Posted: Dec 19, 2011 3:57 am
Hi Calilasseia --
With all due respect, you seem to be confused about prior probability on an epistemic interpretation. There are TWO bodies of evidence: (1) the background evidence (often or usually called background information), and (2) the specific evidence to be explained. The background evidence is what's used to determine the prior probability; the evidence to be explained is what's used to determine the explanatory power. For an example of a Bayesian argument where both the background evidence and specific evidence to be explained are explicitly stated, please see this page:
http://secularoutpost.infidels.org/2011/11/argument-from-scale-as-revisited-part-2.html
That page gives an example of a Bayesian formulation of Nicholas Everitt's "argument from scale" for atheism. I conclude the argument itself fails, but, again, that page demonstrates how prior probabilities can be based on evidence.
I don't think we are talking about the same thing. Let B is our background information (or evidence), E is the evidence for biological evolution (e.g., the fossil record, biogeography, etc.), and H is the hypothesis of evolution. In this context, "explanatory power" refers to the probability that we would find the evidence of the fossil record, biogeography, etc., on the assumption that evolution is true. It's the probability of E conditional upon H & B, i.e., Pr(E|H&B).
The link I provided above contains a worked example.
So what's the problem? It sounds like we agree. People "make shit up" and skeptics show that whatever evidence they offer for their claims doesn't yield a high final probability.
I know you don't believe that theism has a probability greater than zero. What I should have written is that "you seem to be under the mistaken impression that IF theism has a higher prior probability than other forms of supernaturalism, that it somehow follows that theism has a high final probability." This is in response to where you wrote:
Moving on to your next comment.
Those numbers were purely for illustration; I don't claim to be able to calculate quantitative values for the sort of epistemic probabilities involved. In my writings, I never offer or estimate 'absolute' probability values; instead I use words like "high," "low," etc. which represent a range of numerical values.
Again, I made up those numbers simply as an example to find out if, in fact, you think that IF theism has a higher prior probability than other forms of supernaturalism, that it somehow follows that theism has a high final probability. Is that what you think or not?
And, again, if "the real world evidence" shows that some idea is "plain, flat, wrong," I suggest that it is not because the real world evidence entails that the idea is false, but because the evidence makes it highly probable that the idea is false.
See above and my earlier posts.
Cool. I've been accused of having a priori presuppositions also, so I know how annoying it is.
In summary, I think there's been a big misunderstanding here. The propositions I include in B (the background evidence or information) include empirical propositions and so include "real world evidence." As far as answering your questions is concerned, I'm not sure what you are referring to, but I'll take a look at your other posts on this thread to locate them. It's quite possible I haven't answered something.
Calilasseia wrote:jlowder wrote:Calilasseia wrote:Except that Draper misses the point.
Hi Calilasseia -- No, with all due respect, YOU have missed the point. You are simply confused about the difference between prior probability, explanatory power, and final probability.
Oh really? Since I've been exposing canards erected by creationists with respect to these for something like four years, I suspect your above assertion is somewhat premature.jlowder wrote:Prior probability: this is a measure of the plausibility of a hypothesis, i.e., its probability independent of the evidence to be explained.
Tell me, how can this be determined without reference to the evidence? Please explain this. Only this is the whole point of evidence - it provides a means of separating sound from unsound postulates.
With all due respect, you seem to be confused about prior probability on an epistemic interpretation. There are TWO bodies of evidence: (1) the background evidence (often or usually called background information), and (2) the specific evidence to be explained. The background evidence is what's used to determine the prior probability; the evidence to be explained is what's used to determine the explanatory power. For an example of a Bayesian argument where both the background evidence and specific evidence to be explained are explicitly stated, please see this page:
http://secularoutpost.infidels.org/2011/11/argument-from-scale-as-revisited-part-2.html
That page gives an example of a Bayesian formulation of Nicholas Everitt's "argument from scale" for atheism. I conclude the argument itself fails, but, again, that page demonstrates how prior probabilities can be based on evidence.
Calilasseia wrote:jlowder wrote:Explanatory power: this is a measure of how well the hypothesis yields the data, i.e., the probability of the evidence conditional upon the hypothesis and the background information.
Er, once we have a body of evidence available, the probability of that evidence is 1. Because it exists and has been demonstrated to exist. I think you need to inject some rigour into the above, and specify precisely what you mean.
I don't think we are talking about the same thing. Let B is our background information (or evidence), E is the evidence for biological evolution (e.g., the fossil record, biogeography, etc.), and H is the hypothesis of evolution. In this context, "explanatory power" refers to the probability that we would find the evidence of the fossil record, biogeography, etc., on the assumption that evolution is true. It's the probability of E conditional upon H & B, i.e., Pr(E|H&B).
