then he does have some reason. Here intuitions divide. But notice what that reply does: it counts a hunch from nowhere as evidence.
- Swinburne's own special considerations let evidence about reliability bear on how much a seeming weighs, since a bad record takes weight away. If the record counts when it is bad, it ought to count when there is none. So where that evidence is silent, as it is for Norman, the even-handed verdict is no weight yet.
(3) Trusting a source in advance lets it certify itself.
- Swinburne is right that checking perception with perception proves nothing (P1). Yet the principle approves exactly that check. The wall looks red, so it is red. It is red and it looked red, so my eyes got it right. After a hundred such cases, my eyes are reliable. Every step is licensed, and the conclusion is as worthless as a tape measure checked against itself.
- Stewart Cohen named this the problem of easy knowledge: knowledge of your own reliability that comes far too cheaply. It besets the principle and its modern cousins, which work the same way: Michael Huemer's phenomenal conservatism (if it seems to you that something is so, you thereby have some reason to believe it) and James Pryor's dogmatism (a perceptual experience gives you reason to believe what it shows, with no prior reason needed to think your perception reliable).
- Defenders have replies, and the debate is live. The classical view escapes it, because it never grants the first step, that the wall's looking red by itself makes it probable that the wall is red. Its burden lies elsewhere, on the longer road from experience to the world (D1).
(4) Ordinary perception has the reasons it needs.
- None of this ends in doubt about everything. Ordinary perception earns its independent reason by probable reasoning from what a person is certain of. A real world that perception tracks is the best explanation of why his experiences hang together (D1).
- The regress, the chain of reasons behind reasons, stops at what is certain. That means present experience, taken in by a belief that contains it, and self-evident truths, the ones you see to be true just by understanding them, like two plus two making four. These present nothing beyond themselves, so there is nothing further to trust them about.
- You trust a new thermometer once it agrees with the others in the house and with the frost on the window. Not before.
Therefore, no seeming supports what it presents without an independent reason to trust it.
(C)Therefore, the principle of credulity is false.
+The principle of credulity, as Swinburne states it, is false, because no seeming supports what it presents unless the person has an independent reason to trust it. The conclusion follows by modus tollens, Latin for "the way of denying," which runs: if A, then B; not B; therefore, not A. If the oven is on, the kitchen is warm; the kitchen is not warm; so the oven is not on. Here, if the principle were true, a seeming could support what it presents without an independent reason to trust it (P1). No seeming can (P2). So the principle is not true as Swinburne states it.
(1) Experience is still evidence.
- The experience itself, taken as an experience, is something a person can be certain of. From it, together with what else he knows, he can reason to a great deal.
- Nobody is being told to doubt the rain on the roof. The belief that it is raining is justified. What justifies it is an inference from the sound and from everything the person knows about sounds and weather, not the bare seeming.
(2) What lies beyond the experience is weighed, not granted.
- Beliefs about the world are not basic. They are conclusions, held with the probability that the person's total evidence gives them.
- Testimony is evidence too, and it is weighed the same way. The McGrews' own work on miracles treats testimony as evidence whose force depends on how much more likely the report would be if it were true than if it were false. A text from your spouse saying the milk is gone is far more likely if the carton is empty than if it is not, which is why you stop at the store.
- Religious experience is evidence as well. The experience is real, and certain as an experience, but what it shows about God has to be weighed with everything else a person knows rather than granted at once.
(3) Swinburne and McGrew still stand together.
- Both hold that justification lies in the believer's own evidence. Both measure it by probability. Both think the evidence for God is strong.
- They are like two builders who agree on the shape of the house and differ over what goes under the floor. The disagreement is about the first step. It changes how strong the case from religious experience is on its own, not whether there is a case.
Therefore, if (P1) and (P2) hold, the principle of credulity is false, and seemings are evidence whose weight must be shown, not presumed.
See also:
Philosophy / Epistemology: Classical FoundationalismPhilosophy / Epistemology: InternalismPhilosophy / Epistemology: Warrant
Swinburne's view:
Swinburne, Richard. Epistemic Justification. Oxford: Clarendon Press, 2001.
Swinburne, Richard. The Existence of God. 2nd ed. Oxford: Clarendon Press, 2004.
Swinburne, Richard. Is There a God? Rev. ed. Oxford: Oxford University Press, 2010.
