My Photo

The Out Campaign

Atheist Blogroll

Blog powered by Typepad
Member since 05/2005

« Jerry Coyne on "Can Atheism Be Proven Wrong?" | Main | Sex in the City, But Lost in the Desert: Sex and the City 2 »



I think it may be that you could hypothetically be persuaded of the *truth* of a religious claim whereas PZ is arguing that he could never be persuaded to *worship* under any religious paradigm.

I.E, PZ seems to argue that if something exists, it's a real hypothesis about the world, but that if a deity required *worship* he would consider that a moral failing and refuse.

I'm not exactly convinced everybody is arguing the same thing.

Greta Christina

I don't think that's what PZ is saying, Jemand... but I agree that we're not arguing the same thing. I think PZ is arguing that, in any practical sense, based on an overwhelming preponderance of evidence to date, we can comfortably discard the god hypothesis: Jerry and I are arguing that, while we agree that this is so, our atheism still ought to be provisional and falsifiable. PZ is arguing the pragmatic approach, we're arguing the philosophical one.

Or, to put it another way: I think PZ's point is, "Religion has thus far utterly failed to show even a scrap of good evidence supporting its claims" (with him so far). "Therefore I feel entirely comfortable rejecting it as a hypothesis" (still with him). "Therefore, I don't have to accept even the fragment of hypothetical possibility that I might someday be shown to be mistaken." Which is where he loses me.


So far as I can tell PZ's point seems to be that, not only is it the case that no God thus far has had any evidence to support it, but that any kind of God capable of being supported by evidence would not be the sort of entity we would wish to call 'God'. It is almost as if it is part of the definition of God that it is the sort of thing that one must have faith in rather than have evidence for. That is what makes God different from, for example, some really advanced alien. If we were to discover that were were created by aliens, we would not call them Gods precisely because they are the sorts of things that we could potentially study and understand. If you were to try and think up an entity who's existence could possible be supported by evidence, you would necessarily be describing an entity that is arguably not God.


I take PZ's point differently. It seems to hinge on "how godly does something need to be to count?"

If 'god' = 'anything that can be worshiped' then we can declare that gods definitely exist (people can worship mountains).

If 'god' = 'things with super-powers' then the proofs are trivial; Superman would count as a god.

If 'god' = 'actual omnipotent thing' then proofs are hard, as we can't differentiate omnipotent from mere super-potency. This seems to be what PZ is arguing.


From back at the original post that PZ linked to when this thing blew up, the thing that leapt out at me was a kind of argument from information theory/complexity:

Omniscience, as a property, means that an entity with this property is more complicated than our entire universe. From an information-theory kind of angle, where in order to know everything that is happening and has ever happened and will ever happened, the entity has to be at least as complex as everything that's ever happened put together. By being literally the most information rich and therefor complex thing imaginable, a god becomes the least probable thing you could possibly propose to explain anything.

Now if you accept that premise, there's suddenly no evidence that is best explained by god. Cognitive errors and/or aliens become more plausible explanations for anything you might experience. Anything at all.

I'm sure that my grasp of information theory is inadequate to judge whether the premise is sound, but it sure is interesting.

Greta Christina

Jules: Who said anything about omniscience? I actually agree that the Omniscience, Omnipotent, Omnibenevolent god of most Christianity is a logical impossibility, and I think it's arguable that any sort of omniscience -- or any sort of omni-anything -- is logically impossible. But lots of religions haven't posited an omniscient god or gods. It isn't a requirement of religion.

Villa: I agree that, to some extent, we're arguing about terminology. Part of PZ's argument seems to be that any and all definitions of god or gods are either (a) trivial (i.e., it could be super-powerful but still physical, like super-intelligent space aliens), (b) logically impossible (like the Omnimax god), or (c) falsifiable, and has already been falsified (like the god who created the universe 6,000 years ago) and has already been falsified, but whose followers are stubbornly still believing.

I agree that all gods to date have fallen into these categories. And I agree that the possibility of a religion appearing with a coherent, testable god hypothesis -- including a coherent definition of what "god means" -- that's well-supported by solid evidence is vanishingly small... so small that, for all practical purposes, we can dismiss it and move on, the way we've dismissed the "flat earth" hypothesis and moved on. But again, there's a difference -- trivial in practical terms, very important in philosophical and rhetorical ones -- between "I am rejecting this hypothesis" and "I will never accept this hypothesis no matter what happens."

Daniel Schealler


Sorry if I'm being a pain. I just put through a long-ish post on Information Theory that disappeared when I came back to check on it.

Is it in moderation or something? Don't want to double-post if it's still hanging around.

Greta Christina

Daniel: I'm not seeing it in the spam folder. It might have gotten eaten. By a god, probably. :-) Sorry. Please do try again.

Daniel Schealler


Thanks. Call me paranoid, but I've started copying my longer posts to notepad before I hit send.

That's probably a bad sign of something or other. Next I'll be wearing my underpants on my head and collecting stray cats.



I did a bit of Information Theory at Uni as mart of my BSc in CompSci.

I can't remember *any* of the math now, of course. I couldn't differentiate a curve right now without looking it up online either. Fuck integration.

