Can theists be rational?

[Bri]

Thank you for proving my point.

Sorry, but it didn't prove your point -- unless your point is that belief in teapots orbiting Jupiter is a more rational belief than belief in extra terrestrial intelligent life. Was that your point?

I mentioned previously that evidence isn't the only kind of information that is useful to us.

Evidence is the only information helpful in determining the probability of intelligent aliens or anything else. Specifically, a belief without evidence is a faith-based belief rather than an evidence-based belief. Whether or not other information that's not evidence inspires people to pursue actual evidence to support the belief seems to be irrelevant when determining whether the belief is supported by evidence or not.

And you specifically stated that the new evidence to be used for the argument for God (i.e. the value to be plugged in) is evidence that the universe is fine-tuned - that is, the probability that the physical constants necessary to lead to life would have the value that they have due to chance.

I don't understand the above sentence. The new evidence (the probability that the universe is fine-tuned) is actually two variables: the probability that the universe is fine-tuned if a god exists and the probability that the universe is fine-tuned if there is no god. In the form of Bayes theorem used by the argument, these values are represented by P(E|H) and P(E|~H).

Again, here is the form of Bayes theorem used in the argument:

                    P(E|H) * P(H)
P(H|E) = ---------------------------------------
         [ P(E|H) * P(H) ] + [ P(E|~H) * P(~H) ]

P(H) - probability that god exists
P(~H) - probability that there is no god (1.0 - P(H))
P(E|H) - probability that the universe is fine-tuned if a god exists
P(E|~H) - probability that the universe is fine-tuned if no god

On the other hand, every value of the Drake equation, when plugged into the terms of the equation, will influence the result.

As they do in the above equation.

So my question is, how can the presence of fine-tuning be evidence of God if the posterior probability of God is unchanged regardless of whether these physical constants are common or rare?

The prior probability P(H) and posterior probability P(H|E) will be different unless P(E|H) and P(E|~H) are the same (which would indicate no fine-tuning). The amount that they differ depend on the values of P(E|H) and P(E|~H) assumed in the premise. Therefore, I don't understand your question.

-Bri

 

[yy2bggggs]

P(H/E) is a proposition.

Let's see.
From proposition^WP:

...In logic and philosophy, ... Propositions in either case are intended to be truth-bearers, that is, they are either true or false.

P(H|E) ranges from 0 to 1. It is not a truth bearer. Bear in mind that even in the most fatal cases, where P(H|E)=1, it's at best almost surely^WP true, and where P(H|E)=0, it's at worst almost surely^WP false. So even on the border cases, you need something else, completely beyond Bayes Theorem, to make the final step of judging whether or not H actually is true or false.

Also, P(H|E) is a probability that H is true given that E is... true. E is a proposition. In a Bayesian treatment, E can be true, or false. That's it. Only two values allowed, and those are your choices. P(E), on the other hand, is a probability.

None of this even matters, because this:
P(H|E) = P(H)P(E|H)/P(E)
...is an equation. It's a statement that two things are equal. It, in itself, can be considered a proposition. But it is not an argument. There are no premises. There is no conclusion. There's just a value, on the left, and a value, on the right. It could be true, could be false, could be true in certain
circumstances, but it cannot be circular, because it's not an argument.

Do not confuse degree of belief with degree of truth. Bayesian probability has nothing to do with the latter. If you want degree of truth, as I've mentioned before, try fuzzy logic^WP.

 

[fls]

Sorry, but it didn't prove your point -- unless your point is that belief in teapots orbiting Jupiter is a more rational belief than belief in extra terrestrial intelligent life. Was that your point?

Nope. Just that I'm not going to see any examples that aren't silly from you. Unless I somehow missed noticing our species' tendency to place teapots in orbit around planets.

Evidence is the only information helpful in determining the probability of intelligent aliens or anything else. Specifically, a belief without evidence is a faith-based belief rather than an evidence-based belief. Whether or not other information that's not evidence inspires people to pursue actual evidence to support the belief seems to be irrelevant when determining whether the belief is supported by evidence or not.

