| Religion, Rationality, and a letter to a rabbi in response to a phone call. |
[Jan. 15th, 2009|06:38 pm] |
Okay, a bit of background: A while back, the head of a chasidic sect (not Lubavitch, the Rebbe of a different group) was apparently visiting in the area. Anyways, bunch of people (including my parents) went to get blessings from him blah blah blah...
So, apparently when they spoke to him and they had mentioned me, he requested to meet me. Anyways, so my parents asked if I would come. me: "fine..."
Anyways, met him, etc etc, do I do this or that, eventually I explained that the issue is that I'm an unbeliever, blah blah... he was nice, but used some rather poor argments.
Fast forward about half a year. It was thanksgiving evening, we were watching some movie we rented, and we get a call. Turns out to be some rabbi that represents/works for this Rebbe, specifically wanted me, started telling me random stuff about how I have the potential to be a great scholar, blah blah blah. I politely explained him why that was unlikely. (The whole me being an atheist and so on thing..) aaanyways, he managed to get out of me a promise to email him and discuss the issues.
I decided to carefully write a rather comprehensive letter, detailing my basic views and to respond to some of the things he said on the phone. I got a bit technical, but at least what I tried to do was to make it as clear as possible. I didn't want to simply impress with a wall of confusing stuff, but to communicate and argue about actual ideas. Took me a bit to write the letter, then I asked a couple people to look over it just to see if it was comprehensible (not if they agreed with what I was saying, but I simply wanted to verify whether I was at all succeeding in communicating what I was trying to communicate) Anyways, so it was about two weeks or so after thanksgiving when I finally sent the letter. I have yet to see a reply, and I took a while writing it, so why waste that effort? Might as well let others read it too, so here it is:
( Read more... ) |
|
|
| Part 2/? |
[Jun. 30th, 2006|07:22 pm] |
Well, that conversation was a waste of time...
One Rabbi I've been talking to, not really being able to help me out (See original entry about Bayesian religious stuff) gave me a list of others to contact that may be able to handle my questions and perhaps furnish me with pieces of evidence ("evidence" in the Bayesian sense.. not necessarily numerically, but at least could make an argument about the likelihood ratio being one way vs the other..)
Just contacted the first one on the list.
Ugh. Suffice it to say that conversation was an argument _against_ belief. (ie, if this is the best the rabbis can do...)
Anyways, Called him, got a machine, he called back, and when I tried to explain a bit of the background for how I was trying to figure stuff out, was basically told stuff like I'm playing mind games, being pseudo-intellectual, and he brought up Pascal's Wager (ugh... I was almost embarrassed on his behalf for the fact that be brought it up.)
Aaanyways, basically he was just telling me stuff how "either you decide to believe life has meaning or it's just a party to have fun", and so on and so forth. And basically said that it's pointless to try to rationally figure stuff out because it's just rationalizing (the two are, of course, very different) and that you'll always find stuff to support what you want to believe. (Geee... I had no idea... WHY ELSE WOULD I BE TRYING TO USE SPECIFIC RULES ABOUT WHAT I WOULD CONSIDER AS EVIDENCE?! AND WHY WOULD I GO TO OUTSIDE SOURCES THEN?! The entire point of this exercise is to find the bits of evidence I haven't noticed or allowed myself to notice or otherwise haven't considered. As for how to weigh the evidence, as I explained previously, there's a specific principle I'm using for that..)
Also a bit about how to believe is a challenge, there will be always evidence against, that's what makes it free will and worth while, blah blah blah. Basically, there he was just giving the whole "any evidence against is a test of faith" spiel. whoop de doo.
That conversation was worse than useless... Well, I guess one could argue that it was useful in that it furnished me with another piece of in this case perhaps weak evidence against. But anyways... No one can argue honestly I'm not at least making the attempt... |
|
|
| Bayesian religious stuff, part 1/? |
[Dec. 2nd, 2005|03:01 pm] |
My first entry ever here, because now I have something to say. I'm somewhat of an unbeliever now, background is orthodox Judaism, but still tend to drive myself nuts with "but what if really I'm a bad person, blah blah blah." No, that's not what this is about, this is just the setup. I decided enough's enough, let us calculate.
First, I know someone else did this, apparently there's a book, The Probability of God, in which the author tries to use Bayesian reasoning to compute, well, guess. :) Haven't read it. But I do know Bayesian reasoning, so I'll do it myself.
First, what's all this Bayesian kerfungabooboo? Bayes Theorem is an basic principle from probability theory that can be used to deal with questions like "if I have x belief in hypothesis H, and am presented with new evidence E, what should my new belief y be?"