But from what I do remember: Whatever a person is arguing may very well be interesting. But if it doesn't involve an actual bit-string and some math at some point? Whatever else the person is arguing about, it isn't actually Information Theory.

That's not to suggest that PZ and Steve don't get Information Theory. PZ in particular has provided some indications in the past that he actually understands Information Theory better than some CompSci majors - which isn't hard at all, btw.

And even Steve at least tries, in words, to tie his argument back to the level of data.

However, Information Theory relies on certain assumptions about data and the medium and encoding of that data. The whole point of the God argument (to me) is that there is something like a mind that exists outside of any given storage or transmission medium.

If the 'information' isn't actually stored as data in a medium (which I believe to be impossible, btw) then there is no bit-string associated with that data. If there's no bit-string, Information Theory doesn't apply.

To put it another way: Information Theory is all about analysing the encoding of information. If information itself is never encoded, then Information Theory has nothing to work with.

This is of course absurd - from everything we know, information cannot meaningfully exist outside of some form of physical encoding medium. But again - therein lies the rub. That is precisely the nature of the God/soul claim in the first place.

So Steve/PZ's argument adds another straw to the pile of reductio ad absurdum arguments that naturally follow from most of the grander God claims: Not only does the existence of God violate thermodynamics, but it also violates everything we know about Information Theory as well! Great!

But to be honest - violating thermodynamics was enough for me. That particular camel's back is already amply broken as far as I'm concerned. It's interesting and all, but this does feel like a mountain that's been made out of a molehill.

Let me know whether or not I'm making sense - I've been trying to work on my technical communication skills. Constructive critical feedback would be lovely. ^_^


So... it seems you are saying that being atheist involves the rejection of hypothetical, non-existent religions, and that it's falsifiable because one of the hypothetical non-existent religions might hypothetically produce proof of its validity?

Whereas I (and perhaps Meyers, I think) would disagree that atheism includes disbelief in hypothetical non-existent religions - I don't see how it would even make sense to reject a hypothetical, unarticulated hypothesis.

Greta Christina

People in this debate, just letting you know: I'm not ignoring you. I'm finding this fascinating, it's making my head spin in a really good way. But I need to take a break from it for the moment, as I have deadlines and whatnot. Please keep the discussion going if you're so inspired; I'll get back in on it, either here in the comments or in a separate post, as soon as I can.


This debate is beginning to remind me of debates in my ex-religion on whether Jesus was sinless because he *couldn't* sin or if he just *didn't* sin. Or if he was inherently sinless or if it was just his actions. Or what ever that even meant.

Aw hell. lol *THIS* debate is rather more interesting than that one, but still... there are more similarities than I'd like lol.


@Daniel - Thanks, that was very interesting. I was hoping for some commentary from a slightly better informed person.

I can totally see where you're coming from when you say that a god already doesn't fit within the laws of physics, but I still like this as a different angle.

It's so easy to think anthropomorphically that one can miss how ridiculously big a leap it is to go from "I don't know" to "goddidit". It's a little bit different with non-theistic religions, but it's a little bit the same too: e.g. How does karma actually work?

@Greta: I understand Abrahamic religions to account for almost half the world's population, so I naturally assumed that's what we're talking about. This was clearly wrong in a conversation about whether any religion could ever be true :(

I think the question it led me to remains valid, though: Could there be any evidence for a religion that is not better explained by error, madness or deceit, even before you invoke aliens or sci-fi?

For that matter, why do crazy paranoid theories sound... well... crazy and paranoid, while an invisible alternate reality with optional intelligence seems reasonable to so many people?

That leads to an interesting thought experiment: What evidence for any religion could not be generated by a follower of that religion with a mind-control device?

Descartes' devil with an agenda and a sci-fi human skin, if you will ;)

Daniel Schealler



Do not get me started on the notion of karma. It's bad in ways no-one ever seems to notice. Between the notion of sin and the notion of karma, karma is the lesser of the two evils, but not by much.

*pulls on reins*

Okay. Off topic now. Moving back to IT.

There could be an argument in the IT thing if you really wanted to work with it, I suppose.

1) If we presume that the 'information' that comprises God is in fact encoded in a medium somehow, then Steve Zara's original argument applies.

2) But if we assume the opposite? If the 'information' that comprises God is not encoded at all? This violates everything we know about Information Theory.

3) If we assume that the universe itself is the encoding? That gets us to a vague deism/pantheism that might work with IT. But I can't see people wanting that kind of thing.

For example: How does this 'God' make decisions? If we move matter around, are we changing God's mind? Or is it just God playing very complicated solitaire, and we're the cards? If it's all information in the mind of God - then what is heaven if not more of the same?

That's potentially interesting, now that I think about it.

All the same, the consequences of that view wouldn't sit well at with most people who want to argue in favor of an intervening monotheist deity. Still don't see them going for it.

4) If, on the other hand, God doesn't exist at all? Then there was never a problem in the first place. Clean and simple.

The reason I find this line of attack a bit uninspiring is that I can't see that believers would have any problem in claiming that 2) implies that everything we know about Information Theory is wrong. Especially since most people don't really know what IT is in the first place.