McGrew's view and critique:
McGrew, Timothy J. The Foundations of Knowledge. Lanham, MD: Rowman & Littlefield, 1995.
McGrew, Timothy. "A Defense of Classical Foundationalism." In The Theory of Knowledge: Classical and Contemporary Readings, 2nd ed., edited by Louis P. Pojman, 224-235. Belmont, CA: Wadsworth, 1999.
McGrew, Timothy. Review of Richard Swinburne, Epistemic Justification. Notre Dame Philosophical Reviews, 2002.09.10.
McGrew, Timothy, and Lydia McGrew. Internalism and Epistemology: The Architecture of Reason. London: Routledge, 2007.
McGrew, Timothy, and Lydia McGrew. "The Argument from Miracles: A Cumulative Case for the Resurrection of Jesus of Nazareth." In The Blackwell Companion to Natural Theology, edited by William Lane Craig and J. P. Moreland, 593-662. Chichester: Wiley-Blackwell, 2009.
Related views and critics:
Alston, William P. Perceiving God: The Epistemology of Religious Experience. Ithaca, NY: Cornell University Press, 1991.
Bergmann, Michael. Justification without Awareness: A Defense of Epistemic Externalism. Oxford: Oxford University Press, 2006.
BonJour, Laurence. "Externalist Theories of Empirical Knowledge." Midwest Studies in Philosophy 5 (1980): 53-73.
Coady, C. A. J. Testimony: A Philosophical Study. Oxford: Clarendon Press, 1992.
Cohen, Stewart. "Basic Knowledge and the Problem of Easy Knowledge." Philosophy and Phenomenological Research 65 (2002): 309-329.
Fricker, Elizabeth. "Against Gullibility." In Knowing from Words, edited by Bimal Krishna Matilal and Arindam Chakrabarti, 125-161. Dordrecht: Kluwer, 1994.
Huemer, Michael. Skepticism and the Veil of Perception. Lanham, MD: Rowman & Littlefield, 2001.
Huemer, Michael. "Compassionate Phenomenal Conservatism." Philosophy and Phenomenological Research 74 (2007): 30-55.
Jeffrey, Richard C. The Logic of Decision. 2nd ed. Chicago: University of Chicago Press, 1983.
Jeffrey, Richard C. Probability and the Art of Judgment. Cambridge: Cambridge University Press, 1992.
Lewis, C. I. An Analysis of Knowledge and Valuation. La Salle, IL: Open Court, 1946.
Martin, Michael. Atheism: A Philosophical Justification. Philadelphia: Temple University Press, 1990.
Pryor, James. "The Skeptic and the Dogmatist." Noûs 34 (2000): 517-549.
Reid, Thomas. An Inquiry into the Human Mind on the Principles of Common Sense. 1764.
Russell, Bertrand. The Analysis of Mind. London: George Allen & Unwin, 1921.
Siegel, Susanna. "Cognitive Penetrability and Perceptual Justification." Noûs 46 (2012): 201-222.
+Swinburne's skeptical bog: without the principle, no one could ever justify trusting his senses, since every check on perception uses perception. Rejecting the principle leaves us knowing nothing.
1. The worry is the principle's best argument.
If every reason to trust perception had to come from perception, a person would face a grim choice: trust his senses without a reason, or never trust them at all. Swinburne is right that the second option is a dead end. That is why the principle has seemed to so many philosophers to be the only way out.
2. The way out runs through the experience itself.
The classical foundationalist does not check perception with perception. He starts somewhere else, with what he is certain of: his experiences, taken simply as experiences. Then he asks what best explains their order. Why do they come in stable patterns? Why do they agree with each other, and why do they let him predict what comes next? The answer, a real world that ordinary perception tracks, is reached by reasoning from the experiences, not by trusting them in advance.
- Think of a detective who never saw the crime. He does not need to have watched it to know who did it. He reasons from the traces.
3. The reasoning uses principles known by reflection.
The principles of probable reasoning are not learned from the senses. Take the rule that a hypothesis (a proposed explanation) is supported by evidence it makes more likely than its rivals do. Muddy paw prints on the kitchen floor support "the dog came in" over "nobody came in," because a dog coming in would leave them and an empty kitchen would not.
