Wednesday, August 25, 2010

The appeal to "prior commitments" or "presuppositions" re. theism

Here's a bit from a paper forthcoming in Faith and Philosophy. I put it up because it concerns a certain move that's often made re evidence of miracles - that whether it's sensible to accept testimony of the miraculous depends on ones "presuppositions" or "prior commitments". This phrase just cropped up in a slightly bad-tempered interchange I am currently having with Glenn Peoples here.

The Ted and Sarah case

Suppose I have two close friends, Ted and Sarah, whom I know to be generally sane and trustworthy individuals. Suppose that Ted and Sarah now tell me that someone called Bert paid them an unexpected visit in their home last night, and stayed a couple of hours drinking tea with them. They recount various details, such as topics of conversation, what Bert was wearing, and so on. Other things being equal, it is fairly reasonable for me to believe, solely on the basis of their testimony, that such a visit occurred.

But now suppose Ted and Sarah also tell me that shortly before leaving, Bert flew around their sitting room by flapping his arms, died, came back to life again, and finished by temporarily transforming their sofa into a donkey. Ted and Sarah appear to say these things in all sincerity. In fact, they seem genuinely disturbed by what they believe they witnessed. They continue to make these claims about Bert even after several weeks of cross-examination by me.

Am I justified in believing that Ted and Sarah witnessed miracles? Surely not. The fact that Ted and Sarah claim these things happened is not nearly good enough evidence. Their testimony presents me with some evidence that miracles were performed in their living room; but, given the extraordinary nature of their claims, I am not yet justified in believing them.

Notice, incidentally, that even if I am unable to construct a plausible explanation for why these otherwise highly trustworthy individuals would make such extraordinary claims – it’s implausible, for example, that Ted and Sarah are deliberate hoaxers (for this does not fit at all with what I otherwise know about them), or are the unwitting victims of an elaborate hoax (why would someone go to such extraordinary lengths to pull this trick?) – that would still not lend their testimony much additional credibility. Ceteris paribus, when dealing with such extraordinary reports – whether they be about alien abductions or supernatural visitations – the fact that it remains blankly mysterious why such reports would be made if they were not true does not provide us with very much additional reason to suppose that they are true.

Consideration of the Ted and Sarah case suggests something like the following moral:

P1 Where a claim’s justification derives solely from evidence, extraordinary claims (e.g. concerning supernatural miracles) require extraordinary evidence. In the absence of extraordinary evidence there is good reason to be sceptical about those claims.

The phrase “extraordinary claims require extraordinary evidence” is associated particularly with the scientist Carl Sagan . By “extraordinary evidence” Sagan means, of course, extraordinarily good evidence – evidence much stronger than that required to justify rather more mundane claims. The phrase “extraordinary claims” is admittedly somewhat vague. A claim need not involve a supernatural element to qualify as “extraordinary” in the sense intended here (the claims that I built a time machine over the weekend, or was abducted by aliens, involve no supernatural element, but would also count as “extraordinary”). It suffices, for our purposes, to say that whatever “extraordinary” means here, the claim that a supernatural miracle has occurred qualifies.

Some theists (though of course by no means all) have challenged the application of Sagan’s principle to religious miracles, maintaining that which claims qualify as “extraordinary” depends on our presuppositions. Suppose we begin to examine the historical evidence having presupposed that there is no, or is unlikely to be a, God. Then of course Jesus’ miracles will strike us as highly unlikely events requiring exceptionally good evidence before we might reasonably suppose them to have occurred. But what if we approach the Jesus miracles from the point of view of theism? Then that such miraculous events should be a part of history is not, one might argue, particularly surprising. But then we are not justified in raising the evidential bar with respect to such claims. So theists may, after all, be justified in accepting such events occurred solely on the basis of a limited amount of testimony, just as they would be the occurrence of other unusual, but non-supernatural, events. The application of Sagan’s principle that “extraordinary claims require extraordinary evidence” to the Jesus miracles simply presupposes, prior to any examination of the evidence, that theism is not, or is unlikely to be, true. We might call this response to Sagan’s principle the Presuppositions Move.

