•
•

## Diogenes and Satyricon debate Bayes

• Posted by a hidden member.
Hey guys! So, TigerTim and I had a great facebook chat about whether my Bayesian formulation of the supernatural-realm problem was acceptable. TigerTim wanted me to play the shrink to diagnose his problems on why he was reluctant to accept a Bayesian approach. We thought it might be good to post what we have left of the chat for you guys to see (fb cuts off your chat history past a certain length, unfortunately).

Some background info:
--Omega (loosely speaking) is the set of events that have a probability of occurring in a certain context
--Normalization is making sure that the probabilities add up to 1

(TigerTim presents a hypothetical where three guys are in prison, and you know one of them is a murderer)

1:44pm(TigerTim)
;-)
ooo
so
you do some investigating

1:45pm(Satyricon331)
and find?

1:45pm(TigerTim)
and blah blah
you find out some exciting evidence
like....
one of them is a Democrat
;-)

1:45pm(Satyricon331)
lol

1:46pm(TigerTim)
now here is an important question actually about what is admissable evidence or not etc

1:46pm(Satyricon331)
i was just about to say, we'll assume everything is admissible, since that ev would fail under Rule 403 (edit: and probably 404 since it’s a murder case)

1:47pm(TigerTim)
which is tantamount to characterizing the peakedness of the "evidence dist. funct"
and moreover
1) the possibility space is "tom" "dick" or "harry"
2) the normalization constraint is P(tom) + P(harry) + P(dick) =1

1:49pm(Satyricon331)
ok

1:49pm(TigerTim)
now
somebody told you that it could be "someone else"
like me
i'm like "i bet it was that mysterious guy red_vespa"
and you're like
WHOA!
*idea*
now
as soon as you have that
you can (arrest him) and modify your possibility space
P(tom) + ... + P(red_vespa)
but
until you do so

1:52pm(Satyricon331)
well, that "someone else" information means that your previous axiomatization was false - you specified omega incorrectly

1:52pm(TigerTim)
ahhh good point
so the underlying point is though
how to deal with two different (intersecting) subgroups of a single group of possibilities.
i mean for example
dark matter is?
P(WIMPS)+P(Supersymmetric partners) etc
but someone comes up with a new possibility
so again there's a new possibility
the point is

1:55pm(Satyricon331)
ah, i think your problem isn't on how to deal with two different intersecting subsets of a single omega

1:55pm(TigerTim)
until you have a theory, a formulation, a notion of "consistency wth expt"
you can't apply bayes's theorem to P(something else)

1:55pm(Satyricon331)

1:55pm(TigerTim)
hmmmm
ok
so mr psychiatrist
my specification of omega
what do i do? :-)

1:56pm(Satyricon331)
(lol) i would say that when you axiomatize omega, you have to consider it a proposition itself, and consider the likelihood of *its* truth

1:56pm(TigerTim)
ooooo
errrrr
ooooo
*interesting*

1:57pm(Satyricon331)
when you work within an axiom set, you assume its prob is 1

1:57pm(TigerTim)
but
WAIT
it's prob is either 1 or 0

1:57pm(Satyricon331)
when you consider the aciom set's implications, however, you then have to consider the probablity that the axioms are false

1:57pm(TigerTim)
hmmmm
ok
no you're still fine

1:58pm(Satyricon331)
no! consider the state space, {axiom set is true, at least one axiom is false}

1:58pm(TigerTim)
ok

1:58pm(Satyricon331)
you *know* that this state space will hold since it is tautological
(at least under the logical axioms!)

1:59pm(TigerTim)
mmm
hmmm
yes

2:00pm(Satyricon331)
so you have new information that an even occurs that your axiomatization of omega didn't include, you update your probability that the axiomatization is true to zero
ex ante, however, you may have mixed feelings - or perhaps no information at all

2:01pm(TigerTim)
ok

2:01pm(Satyricon331)
it's only when you examine the axiomatization's implications do you proceed *as if* its prob were 1

2:03pm(TigerTim)
that ontological problems have the character of the validity of axioms
or axiom spaces

2:06pm(Satyricon331)
am i back online?
it seems so!