Can you give me an example of someone who believes there are intelligent aliens?

I don't understand the above sentence. The new evidence (the probability that the universe is fine-tuned) is actually two variables: the probability that the universe is fine-tuned if a god exists and the probability that the universe is fine-tuned if there is no god.

Nope. Because you have no information on those variables. The fine-tuning argument consists of noticing that the physical constants that allow life to be present here are very unlikely to have occurred by chance. That is, the entire argument rests on the very unlikeliness of this set of physical constants - p(E).

Now that has led people to propose that there is something that will make those constants more likely - a fine-tuner. And from that one forms assumptions about probabilities in the presence of a fine-tuner and in its absence. But the idea wouldn't gain any traction if the set of physical constants that support life was common and likely - i.e. if there was no fine-tuning.

In the form of Bayes theorem used by the argument, these values are represented by P(E|H) and P(E|~H).

Again, here is the form of Bayes theorem used in the argument:

                    P(E|H) * P(H)
P(H|E) = ---------------------------------------
         [ P(E|H) * P(H) ] + [ P(E|~H) * P(~H) ]

P(H) - probability that god exists
P(~H) - probability that there is no god (1.0 - P(H))
P(E|H) - probability that the universe is fine-tuned if a god exists
P(E|~H) - probability that the universe is fine-tuned if no god

Exactly. There's no place to put p(E) in this equation. So the universe could be non-fine-tuned and leave the argument unchanged.

As they do in the above equation.

How does the probability of fine-tuning due to chance change the result in the same way that the fraction of planets able to support life does?

The prior probability P(H) and posterior probability P(H|E) will be different unless P(E|H) and P(E|~H) are the same (which would indicate no fine-tuning). The amount that they differ depend on the values of P(E|H) and P(E|~H) assumed in the premise. Therefore, I don't understand your question.

-Bri

It makes no difference to the statements that you've made above whether or not this particular set of physical constants is common (no fine-tuning) or rare (fine-tuning). So how can anyone say that fine-tuning serves as evidence for God?

Linda

 

[Darat]

...snip... So how can anyone say that fine-tuning serves as evidence for God?

Linda

Faith.

 

[Bri]

Nope. Just that I'm not going to see any examples that aren't silly from you. Unless I somehow missed noticing our species' tendency to place teapots in orbit around planets.

It was silly on purpose, because it serves as a counter-example to your point that "it matters how far away (and I don't mean physical distance) something would be from our own experience."

I assumed you would agree that it's not more rational to believe in teapots orbiting Jupiter than in intelligent aliens, yet teapots orbiting Jupiter are not "far away" from our own experience at all (unless you really DO mean physical distance).

Can you give me an example of someone who believes there are intelligent aliens?

I would imagine that most of the people involved with SETI are of the opinion that intelligent aliens are probable, or they probably wouldn't be spending millions of dollars to communicate with them.

Nope. Because you have no information on those variables.

Incorrect. We have information that the universe is fine-tuned and therefore would have evidence to support a probability of that occurring by chance. Therefore, P(E|~H) would be based on evidence.

The fine-tuning argument consists of noticing that the physical constants that allow life to be present here are very unlikely to have occurred by chance.

Exactly. That probability would be P(E|~H) in the equation.

That is, the entire argument rests on the very unlikeliness of this set of physical constants - p(E).

Yes. The argument lies on a low value placed on P(E|~H) in the premise.

Now that has led people to propose that there is something that will make those constants more likely - a fine-tuner. And from that one forms assumptions about probabilities in the presence of a fine-tuner and in its absence. But the idea wouldn't gain any traction if the set of physical constants that support life was common and likely - i.e. if there was no fine-tuning.

Again, yes.

Exactly. There's no place to put p(E) in this equation. So the universe could be non-fine-tuned and leave the argument unchanged.

What you're calling P(E) is actually P(E|~H) in the equation (the probability that the universe is fine tuned by chance if there is no god).

How does the probability of fine-tuning due to chance change the result in the same way that the fraction of planets able to support life does?