That there is something awry with the Presuppositions Move, at least as it stands, is strongly suggested by the fact that it appears to license those of us who believe in Big Foot, psychic powers, the activities of fairies, etc. to adopt the same strategy – e.g. we may insist that we can quite reasonable accept, solely on the basis of Mary and John’s testimony, that fairies danced at the bottom of their garden last night, just so long as we presuppose, prior to any examination of the evidence, that fairies exist. Those making the Presuppositions Move with respect to religious miracles may be prepared to accept this consequence, but I suspect the majority of impartial observers will find it a lot to swallow – and indeed will continue to consider those who accept testimony of dancing fairies to be excessively credulous whether those believers happen to hold fairy-istic presuppositions or not.

I suspect at least part of what has gone wrong here is that, when it comes to assessing evidence for the Jesus miracles and other supernatural events, we do so having now acquired a great deal of evidence about the unreliability of testimony supposedly supporting such claims. We know – or at least ought to know by now – that such testimony is very often very unreliable (sightings of ghosts, fairies, and of course, even religious experiences and miracles, are constantly being debunked, exposed as fraudulent, etc.). But then, armed with this further knowledge about the general unreliability of this kind of testimony, even if we do happen to approach such testimony with theistic or fairy-istic presuppositions, surely we should still raise the evidential bar much higher for eye-witness reports of religious miracles or fairies than we do for more mundane claims.

{{ENDOTE It may be said that there is a relevant disanalogy between the application of the Presuppositions Move with respect to religious miracles and to fairies. We have now acquired good empirical evidence that there’s no such thing as fairies. Starting off an assessment of the empirical evidence with the presupposition that fairies exist is one thing. Retaining that presupposition in the teeth of empirical evidence to the contrary is quite another. The Presuppositions Move surely requires that we have come across no body of empirical evidence throwing into serious doubt the existence of what we have been presupposing exists. This blocks the application of the Presuppositions Move in defence of accepting testimony regarding fairies. However, while there’s good empirical evidence that there’s no such thing as fairies, there’s no such evidence against the existence of God. Thus the Move can still be made with respect to testimony of religious miracles.

An obvious difficulty with the above suggestion is the evidential problem of evil (for an assessment, see my “The Evil God Hypothesis” in Religious Studies 46 (2010), 353-373). Prima facie there is good empirical evidence that there is no God. In which case, the above suggestion looks to be no less an obstacle to the use of the Presuppositions Move with respect to religious miracles. So, prior to employing the Move, those theists insisting on the above disanalogy will need to come up with an adequate solution to the evidential problem of evil (a solution not dependent on the truth of religious miracle claims) – not an easy task.


So, my suggestion is that P1 is, prima facie, a fairly plausible principle – a principle that is applicable to the testimony concerning the miracles of Jesus. Note that P1 at least allows for the possibility that we might reasonably suppose a miracle has happened. Of course, I do not claim to have provided anything like proof of P1. But it does appear fairly accurately to reflect one of the ways in which we assess evidence. We do, rightly, set the evidential bar much higher for extraordinary claims than we do for more mundane claims.


Richard Baron said...

I think it would be good to say something more about what counts as extraordinary. Perhaps you cover this elsewhere in the paper, and I agree that whatever our definition, miracles have to count as extraordinary. My thought is that our criteria for classifying a claim as extraordinary (which may amount to indicators, rather than a definition) may influence how one defends your principle that extraordinary claims require extraordinary support.

One option would be to use Quine's fabric of beliefs. The more tensions that a claimed observation (something at the edge of the fabric) would create, and the more and deeper adjustments to the fabric that accommodating the claim would require, the stronger the evidence would need to be.

That approach would separate violations of laws of nature, which would require massive adjustments, from remarkable non-violations, like being dealt all 13 cards of one suit in a game of Bridge, which would require hardly any adjustments, and acceptance of which does not demand extraordinary evidence. Furthermore, the dubious nature of the presuppositions that one would need in order to accept reports of miracles without demanding extraordinary evidence becomes obvious, if we demand that those presuppositions themselves be fitted into the fabric of beliefs.