2:07pm(TigerTim)
phew

philosophy is like heroin

2:07pm(Satyricon331)
hahaha

2:07pm(TigerTim)
i need a dealer :-P
anyway
did you get my last message?
that ontological problems have the character of the validity of axioms
or axiom spaces

2:08pm(Satyricon331)
yes, and i think so
ontological problems ultimately are propositions, and you have to treat them like propositions

(TigerTim)
hmmmmm
ok
i think i am satisfied
but wait
presumably you would agree that the question "which axi
• Posted by a hidden member.
I think that made me fall asleep for a bit... then i woke up at the bottom. My pants were gone, wallet had no cash and i was in a bathtub full of ice with no kidney...
• Posted by a hidden member.
oh guys, this is interesting stuff!

it cut off the end, so here's the rest of the post...

(TigerTim)
hmmmmm
ok
i think i am satisfied
but wait
presumably you would agree that the question "which axiom space is correct" is not formulatable unless a space of axiom spaces is supplied

2:15pm(TigerTim)
so we have in fact made a restriction

2:15p(Satyricon331)
i'm reluctant to asnwer that question directly since i don't know the mathematics of "normalizing" sets with cardinalities greater than continuum... but... (edit: I should add, that the Bayesian consideration is of each axiom set individually, so i don't think that considering the space of axiom spaces is necessary - and I don't think it;s actually a well-posed option anyway, but that's outside the thread's scope...)

2:15pm(TigerTim)
we are ONLY allowed to apply bayesian arguments
to the question "does this space contain the correct answer"?

2:15pm(Satyricon331)
yeah, i would only ever use bayesian args

2:16pm(TigerTim)
in that case i am satisfied
my expectation is that such sets are NOT normalizable
but yes
i am satisfied

2:17pm(Satyricon331)
YAY!
• Posted by a hidden member.
Will this help me get through calc?
• Posted by a hidden member.
I don't even smoke and I need a cigarette after that. That was hot.
• Posted by a hidden member.
You guys *need* to get laid.
• Posted by a hidden member.
• Posted by a hidden member.
Oh Satyricon, I miss you :-)

--- Diogenes the Cynic
• Posted by a hidden member.
well at least they are making some more productive Crack these days or whatever the kids are calling it now...

lol...just kidding.
• Posted by a hidden member.
fluxu8 saidwell at least they are making some more productive Crack these days or whatever the kids are calling it now...

lol...just kidding.

This one time someone like said 'Are you on crank?'. I was like 'Wtf?!'. Then I like punched his neck off. And like, that's The End.
• Posted by a hidden member.
ok
• Posted by a hidden member.
P(The world is going to end) = ∑{P(TigerTim will destroy the world), P(Satyricon will destroy the world)}

P(The world is going to end) = {(P(TigerTim is crazy and will destroy the known Universe) * P(TigerTim is crazy))/P(The world will end anyway)} + {(P(Satyricon is crazy and will destroy the known Universe) * P(Satyricon is crazy))/P(The world will end anyway)}

P(The world is going to end) = {(P(TigerTim is crazy and will destroy the known Universe) * P(TigerTim is crazy))/(P(TigerTim is crazy and will destroy the known Universe) * P(TigerTim is crazy))+(P(TigerTim is crazy but the world will not end)*P(TigerTim is not crazy))} + {(P(Satyricon is crazy and will destroy the known Universe) * P(Satyricon is crazy))/(P(Satyricon is crazy and will destroy the known Universe) * P(Satyricon is crazy))+(P(Satyricon is crazy but the world will not end)*P(Satyricon is not crazy))}

...

• Posted by a hidden member.
We should be grateful to them for debating something more substantive than He Who Must Not Be Named.
• Posted by a hidden member.
That's just the point, the debate is either substantive or not depending upon its position in an axiom space, without that................... . . . .. .. . . ...

TexDef07 saidWe should be grateful to them for debating something more substantive than He Who Must Not Be Named.
• Posted by a hidden member.
Hopefully I am not setting the cat amongst the pigeons by cross-posting this item. However, I found it very interesting.