If it's low, it causes the result to be high. If it's high, it causes the result to be low. It's an inverse relationship, but it can either raise or lower the result. If it's the same as P(E|H) then it won't change the prior probability at all. If it's higher than P(E|H) it will lower the prior probability.

It makes no difference to the statements that you've made above whether or not this particular set of physical constants is common (no fine-tuning) or rare (fine-tuning). So how can anyone say that fine-tuning serves as evidence for God?

Of course it makes a difference. If there is no fine-tuning P(E|~H) will be high, thereby lowering the results.

-Bri

 

[yy2bggggs]

The fine-tuning argument consists of noticing that the physical constants that allow life to be present here are very unlikely to have occurred by chance.

Exactly. That probability would be P(E|~H) in the equation.

Actually, the probability is only P(E|H') in the equation if you set up a false dichotomy.

H' isn't a hypothesis--it's the complement of one. H union H' must be U--in Bayesian terms, it should be the universe of possibilities.

 

[fls]

It was silly on purpose, because it serves as a counter-example to your point that "it matters how far away (and I don't mean physical distance) something would be from our own experience."

I assumed you would agree that it's not more rational to believe in teapots orbiting Jupiter than in intelligent aliens, yet teapots orbiting Jupiter are not "far away" from our own experience at all (unless you really DO mean physical distance).

Ah, then we are in the habit of sending teapots into orbit around planets. Thanks for the information - I had no idea.

I would imagine that most of the people involved with SETI are of the opinion that intelligent aliens are probable, or they probably wouldn't be spending millions of dollars to communicate with them.

So my idea that aspirin may help prevent strokes is a faith-based belief?

Incorrect. We have information that the universe is fine-tuned and therefore would have evidence to support a probability of that occurring by chance. Therefore, P(E|~H) would be based on evidence.

Exactly. That probability would be P(E|~H) in the equation.

The fine-tuning argument doesn't give us a value for p(E/~H). Let 'x' be this particular set of values for the physical constants (which we all agree is a tiny portion of the set of all values for the physical constants ('n')) and 'E' be the set of all values for the physical constants that will support life. The fine-tuning argument says that x/E is one or is close to one. If the universe isn't fine-tuned, then the ratio would be smaller - closer to x/n.

So the fine-tuning argument would modify the value for p(E) by some fixed amount, but this would apply equally to the p(E/H) and the p(E/~H). Whether this ratio is close to one or close to x/n, the result for (p(H/E)) is unchanged.

Linda

 

[Bri]

Actually, the probability is only P(E|H') in the equation if you set up a false dichotomy.

H' isn't a hypothesis--it's the complement of one. H union H' must be U--in Bayesian terms, it should be the universe of possibilities.

It's not a false dichotomy -- it's just a dichotomy. P(E|~H) is the probability that the universe is fine-tuned if there is no god. P(E|H) is the probability that the universe is fine-tuned if there is a god.

What other possibility is there besides a god or not a god?

If you think it likely that there is some natural explanation of a fine-tuned universe, then you would give P(E|~H) a higher value and the result would be different. That's a valid objection to the premise.

-Bri

 

[Bri]

Ah, then we are in the habit of sending teapots into orbit around planets. Thanks for the information - I had no idea.

I believe I said the opposite. Which is what makes the conclusion that such a belief would be more rational than other beliefs which are more "far away" from our own experience.

So my idea that aspirin may help prevent strokes is a faith-based belief?

No, there is plenty of evidence to support such a belief, making it an evidence-based belief rather than a faith-based belief.

Let 'x' be this particular set of values for the physical constants (which we all agree is a tiny portion of the set of all values for the physical constants ('n')) and 'E' be the set of all values for the physical constants that will support life. The fine-tuning argument says that x/E is one or is close to one. If the universe isn't fine-tuned, then the ratio would be smaller - closer to x/n.

Actually if L is the set of constants that support life and A is the set of all constants including those that don't support life, evidence of a fine-tuned universe would indicate that L/A is small. That would probably be the value you would place into P(E|~H) in Bayes theorem.