Should one also look at the way in which some presuppositions themselves look more plausible because they are widely believed? There are, I think, more believers in Christianity or in Islam than in Bigfoot. Presuppositions should not gain extra plausibility like this, of course, but I suspect that they do. It would be an application at the level of general, rather than particular, claims, of the principle that "Three men make a tiger".

wombat said...

Regarding the endnote - surely even if we still work from a theist perspective we are in the same position with respect to miracles (including the Jesus ones) inasmuch as even though we accept the existence of the supernatural as given we have a large body of evidence that suggests that miracles are vanishingly rare, and in fact the sort of miracle that can _only_ be explained by divine intervention working to bypass causality, logic or physical law simply has never been observed.

Consider the evidence that the RCC requires for sainthood. Two miracles. Now how many miraculous claims actually pass muster? Well if John Paul II is anything to go by very few.

e.g. from The Times

"More than 250 claims of miracles, mainly regarding alleged healings from cancer, have been made but the one eventually investigated in 2006 was when a French nun, Sister Marie-Simon-Pierre, was cured of Parkinson’s disease.

Doctors declared her cure “scientifically inexplicable” after her sisters prayed for the late Pope’s intercession and she suddenly found herself able to pick up a pen and write his name. “I was sick and now I am cured,” she said at the time.

After studying the 2,000-page positio produced in the investigation by the postulator Father Slawomir Oder, Pope Benedict is expected to sign the decree tomorrow declaring his predecessor venerable, a source confirmed to The Times. "

So 1 in 250 and a 2000 page file of evidence, with presumably some living witnesses and a professional advocate.
Much better quality of evidence on the face of it than for the Biblical miracles. And this is still for a miracle which does not exclude physical causes - the condition is I understand lacking in conclusive diagnostic tests (absent brain biopsy) and such tests as are done very often simply eliminate other causes such as drugs, toxins or stroke. As a theist one might agree that the probability of divine intervention was greater that the probability of the doctors getting it wrong but its very much on the level of "getting a string of 20 consecutive double sixes playing Ludo" rather than changing a sofa into a donkey.

Interestingly of course the Evil God responds to prayers too. Except they get called curses.

Tony Lloyd said...
This comment has been removed by the author.
Tony Lloyd said...
This comment has been removed by the author.
Glenn said...

FWIW - I wasn't bad tempered. I'm sorry if you were annoyed. But look, something good came of it :)

Eric said...

Professor Law, what if we replace the presuppositional account of Glenn's "prior commitments" move with a "good reasons" account instead?

That is, instead of saying theists 'presuppose' god exists, we say they claim to have 'good reasons' for thinking that god exists, and that this, along with testimony from reliable witnesses, and some connection between the god that is said to exist and the purported miracle, raises the likelihood that a miracle occurred.

This account seems to me to weaken considerably the "fairy" (Big Foot, et al) counterexample, for I think that if we thought we had good reasons for thinking that fairies exist, and if we thought we had reliable witnesses, and if we thought there were good grounds for believing that there's some connection between the claim of the witnesses and the nature of the fairies, we'd be justified in believing the witnesses (or, minimally, we'd have better reasons for believing them on 'good reasons' account of prior commitments than the on the 'presuppositional' account).

Tony Lloyd said...

I think the propositions we hold before judging evidence do effect whether a claim is “extraordinary” or not. But in their analysis the theists wrongly identify the propositions that make miracles extraordinary. (Beretta-online is beretta-offline at the moment so I’m reconstructing the theist-commitment argument from memories of other theist-commitment arguments I’ve read).

I would say that an extraordinary claim is one that our theories about the world would lead us expect wouldn’t be the case. Taking our theories about the world as W and the claim as V, an extraordinary theory is a V where “if W then not V”.

“It rained yesterday” is not an extraordinary claim. Our theories predict that rain will happen every so often. “The rain fell upwards” is an extraordinary claim because it is simply incompatible with our picture of the world as one where unsupported things fall down.

Now the theists that claim “prior-commitments” or “pre-suppositions” are obviously invoking theism as a proposition held that stops, say, the resurrection being “extraordinary”. If theism fulfils this function then it’s a fair point on the theist’s part: the atheist’s insistence that the resurrection is “extraordinary” begs the question.

