Here are seme snippets from volume 1 of the Clipingdale Collection. I don't get much time

Master Index Current Directory Index Go to SkepticTank Go to Human Rights activist Keith Henson Go to Scientology cult

Skeptic Tank!

Here are seme snippets from volume 1 of the Clipingdale Collection. --------------------- I don't get much time to read and post here lately (I hear and appreciate your cries of grief and horror) but I felt compelled to respond to this and in fact to gibber on at some length, for which I apologise. A number of different considerations lend support to the `strong' [atheist] position, and I'll try to enumerate a few. Firstly there is the Popperian observation that only hypotheses which are *falsifiable* are capable of increasing our knowledge about the universe by providing an increase in information (decrease in uncertainty) upon falsification. But why, even assuming that we agree that unfalsifiable hypotheses are informationally useless, does this imply that one *should* buy into the respective falsifiable converse hypotheses rather than merely reserve judgement? There are at least two answers to this at different levels. Firstly, the persistent non-appearance/non-intervention of gods (or unicorns, or whatever) which are claimed to exist (and thus implicitly to appear / intervene at least occasionally), may be legitimately regarded in the context the claim as evidence against it (see Jefferys & Berger, "Ockham's Razor and Bayesian Analysis," American Scientist, Sigma Xi, Jan-Feb 1992 or any text on Bayesian analysis and conditional statistics). The hypothesis that no gods exist is falsifiable because it predicts that events will fall in a narrowly and specifically circumscribed region of the event space ("no gods will ever appear") and one event falling outside this region is sufficient to falsify the hypothesis. The converse hypothesis, that god/s exist, is unfalsifiable because events falling anywhere in the space, including the persistent non- appearance of gods, are consistent with the hypothesis. (I now have to add a disclaimer since Ken Arromdee took me to task over gods which are inconsistently defined, claims of the existence of which are trivially false. So here it is.) As events continue to fall in the region of event space predicted by the specific [and hence falsifiable] hypothesis, our confidence in it must increase by Bayes' Rule, and since the two hypotheses form a partition, our increased confidence in Gods_Don't_Exist implies a corresponding decrease in confidence in Gods_Exist. Before somebody says "Ah, but it's only a *probability*", well, so is all of empirical and scientific knowledge, so while that is undoubtedly true, if you think it's *probably* best to avoid stepping in front of speeding trucks then it should be good enough for you. I adjudge it to be *certainly* best to avoid stepping in front of speeding trucks (and I imagine you do too). This is exactly the same degree of certainty, and of empiricism, as my judgement that, on the basis of the current evidence, I am *certain* that god/s do not exist. But it's a probabilistic thang, sure enough, and I think you do most `strong' atheists a disservice by implying that they regard it as an eternal absolute instead. I can be, and am, certain without being dogmatic about it; evidence that I am wrong will sway me despite my current certainty that I am not wrong. The second answer, which follows on fairly neatly from the speeding truck stuff, is consistency. The agnostic position is logically consistent only if one never/always buys into any/every hypothesis which has an equal lack/ level of evidentiary support, erm, if you see what I mean. If one isn't prepared to reject a particular unsupported, unfalsifiable hypothesis, one must never reject *any* unsupported, unfalsifiable hypothesis no matter how ludicrous it may appear, and even if one just made it up off the top of one's head. Conversely, if one will not accept a falsifiable and as yet unfalsified hypothesis for which there is also evidence (see above), unless and until it is absolutely proven (tricky: remember that its converse is unfalsifiable), then one must never accept *any* such hypothesis until it is absolutely proven. Bang goes all of science for a start, since it is not in the business of absolute proof. Otherwise one is guilty of multiple standards, and of changing one's epistemic rules according to the proposition in question. This has been illustrated here in the last few weeks. Remember this paraphrased exchange?: "Don't confuse the real issue of God with garbage about invisible pink unicorns, God's existence is a different ball game entirely." "How come? Surely this is *exactly* "the real issue"? How is the evidence more compelling for one than for all the other nonsenses?" "Just is so there! I quit, you people have no common sense and can't be argued with seriously..." Now