So the fine-tuning argument would modify the value for p(E) by some fixed amount, but this would apply equally to the p(E/H) and the p(E/~H). Whether this ratio is close to one or close to x/n, the result for (p(H/E)) is unchanged.

Evidence of fine-tuning would give us some idea of P(E|~H) as shown above. The premise of the argument assumes that P(E|H) would be much higher (the probability of fine-tuning would be higher if there was a god). If you disagree with that premise by making P(E|H) lower, you would get a different result.

-Bri

 

[fls]

I believe I said the opposite. Which is what makes the conclusion that such a belief would be more rational than other beliefs which are more "far away" from our own experience.

Wouldn't that make orbiting teapots "far away" from our own experience?

No, there is plenty of evidence to support such a belief, making it an evidence-based belief rather than a faith-based belief.

But I already told you that there isn't evidence to support my idea, just "tools".

Actually if L is the set of constants that support life and A is the set of all constants including those that don't support life, evidence of a fine-tuned universe would indicate that L/A is small. That would probably be the value you would place into P(E|~H) in Bayes theorem.

Well, even if we had that evidence (which we don't), it still wouldn't be confined to ~H.

Evidence of fine-tuning would give us some idea of P(E|~H) as shown above. The premise of the argument assumes that P(E|H) would be much higher (the probability of fine-tuning would be higher if there was a god). If you disagree with that premise by making P(E|H) lower, you would get a different result.

-Bri

I've already said that we agree with the premise that life is more likely in the presence of a god/fine-tuner, than in its absence. I'm just pointing out that the portion of universes that are fine-tuned can be varied without altering that relationship. The p(E/H) can be presented in terms of the p(E/~H) and so any changes effectively cancel out.

Linda

 

[Bri]

Wouldn't that make orbiting teapots "far away" from our own experience?

Farther away than aliens? How so?

But I already told you that there isn't evidence to support my idea, just "tools".

What are you talking about? There is actual evidence to support the idea that aspirin helps prevent strokes. If there weren't, the American Heart Association wouldn't recommend aspirin to help prevent strokes.

Let's pretend for a moment that there were no evidence that aspirin helps prevent strokes. In that case, if you believed that aspirin helps prevent strokes but have no evidence, then of course that is a faith-based belief.

Well, even if we had that evidence (which we don't), it still wouldn't be confined to ~H.

Actually, yes we do have evidence to support a probability of a universe fine-tuned by chance. But there is little or no evidence to support a probability of a fine-tuned universe if a god exists (it can reasonably be assumed to be higher, but it wouldn't have to be).

I'm just pointing out that the portion of universes that are fine-tuned can be varied without altering that relationship.

It sounds like you're assuming multiple universes here (please correct me if I'm wrong). The argument for fine-tuning doesn't assume multiple universes, and the amount of fine-tuning is reflected in the probability of P(E|~H). If the universe is very fine-tuned (if a small range of values of the constants can support life), the probability of the fine-tuning occurring by chance diminishes. If the universe isn't fine-tuned (if a large range of values of the constants can support life), the probability of fine-tuning occurring by chance is increased.

The p(E/H) can be presented in terms of the p(E/~H) and so any changes effectively cancel out.

P(E|H) is actually independent of P(E|~H). They are two different probabilities. I'm not sure what you mean by changes "effectively cancel out." It is not the case that if one is high the other must be proportionally smaller. For example, you might think that even though the universe is fine-tuned a natural explanation is probable, in which case you might set P(E|H) close to (or even lower than) P(E|~H).

-Bri

 

[articulett]

aargh!!

We can speculate about what sort of things that might exist or be true based on stuff we know to exist or be true--we cannot speculate about poorly defined invisible immeasurable entities that "fine tune" universes or whatever-- for the same reason we can't speculate about unicorns making the stars in little magical ovens.

A god is so real to a believer that they just cannot see how a god from an objective perspective is as "imaginary" as IPUs... they are fanciful... we have no means to speculate about how (or how many) non physical undetectable entities are interacting with our world in any way. Why? Because THEY ARE COMPLETELY undetectable-- they are all indistinguishable from IPU's.

