Objections and responses

Sean Carroll: If you found yourself in a world where there was no evil, the commands of God were clearly laid out to all, and all people had always believed the same religion, you would take this as evidence that God existed and his religion was true. In this world we have none of these things, and so you must take their absence as evidence to the contrary.

Response: If I were trying a case and had three eyewitnesses against the defendant, I would take the third one as evidence. By Carroll’s logic, if I only have two eyewitnesses I have to take this as exculpatory.

Some absent evidence proves the opposite of what its presence would prove, but it needn’t.

Carroll 2: The universe could just be a brute fact. Just “there”.

Response: It’s fascinating to argue that the world can only be explained by physical concepts by appealing to concepts that have no application in physics. What good are brute facts in physical theory? Why is there no entry for them in the index of physics textbooks? Perhaps there is no answer to why, say, a radioactive particle decayed at a precise time T, but what this means is that radioactive decay : to precise times :: mixing cookie batter : to the precise locations of chocolate chips. The regularity and pattern is collective across events and not distributed through each particular event. One explains the process better if he drops the assumption of that sort of precision, and so the apparent “brute fact” is really antipodal to brutishness.

Again, some outcomes are unforeseeable and in this sense are without any law-like explanation, like many of the outcomes of complex systems. But even if this counts as a brute fact, physical theory does not call the universe a brute fact in this sense. THE UNIVERSE!! YOU NEVER SEE IT COMING!! At any rate, if one did mean this it is congruent with a theistic explanation. There’s no meaningful prediction of what an artist will make either, especially one with all possible skill. Either he tells us his plans or we have to wait to see what he does. So there is at least one sense of “brute fact” that describes the universe as a procession of the divine art.

As I’m not actually talking to Carroll but to a considerable number of theologians I’d add that there is something creepy and with a whiff of the demonic in this attempt to reduce the universe to the brutish and irrational, which brings to mind the cognitio nocturna of the fallen angels. There is a sense in which physics takes us closer and closer to the irrational so far as it takes us closer to matter, and so to the absence of the actuality and intelligibility, but this can be taken either as a descent into the irrational (as Carroll takes it) or as a deeper and deeper awareness of how the totality of the universe has something essential within it that refers to another for any actuality that it has. To take even the accidents of the world as irrational is always possible, but this itself is congruent with the argument Hopkins makes in Pied Beauty.

Advertisements

20 Comments

  1. August 2, 2016 at 4:45 pm

    On the first point, Carroll is correct, although you are misunderstanding the logic involved.

    If you have two eyewitnesses saying that someone committed the murder, that is evidence against the defendant.

    But let’s suppose you have those two eyewitnesses, and you do not yet know whether or not there is a third.

    Then if you find out that there is a third, that is evidence against the defendant.

    And if you find out that there is not, that is indeed evidence for his innocence, although fairly weak. Because before, you thought there might be a third, even more strongly establishing his guilt. When you found out that there definitely wasn’t, you should become just a little bit less convinced of his guilt.

    All of this can be put into formal mathematics, and is certainly correct. The way it applies to his original argument is something like this:

    Suppose you are raised as a child to believe in God. Your parents, and probably everyone you know, believes in God. So that is evidence that God exists. But as far as you currently know, maybe everyone period believes in God. When you find out that some don’t, that is indeed evidence against God’s existence. And children often treat it as such, when they discover that fact, and rightly, from the viewpoint of evidence, although of course there are also other moral factors involved there.

    • August 2, 2016 at 6:25 pm

      Because before, you thought there might be a third, even more strongly establishing his guilt.

      But that’s nowhere in the example or in what Carroll said (see the link on the previous post, towards the end of the talk). You’re sneaking in the idea of expectation of outcomes which dramatically changes the point. If Carroll could get theists to admit that, as you put it, they “thought there might be” the sort of world he described given theism, then his point would be incontestable, but he doesn’t ask that and no one would grant it.

      All of this can be put into formal mathematics, and is certainly correct.

      I’m sure it can, if you make assumptions like your a priori expectations can never be merely conjectural (i.e. “Let’s see if we can find another witness, just to make sure we found them all”) or discovered to be unwarranted in the light of later findings (“I thought I should have found a third witness, but it turns out that was an unreasonable expectation”). This is why Aristotle was so careful to distinguish what was evident to the wise and to all/ most. Probability of outcomes exists in minds, not in the world, and there are therefore two essentially different sorts of probability. It’s not accidental to your last example that the evidence is evaluated by children.