There're a few different forms of the theorem. Some give specific probabilities as outcomes. Problem is to do that you have to be able to work with the set of all possibilities. That is, you'd need to know all possibile hypothesies and compute values for all of them. But that'd be a bit hard. Fortunately, a bit of mathematical jiggerwocky can be used to show that one can also use similar methods to instead just compare confidence levels. ie, you can validly do "I believe x is twice as likely as y, and now have new evidence E, so now the ratio of probabilities is..." Alternately, you could say there's a version of Bayes theorem that works with odds.
In other words, you can compare different theories, and update your belief in them based on evidence. Perhaps the best way to explain this is to simply show what I mean. I'll be comparing Judasim and Atheisim here. Yes, there're more possibilities, but these are the two I'll be dealing with here. Some of the specifics will be a bit general, some may only work with Judasim. I may later on try to add other religions to the comparison, but right now just working with what we have here. Also, for now, I'll just be trying to work out a Bayes factor (ie, just the strength of the evidence, as I see it), so let's not get too worried yet about what the priors (starting points) should be.
EDIT: Adding mini Bayesian reasoning tutorial to hopefully make this clearer.
Okay, imagine you have two (mutually exclusive) hypothesies, we'll call them A and B. There may be other possibilities, but those are the two that're being looked at.
Now, you don't actually have a probability for either of those to be true, merely relative confidence values. Let's say, oh, that you start with the same amount of belief in both of them. So the actual numbers don't matter, just the ratio. So, let's call it 100 units of belief in each.
Now, consider some event/evidence/whatever you want E to stand for. now, there are a couple questions. What's the probability that E would have happened if A was true, and what about if B was true instead? the notation is p(E|A) and p(E|B)
Well, let's say, for this example, P(E|A)=5%, and p(E|B)=20%
Well, for the A end of things, 5% of that belief would correspond to A and E being true, and 95% to A and not E (~E), right?
So, starting with A and B each having 100 units of belief, you'd then get A&E being true having 5 units of belief, A&~E having the remaining 95 of A, and similarly, B&E having 20, and B&~E having 80.
Now, let's say you actually observe E. So now you have evidence E. So, what does that actually do? it makes all the ~E stuff go away, because you now know that E is true. so what's left? A&E having 5 points, B&E having 20. So, now that you know E happened, you can say that if you started with the A and B being equal, you now have 4 times as much confidence in B than you have in A. Now, ratios are all that matter here, as can be seen, so instead of multiplying each side by 5/100 and 20/100 respectively, I could have just used 5 and 20, since afterwards, the 100s would cancel in the ratios.
And so on and so forth, incrementally adding evidence. the significance of any piece of evidence when comparing A and B is determined entirely by the chances that it would have happened given A or B. (well, given all the previous pieces of evidence as well.. ie, more formally, one would look at these values: for E5, p(E5|A&E1&E2&E3&E4) and p(E5|B&E1&E2...)
Okay? Make more sense now?
Note, most of the numbers I'll be using are produced in a _very_ fuzzy way, as you'll see. So am posting this partly so that anyone who has any suggestions about shifting any of the numbers to more accurately reflect reality, well, give me your suggestions. I'm just trying to get to the truth here as best as I can. Foaming at the mouth fundumentalisim is not welcome though.
Okay, let's begin, it'll probably (heh heh) be more clear what I'm doing here once we start. oh, btw, all that matters is the ratios here, so to keep the numbers easy to work with, well, you'll see what I mean.
First, the problem of suffering in the world. Judaism claims an infinitely merciful, powerful, intelligent god. Now, anyone can say "blah blah, mysterious reasons, needs to be this way, beyond our understanding", but let's actually look at what's more likely, okay?
Here's how we'll do it: If you didn't know the way it was, but did know god was infinitely compassionate, blah blah blah, what likelyhood would you have assigned to the idea that the world he'd create would have all the suffering that it does? I mean, sure, one could say, mysterious reasons, blah blah blah... but I wouldn't have expected it.
In fact, even if there is some mysterious reason, you'd think an infinitely compassionate all powerful thingy would be able to find some other way to achive whatever needs to be achived. I've heard that there's been some religious speculation on what about a _non_ omnipotent diety. but even so, if we accept one that could create the universe, arrange it, etc etc etc, and superduper smart and so on, how likely would it be that the lack of ability is _just_ in that way such that the only way to achive whatever it is that needs to be achieved in the best way is via the general ickyness going on in the world? not that likely.