Aspirin exists... relief of symptoms exist... these things are defined... whether aspirin helps prevent something is a falsifiable claim... it doesn't involve anything that is indistinguishable from imaginary things. Using this as analogy betrays your flaw in thinking... as all your analogies do.

But your god is, Bri... all gods are. Your argument could be used to justify belief in any woo or any story someone might want to posit for whatever really cool thing they don't understand. That makes it a really, really self-serving argument and a totally unconvincing one... unless you have a need to believe.

 

[RandFan]

We can speculate about what sort of things that might exist or be true based on stuff we know to exist or be true--we cannot speculate about poorly defined invisible immeasurable entities that "fine tune" universes or whatever-- for the same reason we can't speculate about unicorns making the stars in little magical ovens.

This is one point I don't think Bri will ever be able to get.

 

[JoeTheJuggler]

Let's see.
From proposition^WP:

P(H|E) ranges from 0 to 1. It is not a truth bearer. Bear in mind that even in the most fatal cases, where P(H|E)=1, it's at best almost surely^WP true, and where P(H|E)=0, it's at worst almost surely^WP false. So even on the border cases, you need something else, completely beyond Bayes Theorem, to make the final step of judging whether or not H actually is true or false.

Also, P(H|E) is a probability that H is true given that E is... true. E is a proposition. In a Bayesian treatment, E can be true, or false. That's it. Only two values allowed, and those are your choices. P(E), on the other hand, is a probability.

None of this even matters, because this:
P(H|E) = P(H)P(E|H)/P(E)
...is an equation. It's a statement that two things are equal. It, in itself, can be considered a proposition. But it is not an argument. There are no premises. There is no conclusion. There's just a value, on the left, and a value, on the right. It could be true, could be false, could be true in certain
circumstances, but it cannot be circular, because it's not an argument.

Thanks for clarifying some of this. Good clear post.

Do not confuse degree of belief with degree of truth. Bayesian probability has nothing to do with the latter. If you want degree of truth, as I've mentioned before, try fuzzy logic^WP.

So pretty much what I've said all along is a good assessment of the Forster & Marten argument. (That's it's circular.) If you start with a prior probability for the existence of God, then use as "new" evidence the assertion that a universe with life is more likely if God exists (and that this universe is a universe with life), then run it through Baye's Theorem you get a conclusion of a higher probability of God's existence. You have to start with a probability of God's existence in order to reach the conclusion of a probability of God's existence.

Neither probability (the prior probability or the conclusion) says anything about the truth value of the proposition "God exists".

cj's describing it as a rational argument for the existence of God is not a good assessment.

 

[JoeTheJuggler]

The Fine Tuning argument:

If things were different than they actually are, they wouldn't be the same. Therefore God exists.

 

[articulett]

How is it that all the roads just happen to go to all the important places and right up to everyone's home? It must be part of some grand plan...

How do the markets just happen to have the brands and foods and stuff that people buy in the right quantities? How can the internet be so finely tuned that no matter what language you speak it understands you and sends your message to others.

By golly, there must be some magical higher intelligence "front-loading" the info in from the get go! I'll call this magical designer, "Zeus".

Hail Zeus. So what's the probability that Zeus exists? Why is it different for god?

We cannot use magical entities or explanations to describe that which we don't understand. To do so is a Humongous and irrational leap of faith. People do it, because they are trained to... encouraged to... indoctrinated... but your god is no more probable than my Zeus when it comes to reality, Bri. Is my Zeus as probable as some material alien intelligence?

Definitely not.

You can't use a problem to "invent" an answer that is a bigger unknowable than the problem it's supposed explain. Yet all gods are exactly that. They are bigger mysteries and impossibilities than whatever it is they are said to explain. If a fine tuned universe is hard to understand, than a magic guy who fine tunes universes so that humans could evolve on one little speck of rock is immensely more improbable. For the same reason "Zeus" is.

 

[Beth]

we cannot speculate about poorly defined invisible immeasurable entities that "fine tune" universes

This is one point I don't think Bri will ever be able to get.