Satyricon331 says (in reference to TigerTim):

"I think our disagreement stems partly from something you’ve overlooked – I already identified god as an object that has a connection to our realm but nonetheless requires the existence of another, super-natural realm. It’s not a complete definition of course, but it does supply a testable prediction, namely that our realm is not self-contained."

Agreeing that your discussion revolves around the specific hypothesis of the existence or non-existence of an empirically-provable supernatural realm, my doubt centers on why the supernatural realm would be necessary in order to apply inverse probability to this problem.

Let me apologize at this point and confess that my highest level of mathematics education is high-school algebra 1, which I failed. Therefore, I am able to read the theory and notionally understand it, but the depth of my understanding can hardly be compared to either of yours.

01. Supernatural Realm does not exist.
02. God does not exist
03. Belief in God is equatable to belief in misidentified naturally occuring phenomena (effect).
04. Belief in God as a supernatural being is equatable to belief in effect.
05. Inverse probability can be applied to hypothesis of cause in complex systems.
06. Both cause and effect are knowable; effect even being easily provable.
07. Causality is, by virtue of cosmological scale, massively complex.
08. Even misidentification (transposing God and Effect in complex Causality) exists within an empirically-knowable data set (Universe sans supernatural realm).
09. It should be possible to posit God(s) = Complex Causality and obtain proof.
10. Always considering cosmological scale, knowability and even proof do not demystify causality.
11. The equilibrium between benefit and destructive force resulting (effect) from belief in God(s) = Complex Causality ought to be equally knowable and measurable within the same closed data set (Universe sans supernatural realm). (I chose the word equilibrium because it represents my belief that such an equilibrium does exist).

Please don't eviscerate me for this non-scientific progression. I said, a priori, that I am an uneducated boob.

On to something I know better.

Anecdotal Evidence:

When the weather doesn't truly suck (as it does now), it is often my habit to make the walk up to the Sanctuary of the Madonna of Monte Berico. The walk itself is pleasant and I focus my thoughts on the Madonna.

Perhaps I am delusional. I don't think so, but I will accept the possibility. My construct of this Goddess is that she represents all life force. Therefore, when I offer my awe to her, I am offering my awe to life force on a cosmological scale.

It is not my actual belief that she is Mary mother of God. That belief is, in fact, too small to fit a cosmological construct so vast as she. However, I can accept the concept that my projection of a Goddess representing the existence of all life could give birth to other Gods who represent absolution.

Therefore, when I think of the Christ, the Buddha, Muhammed, or the pantheon of Greek and Roman Gods (I find a polytheistic belief system that is largely centered on ancient Gods more satisfying than its modern streamlined equivalent) I render them on the cosmological scale (close enough to supernatural - by the way - given my own insignificance in such an enormous system of causality - but without being supernatural - actually natural is more satisfying, more intimate) as the Gods of absolution.

Absolution is important. To me, represents the provable fact that the stupid things I have done are ultimately (and even immediately) forgivable when considered on a cosmological scale.

Admittedly, my theist construct is semantic, even semiotic. However, my theist construct exists within the absolute same data set (Universe sans supernatural plane) as the anti-theist construct and, for that matter, the hyper-theist construct. Certainly as a part of that data set, causality is finally discernable in my belief construct and mine is no more or less relevant than anyone else's construct. My theism, my Madonna, my belief in life on a cosmological scale, and my acceptance of absolution from whichever deity is willing to dispense it are all integral to a cosmological whole that is ultimately finite, but infinite in inasmuch as I am able to perceive with the toolset available to me.

Terry

p.s. On the thread that shall not be named, OW makes reference to the Daimon - in this I do believe as well.

• Posted by a hidden member.
Hey Terry –

I’m not sure I follow your sequence of points 1-11. If you meant it as a restatement of my discussion with TigerTim, I don’t think it captures our argument.