Max 1%, I'd probably be more inclined to .1% if even that, likely far far less... but, we'll call it 1%
so, probability of general suffering/nastyness/etc in the world given that it was created by blah blah blah...
we'll just call this p(suffering|Judaism)=1%
But, this, on its own, doesn't tell us much about the chance Judaism is true.
we also need p(suffering|Atheism).
Not entirely sure what'd be a good number for this, it's open to debate. and I mean that, you guys can toss suggestions with reasons/etc, if you want, but... I'll call it, say, p(suffering|Atheisim)=75%
So, if we started with Judaism and Atheisim equal (ie, 1:1) (which is likely generous to Judaisim, probabily ought be more like Atheisim equal to all religion or all monotheistic religions at least... or something like that... whateve.) So, as I was saying, if we started with them equal, then multiplying each by the appropriate number now gives is 1:75. That is, if we accept a 1:1 starting point and my values for the likelyhood of suffering given each of the two, then now we have Atheisim being 75 times as likely as Judasm.
That's basically how Bayesian stuff works. Strength of evidence is determined by asking "How likely would we have been to observe E given each of these possibilities?", then working with the ratios.
Next, Judasim does more than claim stuff about the origin of the world, it says that people are supposed to believe and do certain things. Okay. So if that's true, you'd have expected that god would have put so much obvious evidence that there's no way we would have missed it, instead of playing this faith game of "tee hee hee, these people say you're supposed to do this, those say I want you to do that, they say I want you to do that, and I'm just gonna hide, only giving hints that can be interpreted in many ways for each of them... whoops, you guessed the wrong one? oh well, to bad, you lose, thanks for playing, try again next time!"
Seriously, I DON'T think so. I always here it justified with "Well, if it was obvious, where'd be the challenge, the free will?" Well, gee... what good is free will without being able to make an informed decision? And heck, the challenge could be, if such was truly needed, to insure there was plenty of temptations for naughtyness, but still at least make it clear what the right answer is, so that those that do care aboud doing Good(tm) could at least know which direction to direct their efforts...
But, still, if Judaisim is true, well, we were really wrong about that suffering thing, right? so let's be generous and say that we'd give a 5% chance to the idea that if all this is true, god would do this whole "gonna sneak, gonna hide, and gonna plant false evidence, but you're in trouble if you can't find me" me thing. So p(lack of obvious evidence|Judaisim)=5%
And obviously, you wouldn't expect that in an atheistic universe for there to be much strong evidence at all in favor of Judaism or such, so, probably should say 99.9999 blah blah, but, again, will still be a bit generous to the religious end and say p(lack of obvious evidence|atheisim)=95%
now, since only ratio matters, well, 95/5=19, so the factor for this specific evidence is 1:19, so doing the multiplication, our current "score" is, drumroll please... 1:1425 (ie, 1*1:75*19)
Next, something a bit different, supporting the other side. The fact that Jews still exist as a group, still in large numbers follow the Torah, etc...
Why's this interesting? Well, Jews have been around for a _LONG_ time, and seems that historically a whole lot of people have wanted to get rid of us.
So the fact that Jews as a group are still around and doing well, when so many other great and powerful groups fell, is actually seemingly evidence in favor of it having some special protection... So let's see....
let's say, oh, p(survival of Jews|Judasim)=95% (we'd expect god'd try to keep his Chosen People(tm) around... but not more than 95 because, hey, we were so wrong on the other stuff, so who knows what mysterious supprises there'd be...)
and as for p(survival of Jews|Atheisim)? well, there have been a few other groups that have been around for quite a while... Those that believe in the Vedas (I guess collectively Hinduisim then?), Buddhists have been around pretty long. Christianity as well, but Judasim still somewhat longer than the latter two. Not sure about the first. Plus, well, again, historically many many have tried to get rid of us. (a friend suggested to me that all jewish holidays follow the mold of "someone tried to kill us. they failed. let's party!" This isn't really true, but I'd have to say that there's at least a bit of truth to it.)
So, what should we say the number should be for this? 5-10%? If anyone has any clever ideas on looking at the statisitics of the survival of various nations/etc historically, to get some idea of how to get a more accurate number, well, lemme know. For now, will just call it 5%
so p(survival of Jews|Atheisim)=5%. Now, no need to do the math here, see, our previous evidence strength was 5:95, and now, 95:5, so they cancel out. So we can go back to where we were before either of those two, specifically, 1:75, without having to actually calculate. (though go ahead and do the multiplication if you want... it'll work out to be the same ratio.)