- On the McGrews' view such principles are known a priori, Latin for "from what comes first," which means by reflection alone. It is the way you know that if Anna is taller than Ben and Ben is taller than Cal, then Anna is taller than Cal, without measuring anyone.
- So the reasoning from experience to world does not quietly assume what it is trying to show.
4. The regress stops at what needs no trust.
You might ask whether the reasons need reasons too, and so on forever. They do not, because they are built from things that present nothing beyond themselves: present experience, taken in as it is, and self-evident truths seen by reflection (P2). There is nothing further to trust them about.
- Nobody needs a reason to trust that his stubbed toe hurts. The hurt is right there.
5. Whether the certain base is enough is the real question.
The strongest form of the worry grants all of this and says the base is too thin. Michael Bergmann presses the point: beliefs as bare as "this is like this" are too meager to support substantial conclusions about the world. If so, the classical view sinks into the same bog by a longer road. Memory raises the same question in its sharpest form (D11).
- Classical foundationalism answers that the base is thin in words, not in content. A belief that contains the experience it points at holds every shade of that experience, and conclusions rest on a branching tree of such beliefs, never on one.
- Think of a jigsaw puzzle. No single piece shows the picture, and yet a thousand pieces fitted together leave no doubt what it is.
- Whether that reasoning fully answers the skeptic is still argued. But the principle does not answer him either. It declares the matter settled at the start, when the skeptic's question is precisely whether it is.
See also:
Philosophy / Epistemology: Classical Foundationalism
+Anyone who rejects the principle does so because it seems to him to be false. So the critic relies on the principle in the act of denying it.
1. This is a serious argument.
Michael Huemer has pressed a version of it for phenomenal conservatism, the view that if it seems to you that something is so, you have some reason to believe it. If every belief rests on how things seem, then rejecting the principle is itself a belief that rests on a seeming. The rejection undercuts itself, like a man who shouts that he cannot speak.
Its sharpest form points at the case against the principle itself. The verdicts on Norman and the jealous husband (P2) are intuitions, and intuitions are seemings. So the critic trusts seemings with no independent reason, in the very act of saying no one may.
2. Seeing that something is true is not a seeming in the principle's sense.
When a person sees by reflection that two plus two is four, or that a conclusion follows from its premises, he is not trusting an appearance that could be quite unconnected with the truth. He is directly aware of the truth itself. Classical foundationalism counts such self-evident truths among the certain foundations.
- Seeing that a key fits a lock is not the same as having a feeling that it will.
- Whether every good judgment can be traced to seeing of that kind is the point in dispute, and phenomenal conservatives deny it.
3. The judgments about the cases are judgments about evidence.
To judge that Norman is not justified is to see that a belief with nothing on its side has not been made probable by anything. That is a truth about probability. On the McGrews' view, truths of that kind are analytic, true because of what probability and evidence are, and they are known by reflection, the way a person sees that a conclusion follows from its premises. The case against the principle is made of judgments of that kind (P2), not of impressions.
4. The argument would reach only truths of reason.
Suppose the self-defeat argument worked. It would show at most that some seemings about truths of reason carry weight without an independent reason. It would do nothing for seemings about the world (the rain, the neighbor, the presence of God), which is where the principle does its distinctive work.
- A man who trusts his arithmetic has given no reason to trust his horoscope.
+Ordinary people and small children trust their senses without any inference at all. The principle describes how real people come to know, and the inferential picture describes no one.
1. The worry is fair.
A three-year-old who hears thunder and runs to the window has never heard of a best explanation. If her belief that a storm is coming were justified only by reasoning she could state, she would be justified in nothing, and that is absurd. Critics press this charge of over-intellectualizing against every internalist view, and it deserves an answer.
2. Inference need not be slow or spoken.
Reasoning to the best explanation can be quick, habitual and unspoken. A driver who brakes at a red light does not recite a syllogism, a formal argument laid out premise by premise, but his braking follows from what he sees and what he knows. In the same way, a person's belief that it is raining rests on the sound and on what he knows, even when he never puts it into words.
3. Internalism asks for reasons the person has, not reasons he can recite.
On internalism, the view Swinburne and McGrew share, a belief is justified by what is present to the believer's mind: his experiences, his memories, what he has learned. A child has a great deal of that, and her beliefs rest on it, though she could not explain it to a philosopher.