To state it informally, it merely observes that concepts of the supernatural world, and concepts of gods or other supernatural beings, necessarily require that our world admits non-material causation (otherwise [and I didn’t develop this point explicitly before], the traditional monotheistic God is just a high-tech alien like Stargate’s Goa’uld). Empirical evidence does not support the existence of non-material causal mechanisms in our world, however, and therefore the supernatural realm likely* does not exist (since by definition it has some connection to our world that allows non-material causal mechanisms to operate here), and similarly for gods.

*(This point is where Bayesian inference came in.)

Your theistic construct has deities existing without a supernatural realm. Two points:

1) My discussion with TigerTim would have relevance to your beliefs only if you believe that these deities’ “finitely discernable” causality is non-material – in other words, they operate through “magic” (for want of a better term since something like “supernatural power” is not available) rather than through something that really high-tech aliens could accomplish.

2) If you believe that they have non-material causal mechanisms at their disposal, then since you’ve dispensed with a separate supernatural realm, your construction seems to suggests (although it doesn’t necessarily) that non-material causality occurs more often than it would if these beings existed in a separate realm that had connections to our own. If so, then they would be less likely to exist than a traditional, monotheistic deity unless for some reason they’re less active.

~~~~~
TigerTim (on another thread): ” Oh, Satyricon, I miss you and your intellectualism.

-- the reincarnated Diogenes the Cynic”

Oh, TigerTim, I <3 you! I miss our discussions too. I’m trying to get more free time in my life, but law school saps the life out of you, and this semester I have an internship and I’m studying for the MPRE and the bar on top of everything. Frankly, I hate law school, although I like my internship so far and most of my papers.

On that note, to take a completely irrelevant turn for those in the know… the Bluebook’s creators deserve gruesome execution. I finished one of my two papers yesterday. I thought I would have had it done by last week, but the bluebooking took a week. Seriously. I had hundreds of footnotes to cite, and somehow I naively though it would take a day. Not once the Bluebook had its say!

I have about ten pages left to write by tomorrow for the other essay.
• Posted by a hidden member.
Satyricon331 said

On that note, to take a completely irrelevant turn for those in the know… the Bluebook’s creators deserve gruesome execution. ... I had hundreds of footnotes to cite, and somehow I naively though it would take a day. Not once the Bluebook had its say!

Believe it or not, the Bluebook is actually your friend, if a tyrannical one. With no uniform system of citation, everybody would have to make up their own. Imagine the chaos.

BTW, don't worry about hating law school. You'll hate being a new associate even more. Good luck.
• Posted by a hidden member.
Hey Terry –

I’m not sure I follow your sequence of points 1-11. If you meant it as a restatement of my discussion with TigerTim, I don’t think it captures our argument.

Nope, most emphatically did not seek to restate your argument.

To state it informally, it merely observes that concepts of the supernatural world, and concepts of gods or other supernatural beings, necessarily require that our world admits non-material causation (otherwise [and I didn’t develop this point explicitly before], the traditional monotheistic God is just a high-tech alien like Stargate’s Goa’uld). Empirical evidence does not support the existence of non-material causal mechanisms in our would, however, and therefore the supernatural realm likely* does not exist (since by definition it has some connection to our world that allows non-material causal mechanisms to operate here), and similarly for gods.

*(This point is where Bayesian inference came in.)

OK the only reference to your argument that I made is to say that the proof or disproof of the existence of deities does not, necessarily, require non-material causation. Deities can have far more banal explanations.

Your theistic construct has deities existing without a supernatural realm.

Two points:

1) My discussion with TigerTim would have relevance to your beliefs only if you believe that these deities’ “finitely discernable” causality is non-material – in other words, they operate through “magic” (for want of a better term since something like “supernatural power” is not available) rather than through something that really high-tech aliens could accomplish.

No, I said the opposite.

I said that causality on a cosmological scale is indistinguishable from magic unless you happen to be a really high-tech alien.

No purpose would be served by participating in a discussion (or posting one for that matter) which is of relevance only to the participating parties.

All I said was (without obfuscatory terminology attached) your argument presupposes the necessity of something which is actually unnecessary, thus overcomplicating something which has far simpler explanations.

2) If you believe that they have non-material causal mechanisms at their disposal, since you’ve dispensed with a separate supernatural realm, then your construction seems to suggests (although it doesn’t necessarily) that non-material causality occurs more often than it would if these beings existed in a separate realm that had connections to our own. If so, then they would be less likely to exist than a traditional, monotheistic deity unless for some reason they’re less active.

Again, that is not what I said. What I said is that there is no supernatural realm and that there is no non-material causality. Causality is just, simply, causality. Cause and effect are complex and difficult to discern because the system is extremely large and densely complex.

These arguments are simplified, radically, by deification.

Furthermore, I said that it is a fundamental error to assume that cause and effect might function across the chasm of a supernatural realm and a natural realm. All causality is interconnected and unified in the multidimensional universe which is complicated enough already, thank you very much. No supernatural realm is required.

By the way, your original post to this thread hyperlinks to another one (which I followed and read) in which any number of arguments where put forth that you imply would be subject to "Bayesian formulation of the supernatural-realm problem".

Most people, I believe, would then break this down into a couple of relevant opportunities:

a.) One could congratulate you on the deflective curve of arcana in the FB conversation you chose to post - and by consequence celebrate your intellect in so doing.

b.) Continue the dialog upon which the reference is made http://www.realjock.com/gayforums/299293/ in the belief that your purpose was to rescue an interesting discussion from what degenerated into one of the wackier flame wars.

I chose B

My argument is very synthetic.

There is no supernatural realm. There are no deities. The universe(s) function on discernable laws of cause and effect. Things that are highly complex seem magical and are, not improbably, assigned to deities rather than cause and effect. This is not particularly harmful.

Terry

• Posted by a hidden member.
TexDef07: “ Believe it or not, the Bluebook is actually your friend, if a tyrannical one. With no uniform system of citation, everybody would have to make up their own. Imagine the chaos.”

Perhaps I should have spoken more specifically. The citation system for practitioners is fine, but Rule 2 kills me. It adds absolutely nothing. I also don’t like the short-form for journal citations… it would be easier on the reader and on the writer just to have the author present the full citation than to have both of them flipping back to find a previous footnote just so the reader can remember what the title was.

TexDef07: “BTW, don't worry about hating law school. You'll hate being a new associate even more. Good luck.”

Thanks! Somehow, at the rate I’m going, I think I’m going to hate unemployment.