Now, finally, (you didn't really expect this wouldn't come up, did you?) the Holocaust. But there're three diffferent things that I'll consider here that I consider holocaust related. So, let's take a look:
First, even with the idea that even given Judasim, well, seems like god intends a buncha sufferng in the world, including all sorts of things happening to the Chosen People(tm) which still are protected enough to survive as a group..... This one can't be ignored.
That god wouldn't say "woooh, this one is a bit too much", well, I'll say p(Holocaust would have happened|Judasim)=5% max.
As for what Atheisim has to say? well, even that side of things, I imagine, wouldn't have predicted it very strongly, and would recoil a bit at the thought of it. say. 10-20%? we'll say p(Holocaust would have happened|Atheisim)=10% (if anyone has suggestions for how to get more accurate numbers for this, again, go ahead and suggest...)
so, 5:10 or 1:2 is the ratio for this piece, so multiplying by where we were last, we have 1:150
But WAIT! before we continue, need to bring up something else. So, according to Judasim, Jews were slaves in Egypt, with the situation _very_ nasty and degrading. exodous happened, then Jews got Israel. yay. So? what's my point? Basically, that would likely be the closest thing to compare in ancient times for the situation of the jews to the more recent thing of the Holocaust. And yeah, at least one rabbi I asked basically said that he'd consider it a tough call which one he'd consider worse.
But, what's my point here? Oh looksey, hrm... Holocaust happens, then Jews get Israel again. Hrm..... interesssssting...
Now, wouldn't have predicted it very strongly, but still, Judasim may have at least given a semihigher number... let's say, 20% to that?
so p(Jews getting Israel shortly after Holocaust|Judasim)=20%, vs, well, wouldn't have expected it all that much in an Atheisitic universe, so, say, 5% no more? Or maybe someone has some better suggestions for the numbers... but, well, 20:5 -> 4:1, so where were we? multiplying that gets us to 4:150 or, simplifying, 2:75
BUT WAIT! hold on your horses here, we're not done with the holocaust yet...
See, if we're comparing it to the biblical situation of the Jews in Egypt, well... One of the hallmarks of that story is the big flashy miracles all over the place. It wasn't just that they got out of egypt, you've got the plauges, the splitting of the sea (commentaries say that it was 12 channels, one for each tribe) the cloud of glory/pillar of fire, etc etc etc...
Why all the big flashy obvious miracles instead of hidden? partly, it's said that Jews were so degraded, physically and spiritually there that what was needed was a bit more than just "okay, you're out of egypt now"
Now, let's look at the holocaust. no big flashyness, just eventually the war was won...
Just wait one dang minute here... Frequently god is compared to fathers or kings.
Now, imagine you're a parent (or don't imagine it if you really are one) Your kids get kidnapped, tortured all sorts of ways, some killed, degraded and humiliated in all sorts of ways... And then, you maybe send someone to free them, but don't even bother saying hi to them afterwards? I don't think so. Hidden miracles don't cut it here. Nuh uh. If the exodus happened at all in the way Torah described it, then you'd _really_ have expected to see something similar for the holocaust. For those that are going "well, subtle miracles, blah blah blah", all I can say is "go learn a bit about the holocaust"... There's a reason why "nazi" has become a semi univesal epithet for ultimate evil.
Nuh uh... after something like that, one'd expect that god would make it _VERY_ clear to his children, somehow "yeah, I know that was horrible, I'm still here though..."
But nope. You can argue till blue in the face about hidden miracles involved the jews being able to get israel after the holocaust, but IMNHO, one'd have expected more.
So, I'll say, oh, p(lack of flashy holocaust ending/post holocaust miracles|Judasm)=5%, while, in an atheistic universe, you really wouldn't expect miracles at all, right? but sometimes things happen by chance. so p(lack blah blah|atheism)=99% at least. probably ought be 99.9999999999999 or somesuch, but we're just working in integer values, and being generous in favor of judasim. so how's it do?
Where were we? 2:75? so, multiplying, 10:7425 or 2:1485, or... given the above considerations, with extremely favorable massaging of numbers in favor of Judasim, we still end up with Atheism being 742.5 times as likely to be true as Judasim.
Ouch.
And I hadn't even touched age of the earth/evolution/other science vs religion stuff.
Note, I may update this based on suggestions for more acccurate numbers or other bits of evidence, so I am still open to the possibility that some pieces of evidence will shift the numbers the other way so strongly that I may end up doing the whole black hat/etc thing. But as things stand now....
Discuss. |
|
|
| navigation |
| [ |
viewing |
| |
most recent entries |
] |
| |
|
|
![[info]](http://l-stat.livejournal.com/img/userinfo.gif?v=92.1){kind=small}
psy_kosh's journal