But atheism is not a necessary condition of any miracles being extraordinary neither is theism a sufficient condition to bring them away from being extraordinary. Miracles are extraordinary because they violate the normal theories that we hold whether atheist or theist (Hume). The resurrection is not extraordinary “on atheism” and mundane “on God” but extraordinary on our common theories of life, medicine, how the human body works etc. Both atheist and theist have an understanding of death as an irreversible process which is perfectly adequate to lead us to expect that people do not rise from the dead.

But neither the theist nor the atheist pre-commit to or pre-suppose that death is an irreversible process. “Death is an irreversible process” is just another theory, not a foundational essential without which an entire Weltansachauung will crumble. Using the language of “commitment” and “pre-supposition” sounds like a disingenuous attempt to bring in a worldview relativism “you have your fundamentals, I have mine”.

Martin said...

Has anyone ever asked why the evidence for miracles is always pecularily weak? Surely there's nothing in a miracle as such that says the evidence for it has to be weak. Why don't we get miracles of the type where God's head appears in the sky for a period of a week? In which time there would be plenty of opportunity for the world's press and scientists to assemble and view the miracle for themselves.

It seems we have accepted a narrative about miracles whereby they are always executed in a furtive fashion. Is God ashamed of His miraculous powers?

Max Daygigger said...

I hope you blast Glenn Peoples out of the intellectual universe.

Stephen Law said...

several copies of Lloyds comment have appeared weirdly.

Gleen - no it's me being bad-tempered, not you.....

wombat said...


Surely having good reasons to suppose God exists (provided that it isn't simply as a source of miracles of course) only increases the probability that God's intervention is a possible explanation of events? It still has to compete with all the other possible explanations (simply mistaken witnesses, deception by third parties, random chance, hitherto unforseen consequence of string theory etc.) but at least it is in there as a candidate. It does not really affect the standards to which the evidence should be held does it?

wombat said...

"several copies of Lloyds comment have appeared weirdly."

Miraculous! The internet version of loaves and fishes?

Or perhaps the Gods smote the Blogspot server.

That aside a small nit pick with Tony's post is that is it not possible to have the view that by definition death is irreversible? I would agree that only shifts things a little because we then have the (possibly equivalently improbable) "he was exhibiting all the signs of death for several days before getting up without apparent distress..."

Tony Lloyd said...

I’ve just read Glenn’s “A New Euthyphro” ( . I don’t think that it (or divine command theory) has an awful lot of effect on your Evil God Argument.

The Evil God Argument is against the existence of an all-knowing, all-powerful and maximally good God.

Glenn’s Euthyphro (“GE”) holds that “rightness is that quality of being willed by God”. So God cannot fail to be good (in fact He cannot fail to be all-good: which resurrects the logical problem of evil, but let’s leave that now). But this makes the claim “God is good empty”. It is the tautology and saying it does not denote a property. The Evil God Argument fails not because it is flawed but it is unnecessary to argue against the existence of an entity with properties A, B, and C if there is no property C for it to be.

Some content to “God is good” is given by GE when replying to Socrates’ complaint that if God willed torture was right then it would be right, but torture is horrid. GE agrees (sort of) and claims that God is benevolent. He is benevolent, but benevolence is not identified with God. Although torture would be right if God willed it God wouldn’t will it because he’s a nice chap.

But isn’t benevolence part of the claim that “God is good”? It is with most Christians and has to be if the claim is to mean anything and the Evil God Argument become necessary. Anyway, just a slight tweak of the name to “the Un-benevolent God Argument” fixes things.

(BTW I’ve blogged on the Euthyphro and the Moral Argument here: )

Bill Snedden said...

Stephen, did you note on Glenn's blog a commenter named "Eric" called your attention to a criticism of the EG argument by Edward Feser who says, and I quote: "Law's argument evidently presupposes a "theistic personalist" or "neo-theist" notion of God and is therefore completely irrelevant to the classical theism of Athanasius, Augustine, Anselm, Aquinas, et al., according to which God is not "good" in the way a human being is good, as if He and we both instantiated the same property; rather, He just is Goodness Itself, and anything less than that is (for a classical theist) necessarily other than God. Hence it is incoherent to suggest that God might be evil."