I thought that the middle paraphrasee here (jcopeland if memory serves) was right on the money, and showed up rather nicely the problem of applying different standards to evidence/arguments depending on what they are evidence/arguments *for*. Pointing out by this kind of reductio ad absurdum where people are applying multiple standards is the entire reason why we have such marvellous stuff as invisible pink unicorns or 19-dimensional divine blue bananas (or gods for you sodding bananaians out there :-) discussed so frequently on this board. While you're at liberty to say "I don't have to be consistent; I can choose my epistemic approach to each question differently and as I see fit", that is not going to make for much of a debate because your position could change instantly. In fact you couldn't really even call yourself "agnostic" because even without any new evidence coming in, what's to stop you changing the rules of interpretation of the existing evidence midstream and declaring yourself a Christian (yikes!) and twenty minutes later an atheist (hurrah!) and so on? As to your implication that `strong' atheism implies an immutability or is in any way less open to revision than `weak' atheism or agnosticism, well, this just doesn't follow if whatever beliefs are held are held in response to evidence rather than quasi-religious or psychotic adherence to dogma; I've already indicated why I think that the currently available evidence supports the `strong' atheist position. *The moment* new evidence arrives, an immediate reevaluation of all affected hypotheses is demanded if one claims to [attempt to] believe on the basis of evidence, as does every single person who I have seen defend this position. To suggest that such atheists have nailed their colours irreversibly and dogmatically to the no-god mast is to miss the point entirely. At the core of any evidence-based belief system must necessarily be constant reevaluation. The strong atheist position *is* supported by objective evidence, unlike the theistic position/s. That evidence is the continuing objective non-appearance of gods. So how come this is evidence *for* atheism rather than just a lack of evidence for any theism? Because a specific prediction of the atheist hypothesis is that events will continue to fall only within a delineated region of `event space', namely no gods will appear. Continued non-appearance `events' fall in the region specifically predicted by atheism, and Bayesian statistics indicate that this must cause estimates of the correctness of atheism to increase relative to those hypotheses which less tightly constrain the event space, including those theisms which do not constrain it at all and are thus unfalsifiable (every point in event space, divine non-appearance included, is consistent with such theisms). Rather than take a thousand words: ...-----------------------------------------------------------... C |-----------------------------------------------| B |---------------| A Here are the regions of a 1-D event space consistent respectively with each of three hypotheses A, B and C. C is unfalsifiable (but see `Aside' below) because an event anywhere on the space is consistent with it. Now here come the observations: ... .... ... .. . .. ... . .. ... .. . .... ... ... ... . ... .... ... .. .. . .. .. .. ..... .. . .. ... .. . .... . .. .. ... ... ... .... . .. . . ... .... .... .. ..... .. ... .... ... .. . .. .. ... ... ... .. . .... ... .... ... .. . ... ... .. . .. ... ... .... . .. .. ... ... ... .. . .... etcetera While these observations are consistent with all three hypotheses, and thus do not falsify any of them, nevertheless they better support hypothesis B than C, and in turn they better support hypothesis A than B. After a large number of observations in the same region and none outside, the probability that hypothesis A in this example is correct approaches unity. [Aside: B and C are thus falsified in the limit, even though C is unfalsifiable in a finite number of observations.] But it's up to the individual to make such estimates (although perhaps not in quite the mechanistic numerical fashion of Bayesian analysis) and to draw conclusions. If s/he decides that it's close enough to label as an empirical certainty, much as the existence of gravitational attraction is an empirical certainty for most of us, and if hypothesis A represents atheism, then the individual is a strong atheist as opposed to weak. I fall into this category and for those reasons. The *evidence* for atheism is just that, and is evidence for neither the strong nor the weak version per se, these being merely descriptions of the individual's level of confidence in empirical conclusions. -- Simon Clippingdale Department of Computer Science Tel (+44) 203 523296 University of Warwick FAX (+44) 203 525714 Coventry CV4 7AL, U.K.


E-Mail Fredric L. Rice / The Skeptic Tank