Calilasseia wrote:jlowder wrote:Final probability: the probability of the hypothesis after assessing the prior probability and explanatory power, i.e., the probability of the hypothesis on the background information and the evidence to be explained.
Once again, I think you need to inject some rigour here. A worked example would be helpful.
The link I provided above contains a worked example.
Calilasseia wrote:jlowder wrote:If someone makes shit up, we can evaluate their claims in terms of the distinction between prior probability and explanatory power, and show that whatever evidence they offer for their claims doesn't yield a high final probability.
Er, excuse me, but this is what people have been doing here before you arrived.
So what's the problem? It sounds like we agree. People "make shit up" and skeptics show that whatever evidence they offer for their claims doesn't yield a high final probability.
Calilasseia wrote:jlowder wrote:Calilasseia wrote:Which consists of demonstrating that anyone can make shit up and pretend that reality conforms thereto, and that there is no essential difference between shit made up recently to provide a demonstration of this, and shit made up 3,000 years ago by people who thought that their magic man actually existed. At least, no one has been able to provide substantive demonstrations that their pet mythologies are anything other than assertion-laden made up shit. Trying to claim, for example, that 3,000 year old mythological made up shit is somehow "more valid" than recent made up shit, because the recent made up shit was deliberately constructed to expose the made-up-shitness of supernaturalism full stop, whilst the 3,000 year old made up shit was concocted by people who actually believed it, misses the target by light years.
That paragraph is not of obvious relevance to anything I or Draper have written. First, "validity" is a concept which applies only to deductive arguments, not inductive arguments. The issue is not validity but inductive correctness.
Ah, if you're going to try and out-Quine me, allow me to step in here. If an argument can be couched formally in terms of the propositional calculus, then a validity test can be applied thereto.jlowder wrote:Second, you seem to be under the mistaken impression that just because theism has a higher prior probability than other forms of supernaturalism, that it somehow follows that theism has a high final probability.
I don't think theism has a higher probability of any sort, prior or otherwise, than other forms of supernaturalism. I regard them all as having zero probability. You're addressing above ideas I never erected.
I know you don't believe that theism has a probability greater than zero. What I should have written is that "you seem to be under the mistaken impression that IF theism has a higher prior probability than other forms of supernaturalism, that it somehow follows that theism has a high final probability." This is in response to where you wrote:
Calilasseia wrote:Trying to claim, for example, that 3,000 year old mythological made up shit is somehow "more valid" than recent made up shit, because the recent made up shit was deliberately constructed to expose the made-up-shitness of supernaturalism full stop, whilst the 3,000 year old made up shit was concocted by people who actually believed it, misses the target by light years.
Moving on to your next comment.
Calilasseia wrote:jlowder wrote:Final probability is the product of prior probability and explanatory power. Let's make up numbers solely for the sake of illustration. Suppose that the prior probability of deism is .1, the prior probability of theism is .15, and the final probability of theism with respect to the evidence is .0001. Would that bother you? If yes, why?
Well first, I'd want to know the rigorous basis upon which you arrived at those numbers. I await your convincing me that it's even possible to calculate a 'prior probability' as you have defined it above, because in the absence of evidence to sort out which hypotheses are supported by reality, and which are not, I cannot see how you can calculate this prior probability. I eagerly await your demonstrating otherwise.
Those numbers were purely for illustration; I don't claim to be able to calculate quantitative values for the sort of epistemic probabilities involved. In my writings, I never offer or estimate 'absolute' probability values; instead I use words like "high," "low," etc. which represent a range of numerical values.
Again, I made up those numbers simply as an example to find out if, in fact, you think that IF theism has a higher prior probability than other forms of supernaturalism, that it somehow follows that theism has a high final probability. Is that what you think or not?
Calilasseia wrote:jlowder wrote:Calilasseia wrote:Likewise, trying to erect some bizarre taxonomy of made up shit, on the basis of some probabilistic calculus, is a wasted exercise. At bottom, made up shit is precisely that, and once it has been determined that a given set of assertions is basically made up shit, and therefore contains no substantive knowledge about the real world,
I am not completely sure what you are referring to by "taxonomy," but, more important, I don't think you realize how self-refuting this paragraph comes across as. How has it "been determined that a given set of assertions is basically made up shit, and therefore contains no substantive knowledge about the real world?" Is that something you know a priori? Or a posteriori? Are you 100% certain that it's made up shit or is there some chance, no matter how small (say 10^-40), that the apparently made up shit could be true? If there is any chance at all, no matter how small, that the apparently made up shit could be true (why wouldn't there be?), then you need to deal with probabilities.