Probably because it's a point where articulett is wrong and Bri is correct. Evidence? This very thread contains hundreds of posts where we are doing exactly that.

 

[articulett]

Excuse me, I meant to say it's irrational to speculate about such things... as irrational as speculating about matrix scenarios and wormhole bigfoot visitors and body thetans causing engrams.

But irrational people CAN and do speculate about all kinds of invisible, unfalsifiable, immeasurable stuff. I call them woos.

 

[Malerin]

P(H|E) ranges from 0 to 1. It is not a truth bearer.

Sure it is. (H/E) is a conditional proposition.* Conditional propositions are either true or false. They don’t range from 0 to 1. The epistemic value we assign to it (Pr(H/E)) is our degree of belief in the conditional proposition. In cases where H is logically necessary, (H=all green emeralds are green), it doesn’t matter what E is. (Pr(H/E)=1)). When H is logically impossible (H= a non-green green thing exists), again, no amount of evidence can change it’s value (Pr(H/E)=0)).

If (H/E) could not be a truth-bearer, (H) could not be a truth bearer, for nearly all (or all) of our (H)’s are really of the type (H/E) and the (E) has become part of our background knowledge. Example:
H=“the coin in my pocket is a fair coin”. This is a proposition. It is either true or false. I think it very likely that the coin in my pocket is a fair coin (Pr(H)>>.5)). But that is just to say, “Based on what I know (my background knowledge), I think it highly likely the coin in my pocket is a fair coin because all such coins have turned out to be fair coins (Pr(H)k>>.5)). That is equivalent to “Given that all the coins in my pocket have turned out to be fair (E), I think it highly likely that the coin in my pocket is a fair coin
(Pr(H/E) >>.5))”.

Bear in mind that even in the most fatal cases, where P(H|E)=1, it's at best almost surelyWP true, and where P(H|E)=0, it's at worst almost surelyWP false. So even on the border cases, you need something else, completely beyond Bayes Theorem, to make the final step of judging whether or not H actually is true or false.

It depends on H. If H is “a non-green green thing exists”, then Pr(H/E) is surely zero. A non-green green thing can never exist. If H is "the square I just drew has 4 sides", then Pr(H/E) is surely 1. It will always happen that any square I draw has 4 sides, no matter what E is.

Also, P(H|E) is a probability that H is true given that E is... true. E is a proposition. In a Bayesian treatment, E can be true, or false. That's it. Only two values allowed, and those are your choices. P(E), on the other hand, is a probability.

E is evidence, not a proposition. E is neither true nor false, it just is. E is just a part of reality that we think supports a particular hypothesis (coin tosses, Mercury’s orbit, die rolls, etc.) It is expressed in proposition form (e.g., “The cat is on the mat”, “The coin landed heads“), but that doesn’t mean the cat is on the mat has a value. H is not confirmed or disconfirmed with propositions. It is confirmed or disconfirmed with evidence, which we express as a proposition.

None of this even matters, because this:
P(H|E) = P(H)P(E|H)/P(E)
...is an equation. It's a statement that two things are equal. It, in itself, can be considered a proposition. But it is not an argument. There are no premises. There is no conclusion. There's just a value, on the left, and a value, on the right. It could be true, could be false, could be true in certain
circumstances, but it cannot be circular, because it's not an argument.

It is circular because it is the fallacy of circular definition (I may have called it circular reasoning before, which is wrong). If P(H/E) is defined as P(H/E) = P(H)P(H/E)/P(E), then it would be circular, in the same way that dog = animal that looks like a dog is circular.






* http://books.google.com/books?id=AK...&hl=en&sa=X&oi=book_result&resnum=1&ct=result

 

[Malerin]

Excuse me, I meant to say it's irrational to speculate about such things... as irrational as speculating about matrix scenarios and wormhole bigfoot visitors and body thetans causing engrams.

But irrational people CAN and do speculate about all kinds of invisible, unfalsifiable, immeasurable stuff. I call them woos.

You'll definitely like this then:
http://www.simulation-argument.com/

But it is not an irrational argument.