      • August 2, 2016 at 7:10 pm

        I used children in my example because they might actually be unaware of religious disagreement. Adults of course are not unaware of this. That does not automatically prove that they have taken it into account in their assessment of the evidence. Carroll’s point would be that if in fact you had discovered that everyone agreed, you would have taken that as confirmation. If that is the case, you should have taken the fact that some people disagree as disconfirmation, whenever you discovered that. Which will usually be as a child.

        The weight of this depends, as you say, on how reasonable it is to expect everyone to agree, given that a religion was true. But the weight never goes to zero; if it does, then you should assert that it would not be reasonable to take perfect agreement as confirmation. Which means that Carroll’s point stands: if you take one side as confirmation, you must take the other as disconfirmation, while admittedly if the expectation of the confirming evidence is very weak, the corresponding disconfirmation will be very weak.

        It is also the case that while adults know of the existence of evil and religious disagreement, they often hold, implicitly or explicitly, variations on the denial of these things. For example, some people at least implicitly believe that the existence of God implies that no serious evil would happen to them personally. And for that reason some people lose their faith when it does happen. Likewise, some people believe, even explicitly, that no decent or reasonable person would disagree with their religion. And while some people believe this stubbornly, in the face of the evidence, others believe it by simple ignorance, because that is what they have been taught. And the latter also sometimes lose their faith when they discover decent and reasonable people who do not accept their religion.

      • August 2, 2016 at 7:19 pm

        if you take one side as confirmation, you must take the other as disconfirmation, while admittedly if the expectation of the confirming evidence is very weak, the corresponding disconfirmation will be very weak.

        But now you’re admitting that you’re talking about expected outcomes when I’ve already explained why this is not and cannot be what Carroll is arguing. Carroll’s point rests on a pure counterfactual possibility that his opponent, by Carroll’s own logic, need not expect.

      • August 2, 2016 at 10:23 pm

        if you take one side as confirmation, you must take the other as disconfirmation, while admittedly if the expectation of the confirming evidence is very weak, the corresponding disconfirmation will be very weak.

        This is not at all correct unless we can make more assumptions than are given in the scenario. If we can assume that we are dealing with something that is not itself a necessary truth (which is confirmed by everything), then when we take one side as confirmation, we must take the other as nonconfirmation. Whether nonconfirmation disconfirms, however, rather than simply not increasing confirmation, requires additional information about the relation between evidences. The difference between having Bob’s testimony that something happened and not having it, where Bob is fairly reliable, is not in itself the difference between confirmation and disconfirmation that it happened, but the difference between having some confirmation and not having that particular confirmation. It only disconfirms if I have reason to think that I would have to have Bob’s testimony if the event happened.

        If angels came down and said I am right, that would confirm that I am right; but, even given that, it does not follow that their failure to do so disconfirms in any way my rightness, unless we have reason to think that angels would come down to say I am right if I am. But we can perfectly well be in a position (and usually are) where we have no reason to think that this will happen at all, no matter what, and therefore the lack of it is not a disconfirmation. We need a reason to accept a negation-as-failure rule.

  2. August 2, 2016 at 7:29 pm

    I have not read the text from Carroll, so I don’t know specifically how he puts it. However, you explained him as saying, “If you found yourself… you would take this as evidence…”

    Now if you say it is not true that you would take it as evidence, of course his point cannot stand. But if you would, it stands, despite the fact that this is a counterfactual. And that happens in this way: in the counterfactual world, you are saying, in effect, “I am more sure that God exists than I would be if there was disagreement about it.” And if that is what you would say, then obviously you should say in the real world, “I am less sure that God exists than I would be if there were no disagreement about it.” Otherwise your counterfactual self disagrees with your real self, and one of you is evaluating the evidence incorrectly.

    • August 2, 2016 at 11:01 pm

      This assumes that conclusions can’t be overdetermined by evidence, which is certainly false, and that all evidence affects certainty of conclusions (rather than, for example, preventing other things from reducing the certainty).