4. The principle asks no less of the child.
Swinburne's own account assigns each basic belief a probability matched to its strength, and weighs it against special considerations. No child recites that either. Both views describe what makes a belief reasonable, not the steps a believer consciously takes.
- A child speaks in grammatical sentences years before she could name a single rule of grammar.
See also:
Philosophy / Epistemology: Internalism
+The jealous husband is already covered. His jealousy is a special consideration, so the principle does not give his seeming any weight.
1. If he knows his own jealousy, both views agree.
Swinburne frames his special considerations as evidence: a subject or conditions found unreliable before, similar claims shown false, background evidence against. A husband who knows his temper, and how it bends what he sees, has exactly that kind of evidence. On the principle, his seeming loses its weight. On McGrew's view, it never gains any. Either way the verdict is the same.
- A man who knows his bathroom scale reads five pounds light does not need a philosopher to tell him what to make of the number.
2. If he does not know, internalism judges him by what he has.
Swinburne's sympathies are broadly internalist, and on that reading a special consideration defeats a seeming only when it is part of the person's own evidence. An onlooker who knows the husband's temper may rightly discount his report. For the husband's own justification, though, what counts is what he has.
Suppose the principle were changed so that a seeming produced by a bad process counts for nothing, whether or not the person knows it. Then what justifies his belief would depend on facts outside his mind. That is an externalist sense of "justified," which McGrew rejects outright. Swinburne allows that the word has such senses, but his principle, as he states it, turns only on how things seem to the person and what else he knows.
- Picture two men with exactly the same experience. One man's eyes are healthy. The other's are failing in a way neither of them could detect. On internalism they are equally justified, because from the inside nothing separates them.
- So on both views, a husband who does not know his own jealousy, and whose evidence all speaks for seemings like his, may be justified in a belief that is false. Internalists accept that cost.
3. The husband shows why trust has to be earned.
On its own, the case does not separate the two views. What it shows is the first point of (P2): a seeming shaped by fear can feel exactly like one shaped by sight, so a seeming cannot vouch for its own source, and its link to the truth has to be learned. The views part company where a person's evidence is silent both ways, as Norman's is (P2). There the principle gives the seeming weight, and the classical view does not.
- Two letters in the same handwriting, one of them forged, read exactly alike. Whether to trust either is settled by what you know about where it came from, not by the ink.
+Without the principle, religious experience counts for nothing, and millions of believers who have known God's presence are left with no evidence.
1. Religious experience still counts.
On McGrew's view the experience is real evidence. A person who has experienced what seemed to be God's presence knows that with certainty, as an experience, and that fact has to be explained. Its force depends on how much more likely such an experience is if God exists than if he does not, and on how it fits with everything else the person knows.
2. There is no double standard.
The strongest form of the objection, pressed by Swinburne and by William Alston, is a charge of inconsistency. A critic who trusts his senses without independent checks, and then demands such checks for religious experience, applies one rule to the world and a stricter one to God. That charge is fair against such a critic. It does not touch the classical view, which asks the same of sense perception as of religious experience: each earns its weight by reasoning from experience.
- A referee who checks the replay for both teams is not biased against either.
3. The cost should be admitted.
On McGrew's view, a single experience taken alone does less than it does on Swinburne's. A believer whose only evidence was one such experience would have less than the principle promises. Most believers are not in that position, because their experience sits inside a larger body of evidence.
4. Inference sorts the experiences.
Swinburne's principle gives starting weight to every religious seeming: the Christian's sense of Christ's presence, the Hindu's experience of Brahman, the Buddhist's experience of emptiness. Those cannot all be veridical, which means true to how things really are, so the principle must sort them afterward by special considerations. Weighed as evidence from the start, they can be compared: what they report, how they fit with the rest of what is known about God, how they change the people who have them, and whether they point to a God who has acted in history.
- A judge who believes every witness until contradicted soon finds himself believing both sides.
- Michael Martin pressed a negative principle of credulity: if seeming-present counts for God, then seeming-absent should count against him. Defenders of the principle reply that seeming-absent counts only where a person would expect to perceive the thing if it were there, and a God who chooses when to make himself known need not be perceived on demand. The reply has real force. But notice how it works, by appeal to what a person knows about God. That is the classical point: the weight of a seeming, for or against, is set by the background.