Feser goes on to flesh out his meaning, but it really seems to me to reduce to nothing more than "classical theism defines God as good, so Law's argument is false".

Can you make any other sense of what he's saying?

Oh, and as an aside, your old friend Sye is going to be coming to my hometown to participate in a debate...oh joy.

Victor Reppert said...

I have argued that, in the absence of a solution to the problem of the single case, we really don't have an objective foundation for prior probabilities. Hume's way of deriving one leads to some unacceptable results. And of course, in order to assess miracle claims, we have to assess them relative to some background belief system or other, part of which might include a response to the evidential problem of evil, but also might include the evaluation of the whole case for and against God independent of the miracle-related evidence. You may be suspicious that, say, someone who has a high prior for fairies has somehow gotten there by way of some sort of irrational process, but proving that no reasonable person could possibly have such a belief system would be a daunting task, to say the least.

unkleE said...

"“It rained yesterday” is not an extraordinary claim. Our theories predict that rain will happen every so often. “The rain fell upwards” is an extraordinary claim because it is simply incompatible with our picture of the world as one where unsupported things fall down. "

I'm not sure this solves things. As a theist, I find the claim that the universe appeared out of nothing for no apparent reason (or that it has always been in existence), and that it appears to be so finely-tuned for life, to be extraordinary claims. Our experience indicates things don't normally happen without a cause, nor that the improbable is very likely to happen. And I find those facts to be good evidence for the possible existence of God.

I think we are still back where we started.

Stephen Law said...

Bill - yes I spotted that. I think you are right.

BTW Tony I actually published Glenn's New Euthyphro in THINK so I am familiar with it. I agree with you it's not persuasive, obviously.

Victor. Even if what you say is true it fails to deal with the points I raise here.

Joshua Blanchard said...

"we may insist that we can quite reasonable accept, solely on the basis of Mary and John’s testimony, that fairies danced at the bottom of their garden last night, just so long as we presuppose, prior to any examination of the evidence, that fairies exist."

I agree with Eric and others that the theist is in an at least slightly different situation, in that the theist doesn't merely presuppose theism, but starts with theism for other reasons. If they are good reasons (or if the theist is in any way epistemically entitled to her theism), then the fairy analogue is stripped of some force.

Also, and perhaps to highlight this issue, a claim like "Jesus rose from the dead" can be broadened to "God raised Jesus from the dead." Christian apologists occasionally make this point. If this is the claim someone is making, then the connection to the probability of theism is clear.

Glenn said...

Tony says: "But isn’t benevolence part of the claim that “God is good”?"

Yes, absolutely. Not morally good, of course, but yes, you're right about this. And as I hope is clear, "A New Euthyphro" does not attempt to make a moral argument for God or to respond to anything like the Malevolent God challenge. The only thing that paper does is show why the Euthyphro dialogue does not, contrary to the claims of some, present any problems for a divine command theory of ethics.

Eric said...

"Feser goes on to flesh out his meaning, but it really seems to me to reduce to nothing more than "classical theism defines God as good, so Law's argument is false"."

"Bill - yes I spotted that. I think you are right."

I don't think this is right at all. It seems to me that Feser, following Aquinas et al, isn't simply defining god as good, but is following to their logical conclusions arguments about the sort of being god must be given certain empirical (e.g. some things change) and metaphysical (e.g. change involves a reduction of potency to act) premises. Now you may of course conclude, after reviewing the arguments, that they're problematic and so on, but I don't see how you can say that those who defend the god of classical theism are just "defining" god as good.

Tony Lloyd said...

I've written a follow up to Glenn's "A New Euthyphro".

Comments very welcome.

Tony Lloyd said...

Bugger, wrong link.

Should be

stuartm said...

It has been said that if God exists then resurrecting Jesus is exactly what you would expect Him to do. There is one problem with this: if you would expect God to do this then you must know why God would do it. According to Christian doctrine, Jesus died for our sins. If that's true then wouldn't it make more sense if Jesus hadn't been resurrected? After all, if the price to be paid for our sins was Jesus' life, and Jesus was given his life back, then presumably the price hasn't been paid, and we're not saved after all.