      (1) Conclusions can be overdetermined by evidence. Suppose argument A and argument B are both good arguments proving, rigorously, that C is true. (This is a kind of situation that shows up a lot in mathematics.) Having A is enough to take C as perfectly certain, regardless of whether I have B or not; adding B does not make C more certain; taking it away does not make C less certain.

      A different kind of scenario. Suppose A and B give good reasons, but not rigorous proof, that C is true. Does it automatically follow that lacking B makes C less certain than having it? No; for instance, B, despite being different from A, may only give the same good reasons as A for thinking that C is true, not different ones — e.g., A and B perhaps are evidence of C by logically implying something else, and they each on their own logically imply that something else.

      (2) In at least extreme cases, some evidence may be evidence but not affect the certainty of the conclusion. Consider the rigorous proof case again. If a rigorous proof establishes the conclusion as certain, why have other proofs? One possible reason is that while they do not affect the certainty of the conclusion, they do help establish that nothing else can reduce the certainty of the conclusion, for instance, by eliminating the possibility of arguing that the rigorous proof is not a proof for some esoteric elusive reason. Adding B doesn’t confirm the conclusion (which, by definition, is already as confirmed as possible); it simply confirms that the confirmation is a confirmation.

      Things become more controversial, of course, if we talk about non-proof evidence. However, unless we are Bayesians of some kind, we may very well have reason to think that there are (for instance) satisficing thresholds for evidence. If A tells me something happened, and then B tells me something happened, that might make me sure, but if a million eyewitnesses have told me it happened, is it really the case that adding one more makes me more certain? Yet each eyewitness, is, at least to that extent, equivalent to any other eyewitness in terms of being evidence. Am I really entitled to be more certain every day the sun rises that it will rise tomorrow? You can make assumptions that require this, but it seems perfectly reasonable to say that at some point we are as reasonably certain as we will ever be. Each new day is evidence, because (considered as evidence) each day is interchangeable with any other. But whether each new day actually gives us more certainty depends on assuming that there is no point at which reasonable certainty maxes out (which seems to require assuming that our capacity for evaluating evidence is infinite and infinitely precise).

      • August 3, 2016 at 12:08 am

        Brandon, none of your supposed counterexamples are true counterexamples, even if they are taken at face value.

        Take your first case — you have two proofs, and you consider both of them completely conclusive. Then removing one of them does not reduce your certainty, but it is equally true that if you have one of them, the other does not increase your certainty, because by hypothesis you are already as certain as possible. So Carroll’s point stands: if one side increases your certainty, the other decreases it. If one side does not decrease your certainty, the other side does not increase it.

        The same thing is true in your last case. If you reach a supposed maximum of evidence, indeed an additional case will not give you greater certainty, and likewise lacking that case will not make your certainty any less.

        So Carroll’s point is that if one side of a claim increases your certainty, the other side should decrease it. Similarly if one side has no weight, the other side has no weight. And this is true even accepting your reasoning at face value (which I don’t, but that is another discussion.)

      • August 3, 2016 at 8:48 am

        As I explicitly pointed out, your comment seems to require (1) that conclusions can’t be overdetermined by evidence; and (2) that all evidence affects certainty of conclusions. The cases in question are counterexamples to these. And, indeed, your comments on them require that they actually be so. Thus, you either need to establish (A) that your argument does not depend on these assumptions at all; or (B) that it does so in such a way that it can avoid the problems indicated by these counterexamples. Anything else is merely pretending that logic doesn’t apply to you.

        Now, given your response, the question has to be whether (B) is a possible route of reasoning here; and any such argument has only a limited number of assumptions available that gets the relevant result, because we are not talking about this in a void. Remember that there is a specific scenario that is in view here:

        If you found yourself in a world where there was no evil, the commands of God were clearly laid out to all, and all people had always believed the same religion, you would take this as evidence that God existed and his religion was true. In this world we have none of these things, and so you must take their absence as evidence to the contrary.

        So let’s assume this exact case. And let’s assume that the theist has some argument, not based on any of these things, that he regards as a demonstrative proof of God’s existence, and has it in both possibilities. It then follows that he can be perfectly reasonable in taking it to be true that (1) the situation in the one world is such that it could be considered evidence for God’s existence and (2) lacking it will not in any way be evidence that God does not exist, because God’s existence is already proven.