- Weighing is a stronger position for a Christian, not a weaker one, because the Christian experience of God fits into a case with other evidence behind it.
See also:
Christian Evidences / Resurrection of Jesus: Maximal Data Method (McGrew)
+Almost everything we know comes from testimony. If we needed a reason to trust each source first, we could know almost nothing, so a principle of testimony like Swinburne's is unavoidable.
1. The debt to testimony is real.
Thomas Reid and, more recently, C. A. J. Coady argued that we could not check most of what we believe for ourselves. Coady's strongest point is that the reasons we might give for trusting testimony, such as the record of newspapers and encyclopedias, are themselves known mostly through testimony. So no one could back testimony as a whole with his own observations. That should be granted: a check of all testimony, from one person's own eyes and ears, is out of reach.
2. The reasons needed are local.
We do not need to vouch for testimony as a whole before believing anyone. What we need are reasons to think that this kind of report, on this kind of subject, from this kind of source, tends to be right, and Elizabeth Fricker has argued that a hearer usually has them. We have a great deal of that: the agreement of independent reports, the way testimony fits with what we see for ourselves, and what we know about when people are in a position to know and have no motive to mislead.
- You trust a stranger who gives you directions in his own town because you know how often locals are right about their own streets.
3. Independent agreement does the heavy lifting.
In the McGrews' work on miracles, the force of a report depends on how much more likely it would be if it were true than if it were false, and independent reports multiply that force. That is how testimony can establish even surprising facts, without a principle that grants it weight in advance.
- A neighbor's call saying your garage is on fire is far more likely if it is burning than if it is not. A second call, from someone across the street who has not spoken to the first, makes the case much stronger than one call twice as loud.
- How a small child earns her first trust in testimony is a hard question on every view. The classical answer is that it is earned quickly, from the steady match between what she is told and what she then sees: the cookies are in the jar, just as her mother said.
+Probability does not need certainty at the foundation. A seeming can raise a probability directly, as Richard Jeffrey's rule for uncertain evidence shows.
1. The mathematics is sound.
The philosopher Richard Jeffrey gave a rule that tells a person how to update his other beliefs when an experience shifts his confidence in something without making it certain. It is a real tool, and Jeffrey drew the bold conclusion himself: on the view he called radical probabilism, judgment never needs to rest on anything certain.
- Jeffrey's own illustration is a piece of cloth seen by candlelight. The glimpse leaves you unsure of its color, but more confident than before that it is green, and the rule tells you how that change should spread to everything that depends on the color.
2. The rule does not say where the new confidence comes from.
If a glimpse of cloth by candlelight moves a person's confidence that it is green to seventy percent, something must justify seventy rather than thirty. Jeffrey's rule takes the shift as given. It passes a change along and does not create one, so it leaves Lewis's point standing (P2).
The natural source of the shift is the experience itself, which the person is certain of, together with what he knows about candlelight and colors. That is the classical foundationalist's answer. It is not the principle of credulity, which would give the seeming its weight without the background.
See also:
Philosophy / Epistemology: Classical Foundationalism
+The principle of credulity and classical foundationalism agree on so much that this is a quarrel over details. Both are internalist and evidentialist views, and both are used to defend the case for God.
1. The agreement is real.
Both hold that the justification that matters, for a person deciding what to believe, lies in his own evidence and is measured by probability. Both stand against the externalism of Alvin Plantinga and Alvin Goldman on that question, though Swinburne allows externalist senses of "justified" that McGrew does not. And both think the evidence for Christianity is strong.
2. The detail is the foundation.
Where knowledge begins decides what counts as evidence and how much it counts. On the principle of credulity, every seeming starts with weight. On classical foundationalism, only what is certain starts with weight, and everything else must be earned. That changes how religious experience, testimony and perception enter the case for God.
3. The difference changes what a believer can offer a skeptic.
A case that rests on what a person is certain of, and on reasoning anyone can check, can be laid before a skeptic step by step. A case that starts from the principle asks the skeptic first to grant a rule he may doubt, and then to apply it to experiences he has never had. The classical case may give religious experience less weight on its own (D5), but every step in it can be checked.
- It is the difference between an accountant who shows you the receipts and one who asks you to trust his sense of the books.
+The principle only sets a starting point, and real people always have background knowledge. In practice no seeming is trusted with nothing behind it, so the quarrel is idle.