Also, a popular argument amongst Christian apologists is that the disciples couldn't have made up the story of the resurrection because there was nothing in Jewish tradition to make them expect it. So which is it? Is the resurrection such an obvious thing for God to do that the disciples could easily have convinced themselves that it happened? Or is it so inconceivable that we would have no reason to believe it happened even if we knew God existed?

Ali G said...

This Glenn seem his argument go something like.

1,Sky could not exist with out birds
2,There is birds
3,Therefore sky exists

What a munta

The Atheist Missionary said...

unkleE, why are these facts any less "good evidence for the possible existence of":

1. an evil god;

2. Thor; or

3. You are god. Everything and everyone are just figments of your imagination.

Pvblivs said...

     "Notice, incidentally, that even if I am unable to construct a plausible explanation for why these otherwise highly trustworthy individuals would make such extraordinary claims – it’s implausible, for example, that Ted and Sarah are deliberate hoaxers (for this does not fit at all with what I otherwise know about them), or are the unwitting victims of an elaborate hoax (why would someone go to such extraordinary lengths to pull this trick?) – that would still not lend their testimony much additional credibility."
     Actually, were you to experience such an event and successfully ruled out mundane explanations as per your description, you would have an extraordinary event. It would qualify as extraordinary evidence. Whether you would consider it sufficiently extraordinary to justify belief that their tale was true is a question that I cannot answer. However, if you were to present that event as having happened, you would be met with the more mundane suggestions that you might have misjudged the trustworthiness of Ted and Sarah and that they might, indeed have been the target of a prank. The described event would, itself, be extraordinary.
     Personally, I am skeptical about all (or nearly all) claims. But I suspect that in P1 you are actually talking about dismissing claims rather than being skeptical. Feel free to correct me if I am wrong. But the fact is that most of the example claims are, for all intents and purposes, moot. Suppose I were to accept as a fact that bigfoot really exists (I can't really rule it out.) The first question is "so, what?" Even if he's out there, he's not likely to show up in my living room any time soon. It doesn't really matter to me if faries danced at the bottom of Mary's and John's garden last night. And, unless they stole the plants, it probably doesn't make much difference to Mary and John either. I may be unconvinced by these reports, but they are not important enough for me to declare them false, either.

Stephen Law said...

Hi -0 sorry been away.

The suggestion that there are good reasons for supposing the Judeo-Christian God, say, exists, is irrelevant here. It's false. But irlevant as I was addressing a quite different suggestion.

Puvblivs - the event of S and T giving such testimony would in a way be extraordinary. But so what? I don't understand the relevance. After all, we do have excellent evidence that it happened and that similar such events do happen (read Skeptical Inquirer) I don't know where you going with this.

Feser - some just define God as good. Others have arguments (bad ones). This looked to me to involve at least an element of the former.

Pvblivs said...

     "[T]he event of S and T giving such testimony would in a way be extraordinary. But so what? I don't understand the relevance. After all, we do have excellent evidence that it happened and that similar such events do happen (read Skeptical Inquirer) I don't know where you going with this."
     We, indeed, have excellent evidence of hoaxes. The assertion "extraordinary claims require extraordinary evidence" suggests that something would qualify as that extraordinary evidence. You have described what would, if real, be extraordinary evidence. (It would be extraordinary evidence because of the successful ruling out of a hoax.) Things like the bible are not extraordinary evidence because there is no reason to rule out deception or hallucination.
     A lot of people supporting their religions will say things like "nothing will convince you." I often have mixed feelings about the statement. I often think that they jump to it too quickly because they won't have presented any evidence for me to consider. But there is an unfortunate truth hidden as well. For most believers, no evidence is needed. For most outsiders, no amount of evidence will suffice. If, one night, the stars were rearranged to spell out DEISVNT, it would certainly give me pause. (I'm not talking about "connect the dots" here. This would be stars filling in the areas of the letters and emptiness otherwise.) The event I would be witnessing would be no less unexpected than the claim it supports.

Stephen Law said...

Incidentally why is Glenn's website hardly ever there?

Brian said...

Hi Stephen. Have you heard of Dennett's 'deepity'? I think it's synonymous with pseudo-profundity.

One buzz word you could include is 'synergy'. Meaningly management word that some people think is deep.