        Now, you have been assuming that evidence has to be analyzed in terms of changes in certainty of what it is evidence for. As I’ve noted, this does not appear to be universally true, but let’s assume it for now. But immediately we are faced with an ambiguity: does this change of certainty involved in evidence have to be such that (i) the evidence changes the certainty in this particular case or (ii) it would change the certainty in some possible case? Consider again overdetermination. Taking away demonstrative proof B from a situation in which we already have demonstrative proof A does not in any way reduce the certainty of conclusion C; but this does not mean that B is not evidence for C. It would, in a situation in which there was no A, make C as certain as A does. Thus it has to be (ii) that is in view here. But notice the situation — whether B actually increases or decreases the certainty of C depends on the holistic status of the evidence, not on any question about what the content of B is, but whether it is evidence (whether it could increase the evidence considered only in its own right) depends solely and entirely on the relation of its content to C. The two come apart, and it is illegitimate to assume, as you are doing, that you can move directly from one to the other.

        You keep assuming (in fact, you explicitly appealed to it in a previous comment) that in the above scenario (with the evidence) what makes no-evil, etc., evidence is that it increases the certainty of the conclusion in the world with the evidence. But it can be evidence without actually doing so, simply because it is such as to increase the certainty if it were only considered on its own. Suppose the theist has something else he regards as demonstrative proof. Then taking this no-evil etc. evidence away (1) does not in any way reduce the certainty of the conclusion; (2) even though it is quite certainly evidence for the conclusion; (3) and even though it would increase the certainty of someone who did not have the demonstrative proof.

        Consider in addition a point I already mentioned before — a case in which the truth in question is a necessary truth. Then it follows that adding some kind of evidence can increase the certainty of the truth (perhaps I had not thought about it at all before) but taking it away does not reduce the certainty of the truth (because it can be confirmed by the absence of the evidence as much as it is confirmed by the presence of evidence).

        Thus your rule of evidence is simply defective, and it’s unclear why you keep dogmatically asserting it without adequate evidence to establish that it is universally true.

    • theofloinn said,

      August 10, 2016 at 2:56 pm

      Then, I suppose the disagreements among the Nine Blind Men would imply that the elephant does not exist.

  3. August 3, 2016 at 9:48 am

    Brandon: if something is considered evidence in favor because “it is such as to increase the certainty if it were only considered on its own,” but it does not actually increase the certainty because the thing is already certain, then likewise the contrary proposition will be evidence against because it is such as to decrease the certainty if it were only considered on its own, but it will not actually decrease the certainty because the thing is proven. In other words, again, your examples fit exactly into my paradigm even without modifying them.

    In any case, I am explicitly assuming a Bayesian understanding of probability and certainty, which you obviously reject, and I don’t wish to argue this particular matter. Under the Bayesian understanding, that absence of evidence is evidence of absence, is a mathematical theorem.

    • August 3, 2016 at 10:24 am

      In other words, again, your examples fit exactly into my paradigm even without modifying them.

      As you note, you are assuming a Bayesian epistemology (an unrestricted Bayesianism rather than a domain-restricted one, since otherwise your response would be irrelevant), and the claim in question is inconsistent with an unrestricted Bayesian epistemology, so it is utter nonsense to claim that it also “fits exactly into it”.

      In any case, I am explicitly assuming a Bayesian understanding of probability and certainty, which you obviously reject

      This is the first time you’ve even mentioned Bayesianism (I was the only one to have mentioned it before, and your response did not confirm the matter), so what you actually mean is that you are now making your assumption of it explicit. But this is utterly irrelevant; we are not chitchatting about your personal views (nor my personal views — about which you know nothing, because they are not relevant to the question) of evidence, but discussing the argument as laid out in the post:

      If you found yourself in a world where there was no evil, the commands of God were clearly laid out to all, and all people had always believed the same religion, you would take this as evidence that God existed and his religion was true. In this world we have none of these things, and so you must take their absence as evidence to the contrary.