When a particular collection of supernaturalist assertions, such as those found in a mythology, includes assertions that are known to be plain, flat, wrong, because the real world evidence says that they are plain, flat, wrong, then the idea that said mythology might somehow become admissible as a purported "explanation" for the universe and its contents, if only we can exercise the requisite apologetic cunning, is untenable. Allow me to give you an example. Genesis 30:37-39 contains an assertion, to the effect that it is possible to effect wholesale change in the genomes of living organisms, simply by having the parents shag alongside different coloured sticks. We know that this is complete hooey. How do we know this? Oh, that's right, evidence. Courtesy of thousands of diligent experiments conducted by an Austrian monk. Who alighted upon the real basis of inheritance.
And, again, if "the real world evidence" shows that some idea is "plain, flat, wrong," I suggest that it is not because the real world evidence entails that the idea is false, but because the evidence makes it highly probable that the idea is false.
Calilasseia wrote:Now, the existence of crassly stupid assertions such as the one I've just expounded upon above, within this mythology, tells me that supernaturalist claims about this mythology, to the effect that it is the product of some fantastically intelligent magic entity capable of fabricating a universe, are themselves complete hooey. Because supernaturalist assertions about this entity include the assertion that this entity is in possession of purportedly "perfect" knowledge. Yet apparently this entity, purportedly being in a position to know that human beings would find out in the future that the assertion I've expounded above is complete hooey, allowed the backward Middle Eastern nomads he chose as his ghostwriters to leave this nonsense in. It's basic, elementary facts such as this, that lead me to conclude that the probability of the fans of this mythology being right, when they claim that their pet mythology somehow dictates how reality behaves, is precisely zero. Once again, I'd like to know how the hell you arrive at non-zero numbers for your 'prior probability' calculations, given basic facts such as this.
See above and my earlier posts.
Calilasseia wrote:jlowder wrote:Note also, BTW, that if you believe theism is "made up shit," then by definition and on your own admission, you not only lack the belief that God exists, but you also hold the positive belief that God does not exist.
Actually, I've reinforced the distinction with respect to this before you arrived here. Namely, that from a rigorous position, I begin by refusing to accept uncritically unsupported supernaturalist assertions, precisely because they are unsupported. However, my concluding that these assertions are made up shit arises from a different process, namely evaluating relevant evidence. It is that evidence that tells me that supernaturalist assertions are made up shit, not any a priori presuppositions, which I emphasise because all too often, I see the tiresome erection of canards by supernaturalists, to the effect that my refusal to accept uncritically their unsupported assertions as puportedly constituting established fact, somehow constitutes "presuppositions" on my part. It doesn't.
Cool. I've been accused of having a priori presuppositions also, so I know how annoying it is.
Calilasseia wrote:jlowder wrote:Calilasseia wrote:do we really need to engage in navel gazing with respect to the taxonomic position of one brand of made up shit against another?
I understand you think it's "navel gazing." I think it's the logically correct way to approach these issues. If you find it boring, that's fine. I'm not claiming you have to do anything. But if you are going to claim that all supernatural claims are equally implausible, then you need to be able to back that up and deal with objections that you consider "navel gazing."
Which once again brings me back to real world evidence. Which is the final arbiter in these matters. At bottom, it's reality that decides whether our assertions about it are right or wrong. Unfortunately, this is a lesson that I have a hard time teaching supernaturalists, because all too frequently, they want the discoursive door left open to admit all sorts of specious claptrap and fallacious reasoning. All too often, supernaturalists seem to think that the usual standards of proper discourse shouldn't apply to their wibble. I disagree in the strongest possible terms with this, which is why I subject their assertions to a certain degree of calculated discoursive brutality, not least because it seems to be the only way to shock some of them into waking up and smelling the roses.
Now, I really would like to see you explain your approach a little better than you have done, because from where I'm standing, you appear to be saying that you possess some sort of calculus, with which you can evaluate the plausibility of a hypothesis, before bothering to determine if the real world evidence tosses that hypothesis into the bin or not. I'd like you to demonstrate this calculus in action if you possess it. Not that I'm optimistic about this, because you never bothered answering my previous questions.
In summary, I think there's been a big misunderstanding here. The propositions I include in B (the background evidence or information) include empirical propositions and so include "real world evidence." As far as answering your questions is concerned, I'm not sure what you are referring to, but I'll take a look at your other posts on this thread to locate them. It's quite possible I haven't answered something.