1. The fact should be granted.
This is the best reply open to a defender of Swinburne. Adults bring a lifetime of background to every seeming, and most ordinary seemings would pass the test of (P2) anyway. If the principle only ever operated with that background in place, the two views would rarely disagree about the rain on the roof.
2. The principle still puts the seeming's weight first.
On the principle, background adds to a seeming's weight or takes it away. It does not supply the weight. And on Swinburne's account the background itself was built from seemings the principle granted at the outset, so the principle stays at the root of everything above it.
- A building's upper floors can be solid steel and still rest on whatever was poured first.
3. The quarrel matters where the background is silent.
The principle does its distinctive work exactly where no track record exists: a new kind of experience, and above all an experience of God's presence, where the question in dispute is whether seemings of that kind are ever right. That is where Swinburne needs the principle, and where (P2) says a seeming has not yet earned its weight.
- A person's first experience of anything, whether a new instrument or a new sense, comes with no file at all.
4. A starting probability still has to come from somewhere.
If the principle is a claim about what a seeming makes probable before any other evidence, then it says reason alone favors the seeming's being right. That is the claim (P2) disputes.
A defender might answer that the simplest explanation of a seeming is that it is true. But then the principle has become a piece of explanatory reasoning rather than a basic principle, and explanatory reasoning weighs rivals (a dream, a fear, a wish) against everything else the person knows. That is the classical view's method.
The answer also leaves simplicity needing a ground of its own, since something has to say why the simplest explanation should be the likeliest. McGrew's review faults Swinburne's 2001 book on this point, for treating its criteria of simplicity as starting points that philosophy can clarify but not justify.
- A detective does not name the butler before he has heard who else was in the house.
+Every seeming belongs to many kinds, so a "reason to trust seemings like it" has no fixed meaning. Swinburne pressed this generality problem against reliabilism, and it sinks the demand for independent reasons too.
1. The generality problem is real.Reliabilism is the externalist view that a belief is justified if it comes from a process that usually produces true beliefs. It has to say which process counts. The sound of rain at night is heard by ears, by a tired man, through a tin roof, in a storm, and each of those kinds has its own record. Picking the kind that counts is the generality problem. Swinburne is among those who pressed it, and against reliabilism the choice does look arbitrary.
2. Evidence does not need one kind.
On McGrew's probabilistic internalism nothing has to be chosen. The question is how probable the seeming's content is on everything the person knows, and that includes every feature of the seeming and its circumstances he is aware of. Talk of "seemings like it" is shorthand for that whole body of evidence.
- A doctor reading an X-ray weighs the machine, the patient's age and the shape of the shadow together. She does not first file the case under one heading.
3. The problem bites the principle too.
Swinburne's second special consideration asks whether similar experiences in similar conditions have proved false. Similar in what respect? Any view that lets the past count must answer that, and the answer is the same for both: weigh everything you know. Where features pull different ways, the weighing is not mechanical. But that is the ordinary work of judging evidence, not a defect peculiar to the classical view.
+The reasons to trust perception are built from remembered experiences, and memory can be checked only by memory. So the classical foundationalist must trust memory without reasons, exactly as the principle says.
1. Memory is the hardest case.
Every pattern that (P2) appeals to stretches across time, and yet a person is aware only of the present. If the classical view has a soft spot, this is where a critic would press.
2. A memory is a present experience.
When it seems to you that you locked the back door, the seeming of remembering is here now, and you can be certain of it as an experience. Whether the door was locked is what it presents, and that is reached by reasoning.
- You can be certain you remember locking it and still walk back to check. Most people have.
3. The past is the best explanation of present traces.
Present memories agree with each other, with what you see now (the mud on your boots, the receipt in your pocket), with records, and with other people's memories. Bertrand Russell pointed out that nothing logically rules out a world created five minutes ago, complete with false memories and every trace. But such a world can match the evidence only by copying the whole structure of a real past into its own story, and that makes it the worse explanation.
4. The principle does no better.
The principle trusts memory by default, but it still has to sort true memories from false. It does so by the agreement of memories with each other and with traces, which is the classical method. Whether present evidence alone is enough to reach the past by reasoning is still debated, and the classical view claims only probability for it, never certainty.
See also:
Philosophy / Epistemology: Classical Foundationalism