      Thus the argument, as set up, is about what the person to whom the argument is addressed must do in evaluating evidence. This does not depend on my view of evidence nor on your view of evidence. That we can posit assumptions that would require it is merely trivial; we can posit assumptions that would require anything (including not accepting it at all). The question is whether the kind of person to whom the argument is addressed can reasonably evaluate evidence in such a way that he doesn’t, in fact, have to draw this conclusion. I have argued that your attempt to claim that he does fails; you have argued almost nothing, since you just keep asserting them as if there weren’t an entire field of theories of evidence other than unrestricted Bayesianism.

      • August 3, 2016 at 10:35 pm

        “it is utter nonsense to claim that it also “fits exactly into it”.”

        No it isn’t. This is what happens in a Bayesian framework if you assign a probability of 0 or 1. Nothing can later change that probability one way or the other.

    • August 3, 2016 at 10:56 am

      then likewise the contrary proposition will be evidence against because it is such as to decrease the certainty if it were only considered on its own, but it will not actually decrease the certainty because the thing is proven.

      I should also have noted in addition that you have again ignored the necessary truth examples, which I have given more than once, to which this claim arguably cannot be applied. (It is at least a common view that all evidence properly evaluated implies the truth of necessary truths, so when we are dealing with contingent evidence, which can be absent or present, both the absence and the presence are evidence for the necessary truth in all possible situations. As I already pointed out, this on its own, if you have no response to it, would suffice to show that your claims do not apply universally. The question then would be why they must apply in this particular kind of case.)

      • August 3, 2016 at 9:54 pm

        That all statements imply necessary truths is true of logical implication, where logical implication means “if A, then B,” because if B is a necessary truth, it is true regardless of what A is. But this has nothing to do with evidence.

      • August 3, 2016 at 11:27 pm

        This is what happens in a Bayesian framework if you assign a probability of 0 or 1. Nothing can later change that probability one way or the other.

        This was not the point of the passage to which you were referring, though; the passage was about the distinction between the question of whether something is evidence and whether it actually changes the probability of the conclusion.

        All of your comments suggest that you are muddled about what the actual argument here is; that’s the only possible explanation for your robotic repetition of Bayesian assumptions in a context that logically requires not that we assume a particular theory of evidence but that we consider what theories of evidence can be counted as at least reasonable. You keep starting by assuming Bayesianism and then treating everything as if it were about Bayesianism. As I noted before, this makes the argument about an extraordinarily trivial thing, namely whether if you assume a theory of evidence that requires X that this requires X; even basic charity should lead you to assume that I am not that freaking stupid. The question actually at hand is, as I also noted before, what a reasonable evaluator of evidence must regard as evidence. You cannot address this by assuming the answer from the get-go. If one were to take a particular theory of evidence as necessary for all reasonable evaluators, the only thing that is relevant is what the reasons are for taking that theory of evidence to be the only one a reasonable evaluator can use.

        To that end I have (explicitly) noted that what seem to be your reasons for holding that reasonable evaluators must evaluate on the rule that if you take one side as confirmation, you must take the other as disconfirmation. I still don’t know whether you actually take these to be the reasons in question, because you only have responded in piecemeal fashion. The principles that you seemed to be assuming were, again, that (1) conclusions can’t be overdetermined by evidence; and (2) all evidence affects certainty of conclusions. Everything I have said has been about the problematics of these two things — why they cannot merely be assumed — because they are the only two things that are actually relevant, assuming that they are indeed principles behind your taking your evaluation rule to be the only one possible for reasonable people.

        That all statements imply necessary truths is true of logical implication, where logical implication means “if A, then B,” because if B is a necessary truth, it is true regardless of what A is. But this has nothing to do with evidence.

        Yes, I am aware of the point; contrary to what you suggest here, it is not true in all logical systems (for instance, it is not true for relevant logic). Unlike you, I haven’t been assuming that a particular logical system or theory of evidence must be the one every reasonable person uses to evaluate evidence. Likewise, other theories of evidence take evidence to be understood precisely in such a way as to include logical implication. So you are assuming that this, too, is not a reasonable possibility for evaluation of evidence, although who knows why, because again we just have to follow along with your dogmatic wisdom rather than getting an explanation as to why a reasonable person can’t take it to be true that if X is implied by something true, this is evidence for X.