You're saying that some vague math-y thing a bunch of egg-heads came up with that I don't understand says that God can't exist? Well - too bad for your egg-heads then! Have you taken Jesus into your heart?

Blegh. I'm getting cynical in my old age. Well... more cynical.


@Daniel: I'm afraid I have to disagree. Information theory certainly talks about encodings a lot, and all information it describes is encodable, but the whole point is that that it can meaningfully measure and talk about information independent of the specific encoding chosen.

Indeed, this is a strength of PZ's argument; following yours, all we can say is that the theory does to apply outside our physical world, and therefore your argument can be ignored. Following his, we can make statements about an omniscient god without knowing its nature (i.e. encoding).

Now, IT, as a branch of mathematics, is fundamentally just a game with made-up rules; it's internally unfalsifiable, and the surprise is "the unreasonable effectiveness of mathematics". in describing the natural world.

But by advancing a complexity-based argument, a theist is implicitly admitting the applicability of IT to their god. In other words, they chose the weapon, so can't complain about the fairness of the duel.


In the above, second paragraph, that should be "the theory does not apply outside our physical world." Somehow my fingers typoed "not" to "to".


I know I'm way late to the party but the reason I largely agree with PZ here is that you can't prove to someone, something that cannot even be quantified. A god or gods simply fills in a gap in human understanding and the goal post moves every time you try to get to it. Something happening that is inexplicable or even amazing is still not proof of a god. If it were, then every supposed miracle would prove the existence of a god.

The problem is really that the conclusion is flawed. The idea that the answer to the question is a vague undefined version of a god that one small portion of mankind invented. That the answer to the question of "how" that happened is something that not even two believers in the same religion can agree on, seems suspect and frankly, unscientific. It is possible but not probable.

I'd be more inclined to say that there is the possibility that something greater, more advanced, larger or otherwise beyond our personal understanding. Again, this is not a god, simply something more than what we have ever experienced to date. The minute you throw up your hands and declare that it's a god, is the moment you stop looking for an answer that can expand your understanding of the world around you.

Daniel Schealler


Sorry, late back to this thread - never got a notification email.

I don't understand why people fail to understand this.

You've entirely missed my argument.

The specific encoding isn't particularly important. I'm in full agreement there.

But there still must be an encoding, or else IT has nothing to work with and it's all just hand-waving.

Let's take a look at a real example:



Sorry, late back to this thread - never got a notification email.

I don't understand why people fail to understand this.

You've entirely missed my argument.

The specific encoding isn't particularly important. I'm in full agreement there.

But there still must be an encoding, or else IT has nothing to work with and it's all just hand-waving.

Let's take a look at a real example:


The use of symbolic measures of complexity to derive quantitative characterisations of temporal structures in time series is proving potent for a variety of areas. Such techniques have particular attraction in the medical field, where clinical diagnosis and/or application of therapies is reliant on electronic patient monitoring systems. This paper explores the application of a relative new comer to the area of non-linear measures, using T-entropy computed from electroencephalogram (EEG) and electrooculargram (EOG) signals to derive indicators of sleep state.


What Titchener did here was to take the EEG (measure of electrical activity) linear readout as well as the EOG (measure of eye-motion) and encoded both as linear readouts.

He then took those linear readouts and sampled them over time to digitize them into bit-strings.

He then applied algorithms of IT to those bit strings to work out how much Information was encoded therein at any given point in time.

He then compared the change in Information over time between the two surfaces, and found that certain patterns correlated with sleep-states.

None of which could have been done without the very important step of creating a bit-string first.

I find it very frustrating when people treat IT as if it was just a-priori logic and handwaving, like Rene Descartes trying to understand the bit bad universe by sitting in a dark room, rubbing a piece of wax and thinking to himself.

IT is a practical application of mathematics within Computer Science. It deals with evaluating real-world bit-strings to diagnose and solve real-world problems.

IT is very much connected to reality - and the reality is that IT cannot meaningfully talk about Information that is not encoded in some way.

The specifics of the encoding may not be important. ASCII vs. Unicode doesn't matter in the slightest. If we look to Lempel-Ziv or T-codes we can see that IT actually allows us to construct our both our encoding and our alphabet on the fly! The specifics of the encoding are indeed a side-issue.

But all the same, try for me now. Transmit the information in a blog post without encoding it first.

You can encode it as ASCII bytes, you can encode it as ink on paper, you can encode it as chiseled characters in stone, or you can encode it as a modulated pressure-wave in the air around you by straining your vocal chords just right.

But what you won't be able to do?

You won't be able to transmit 'information' without encoding it first.

And you won't be able to measure or meaningfully talk about information that is not presently encoded in some medium using Information Theory.

The comments to this entry are closed.

Subscribe/ Donate to This Blog!

Books of mine

Greta on SSA Speakers Bureau

  • Greta Christina is on the Speakers Bureau of the Secular Students Alliance. Invite her to speak to your group!

Your email address:

Powered by FeedBlitz

Powered by Rollyo

Some Favorite Posts and Conversations: Atheism

Some Favorite Posts and Conversations: Sex

Some Favorite Posts: Art, Politics, Other Stuff