~~~~~
Hey Terry, thanks for the reply.

ursamajor: “ No purpose would be served by participating in a discussion (or posting one for that matter) which is of relevance only to the participating parties.

All I said was (without obfuscatory terminology attached) your argument presupposes the necessity of something which is actually unnecessary, thus overcomplicating something which has far simpler explanations.”

I believe that good-faith debating requires people to put their opponent’s position in the strongest possible light before they attack it. If someone wants to argue that a deity is more likely than not to exist (the issue that was the subject of the conversation), then that person has the stronger position if he or she asserts that the deity resides in a supernatural realm, which is the main reason I constricted the term “deity” to such a definition, although as I said, I did not develop this point explicitly.

If the deity resides in this realm and operates with only material causal mechanisms, then its powers could be much greater than humanity’s powers are today,* but empirical inquiry has not observed these real-world deities, or their effects. We can therefore use standard inferential reasoning to conclude directly that they do not likely exist, rather than inferring indirectly that they do not likely exist due to this world’s being a closed system (see this thread for the original discussion). (As I recall, TigerTim has a great discussion somewhere on RJ, although it might have been over facebook, why we don’t believe in unicorns.) In other words, we can induct directly on the set of entities** rather than simply their effects on this world, which suggests that constricting the definition of “god” to one that in turn necessitates a separate, supernatural realm formulates the deist’s position in its strongest possible formulation. The conclusion is that even this formulation does not render a deity more likely than not.

As I emphasized in the other thread, my point is not to disparage religion, but instead to explicate my atheism and to illustrate why religious or spiritual belief requires faith rather than empirical inquiry.

As a secondary matter, most deists do not share your definition, and in order to increase the very “relevance” that concerns you, I constricted the term “god” to the formulation that the vast majority of religious persons believe – that God is something more than a Goa’uld or the like – the definition that the Christian in the clip almost certainly believes. In this respect, this definition is not necessary, but simply increases my discussion’s relevance to the vast majority of people.

Thus, no “deflective curves” occurred.

As an aside, your post admits the interpretation that you view my discussion of Bayesian inference as obfuscating, but to the contrary, it improved the discussion’s rigor, and nobody on the two threads I’ve linked to in this post seemed to have a problem understanding anything. If you didn’t intend this implication, pay this paragraph no mind.

*One limitation is that it, with overwhelming probability, could not act over cosmological distances faster than the speed of light, as most people would expect of a deity.

**Scientists sometimes offer models of the likelihood that extraterrestrial evolution will produce other intelligent life in this galaxy, TigerTim might be able to offer specific cites. I just remember reading about it on Scientific American. This article might have been one, but it’s too old to be the one I’m thinking of.

~~~~~
Another point, one major reason I had difficulty interpreting your original post, Terry, is that you did not discuss materiality explicitly. Now that you have, it’s much clearer, although at the onset it was unclear because many spiritualists or religionists believe that this very planet has non-material causation in abundance, and it wasn’t clear where you stood when you were discussing things like "a natural realm." This reason was why I tried to keep things in hypothetical “if-then” statements, so I apologize for the confusion. I was reluctant to ask explicitly because I’ve found on RJ that many people get very touchy if you dare suggest they posted something anything less than clear, but I certainly try to avoid that sin since often what I post isn’t clear either.
• Posted by a hidden member.
Satyricon331 said

As an aside, your post admits the interpretation that you view my discussion of Bayesian inference as obfuscating, but to the contrary, it improved the discussion’s rigor, and nobody on the two threads I’ve linked to in this post seemed to have a problem understanding anything. If you didn’t intend this implication, pay this paragraph no mind.

As you will have surmised, I am just having a bit of fun with your dazzling cuteness and earnestness. To that you add the pursuit of clarity in a package of civility.

Frankly, under these conditions, my proposed experiment would be a nice drunken picnic with you and T.T. to take place before either G.W. or Gordon Brown render picnicking FUBAR. I wager that by the end we would all see God and not care.

Terry
• Posted by a hidden member.
I love picnics Its more likely that it'll be Boris Johnson who bans them --- the man is a nutter! You'll have to explain to us why you're so anti-British, ursa :-)

I think its probably worth mentioning that the reason I liked this debate so much is that it's a really quite technical point which is rather inconsequential: is there a meaningful way of expressing ontological problems within the Bayesian---Popperian framework? We didn't set out to solve any great questions like WHAT IS MASCULINITY? or HOW TO FREE GAY PEOPLE. Not to say those dialogues are invalid, but the scope is so great and the constraints of the medium so severe that it's not surprising to see the fracas they generate, and the strident whining of the ignorant who are so convinced that they are right.

Instead, I had the most beautiful conversation once at a gay club here in Cleveland who is a philosophy grad student from Boston College. We talked about how we both liked pine cones. They are so elegant, and yet unnoticed by almost everyone who sees them.

I think sometimes I'm turning into a (poorer) version of G H Hardy, who famously said

Mathematicians may be justified in rejoicing that there is one science at any rate, and that their own, whose very remoteness from ordinary human activities should keep it gentle and clean.

• Posted by a hidden member.

Thanks Satyricon and Ursamajor. I enjoyed every second spent reading your replies to each other, and think I understood much of it. (total bonus!) My grandfather, passed way long years now, is an atheist and us kids would listen as he argued these kind of points with my Roman Catholic grandmother. This brought back a lot of nice feelings.

-Doug
• Posted by a hidden member.
I give the guy above me a 10.

Ok someone do me now, but be nice...

wait... what thread is this again?
• cowboyathlete

Posts: 1346
mindgarden saidYou guys *need* to get laid.
Truer words were never spoken.
• Posted by a hidden member.