  4. The Lambton Worm said,

    August 4, 2016 at 2:43 am

    Um, hello. I’m sorry for butting in on this argument, but, it seems like you’re disagreeing at much greater length than is warranted by your actual positions, at least insofar as I can make them out. Could it maybe be resolved (in proper Scholastic fashion, I guess) by making a distinction? I mean, forgive me if I’m wrong, but each of you seem to be taking the question at issue in a different sense.

    Brandon: you say “The principles that you seemed to be assuming were, again, that (1) conclusions can’t be overdetermined by evidence; and (2) all evidence affects certainty of conclusions.”

    entirelyuseless has previously said: “if something is considered evidence in favor because “it is such as to increase the certainty if it were only considered on its own,” but it does not actually increase the certainty because the thing is already certain, then likewise the contrary proposition will be evidence against because it is such as to decrease the certainty if it were only considered on its own, but it will not actually decrease the certainty because the thing is proven.” This seems to be a straightforward contradiction of those two theses, at least as they are read naively.

    It seems to me to be very, very intuitive that if a piece of information could never under any circumstances change anyone’s certainty about a matter, I wouldn’t ever think to call it ‘evidence’ with regard to that matter. I don’t just mean here that it couldn’t change my certainty after I had heard other relevant arguments – my claim is only that if we admitted that something could never under any circumstances change anyone’s level of certainly about a claim, even if it were the first thing relevant to judging the certainty of the claim that had ever been brought before that person, we would never have any reason to call that thing ‘evidence’ for that claim. This seems to me to just be so close to the definition of the word that it would be ludicrous to deny it – you (Brandon) certainly haven’t given any arguments strong enough to budge my intuition in that direction.

    But in any case, this brings me to my main point – it seems to me that you (Brandon) aren’t taking ‘everyone believes in God’ to be evidence for God in the sense that Carroll and entirelyuseless are taking it to be evidence. If I’ve read you right, you take both the state of affairs in which everyone agrees and the state of affairs in which many people disagree to both be evidence for God, because God is demonstrated by argument to be necessary for either state of affairs to obtain. (I’m aware this isn’t the only reason you think we could count both as evidence – but this is the kind of thing, right?)

    But it would never occur to someone unacquainted with such arguments that what you might mean by ‘everyone agrees that x is a reason to believe x’ might mean ‘everyone agrees that x is a reason to believe x, because x is a necessary condition for beliefs or disagreements to happen’. **The key thing I want to press here is that a single fact under different descriptions can and often does count as evidence in different ways or for different things.** For example, ‘everyone agrees that intellects exist’ is a reason to believe in intellects both because it is reasonable to think that a thing is true if everyone, wise and common folk alike, affirm it: and because intellects are a condition of agreement existing at all. If we are able to recognise the second piece/description of evidence, we don’t need the first one for the conclusion to be established: but if someone only recognised the first one they would have some evidence even if they didn’t recognise the second.

    Thus I take it that Carroll and entirelyuseless had been ‘using universal agreement is evidence/reason to believe’ only in the first sense, while Brandon is using it in both senses but primarily the second one; and Brandon is taking ‘piece of evidence’ to refer to the fact itself and what the fact can establish, rather than taking ‘piece of evidence’ to refer to a description of the fact and what the fact can establish under that description.

    If I am correct, it can be the case that both agreement and disagreement about God could count as sufficient evidence for God in the senses that Brandon has been using all of the words, and that disagreement about God under it’s naive description, and to a person in a certain state, could count as evidence against God/a reason to disbelieve. In other words: disagreement about God is evidence for God simpliciter; but can be evidence against God secundum quid.

    And thus (fingers crossed) the argument is solved.

    • August 4, 2016 at 7:58 am

      There is in fact, as far as I can see, no definite territory of disagreement; entirelyuseless has been taking the dispute to be about what follows on a Bayesian theory of evidence, which I take to be trivial in this context and not in dispute, and I have been saying that the question actually at hand is what a reasonable evaluator of evidence must assume in evaluating evidence, so we are back one step at the reasons why someone must accept this or that particular approach to evidence rather than another. I am indeed aware that entirelyuseless is assuming a particular view of evidence; the point is that this is precisely to fail to address the only actual question of dispute that does not make the matter trivial: why there is no other reasonable way of looking at evidence.

      It seems to me to be very, very intuitive that if a piece of information could never under any circumstances change anyone’s certainty about a matter, I wouldn’t ever think to call it ‘evidence’ with regard to that matter.

      This is reasonable, but unless you are meaning ‘under any circumstances’ in a very extreme sense, going further, as you do, and claiming that it is ludicrous to disagree, would require other things that other people sometimes regard as counterintuitive– for instance, that rigorous proofs are not evidence for a conclusion if one already has other rigorous proofs for the conclusion; that probable reasoning can never actually settle a question conclusively, with full moral certainty; that learning that X is implied by something true is not learning evidence for X; that all evidence directly affects the certainty of a conclusion; that evidence is to be understood as a normative-causal notion about how information affects us rather than as an objective logical or probability-theoretical feature of information that does not depend on our responses to it; that there is no such thing as evidence that no one will ever know about; and so forth. All of these are obviously controvertible; it is not obvious, however, that they are dismissible out of hand as ludicrous. One cannot determine disputed questions of reasonableness by intuition or assumption; the question is: what are the actual reasons for thinking that nobody can have different intuitions or assumptions and still be evaluating evidence reasonably — that every reasonable person must proceed on these grounds?

      If I’ve read you right, you take both the state of affairs in which everyone agrees and the state of affairs in which many people disagree to both be evidence for God, because God is demonstrated by argument to be necessary for either state of affairs to obtain.

      I have made no claims about how one must assess the claims in question, doing so would require already having a definitive answer to the point actually in dispute. (Admittedly, this is obscured by my only having slowly realized that entirelyuseless was taking me to be opposing a particular view to his own rather than raising questions about the idea that everyone must accept his assumptions.) But I do think your point about how the status of a thing as evidence is always under a description is very salutary, and using the simpliciter/secundum quid distinction in this context is actually quite interesting.

      • The Lambton Worm said,

        August 4, 2016 at 11:02 am

        Hello again.

        “unless you are meaning ‘under any circumstances’ in a very extreme sense, going further, as you do, and claiming that it is ludicrous to disagree, would require other things that other people sometimes regard as counterintuitive– for instance, that rigorous proofs are not evidence for a conclusion if one already has other rigorous proofs for the conclusion”

        I glossed my characterisation of what would count as evidence with “even if it were the first thing relevant to judging the certainty of the claim that had ever been brought before that person” – I think I stand by something like that one. Minimally, I would say that for there to be any reason to call something evidence, there must be at least some possible (or at least some imaginable, or some ‘in principle’) circumstance in which we recognise that it could make something seem more likely to someone. I think entirelyuseless’s characterisations end up at close to the same point. But I’m not especially concerned to run this point into the ground because evidence is a word in natural language and probably someone thinks it’s obvious that it works the other way. To be honest, I think probably the question of whether evidence is normative-causal or not is a pseudo-question: when the detective takes note of the scuff-mark, was it in some sense evidence all along even though the Watson saw nothing in it, or does it count as evidence because it stands (or, alternatively, could in principle stand) in a being-evidence kind of relation to some investigation? I don’t really know and probably you could argue forever about it, like William James’s walking around a squirrel. All that’s really required for what I’ve said is the very weak claim that if someone does use ‘evidence’ in the normative-causal way, that is one reasonable way of using the word, even if there are other competing reasonable ways to use it.

        You’re quite right that I misrepresented you with regard to how things counted as evidence – I wasn’t watching my words carefully. Sorry about that. In any case, I’m glad you found some help or interest in what I wrote.

        Lambton

      • August 4, 2016 at 11:55 am

        The difficulty, of course, would be that if you extend ‘under any circumstances’ or ‘in principle’ far enough, there’s nothing it does not exclude — for instance, even necessary truths are not things of which we are certain from all eternity, but come to be certain about, so anything (including contrary propositions, contrary to entirelyuseless’s claim above) could be counted as making a necessary truth more likely to someone, if one sets the conditions to allow that much. Make the bounds too generous and nothing is ruled out; too narrow and it becomes increasingly controvertible.

        But my claim throughout has been that it is at least not obvious that there is one and only one way for a person to classify and evaluate evidence and still be reasonable, so I wouldn’t deny that people can reasonably approach the matter as you suggest.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: