|
Users viewing this topic:
none
|
|
Login  | |
|
RE: Constant balance - 11/25/2008 3:56:06 PM
|
|
|
robto
Posts: 103
Joined: 10/10/2006
Status: offline
|
quote:
ORIGINAL: BookerG
A few words on probability and the anthropic principle: Probability is not a measure of likeliness, but a measure of one’s knowledge about likeliness. Three forecasters could say the probability of snow tomorrow is 10%, 50%, and 100%, and they could all be absolutely correct.
All knowledge plays in to the probability, including assumptions, even faulty ones.
You can say that if you like, but for the probablility to be meaningful, we want our assumptions to reflect reality as closely as possible.
quote:
There are two problems raised by the anthropic principle. First, we know we exist, so the probability is automatically one. You can get rid of that knowledge by assuming the question was asked by someone else, posit an objective observer who would ask the question whether we exist or not.
[snip]
If our objective observer asked what are the odds that bipedal, self-aware creatures would exist on planet earth, the odds are incredibly low.
If he asked, what are the odds that intelligent life capable of asking the question would exist anywhere, the odds are higher, but still incredibly low.
If he asked about the odds of single-cell life forms...if he asked about any self-organizing, self-replicating objects that could fit a broad definition of life, that raises the odds considerably. But I'd still bet against it.
Of course if the observer has observed us, then he too will conclude the probability is one. I think what you meant to ask was the probability the observer would assign to the existence of life on a random, earth-like planet. But then we are back to the question of what assumptions the observer is able to make. Perhaps he has observed 1000 earth-like planets and found life on 999 of them. Then he will give a high probability to life on the new planet. Perhaps he has no such observations, and is going purely on theoretical assumptions. Then the probability he will assign will depend heavily on what sort of theory he is using. If, for example, he knows of laboratory experiments that show molecular evolution can occur for a wide range of conditions, then he might again assign a high probability to life on the new planet.
So, I think your claims here are ignoring your own point about the need for assumptions in estimating probability.
_____________________________
The Theory of Almost Everything: Everything you always wanted to know about elementary particles but were afraid to ask
|
|
|
|
|
RE: Constant balance - 11/25/2008 7:02:57 PM
|
|
|
BookerG
Posts: 91
Status: offline
|
"You can say that if you like, but for the probablility to be meaningful, we want our assumptions to reflect reality as closely as possible."
If I’m reading you correctly, then you’re saying it’s pointless to imagine our universe’s existence as anything other than the certainty that it is. Every reality has a probability of one. Every non-reality has a probability of zero. For the very concept of probability to be meaningful, we need to assume an imperfect or incomplete knowledge of reality. And to distinguish between a random universe and a fine-tuned one, we need to eliminate as much knowledge and as many assumptions as we possibly can. Okay, another analogy. I think the lottery is rigged. The last ten times, the winner has been a family member of the guy who does maintenance on the number-picking machine. But now that the law is becoming suspicious, he won’t do it again. We have no proof or evidence other than the past lottery results and whatever probability can tell us.
You seem to be saying that the reality is that his sister won, with a probability of one. His brother won with a probability of one. His father, mother, uncle, daughter-in-law’s second cousin twice removed all won with a probability of one. Therefore it was a sure thing. He’s innocent. That’s making probability meaningful by making our assumptions reflect reality as closely as possible. Assume they all won, since they did. Therefore no evidence or likelihood of tampering.
If you want to figure out the likelihood of tampering, or fine-tuning, you have to ignore the knowledge of what happened and focus on the probabilities of what could have or should have happened.
The multiverse claims to solve the fine-tuning problem, but it only goes half-way. It says, let’s assume we didn’t know that we exist (That’s the only way probability calculation makes any sense). The probability of fine-tuning one universe might be very low, but multiply it by an infinite, or nearly infinite number of universes, and the probability skyrockets. (In our analogy, that’s like assuming that everyone in this guy’s family bought a million dollars worth of lottery tickets every week. It makes the results less surprising; but it’s a pretty big assumption).
I’m saying, take it one step farther. Don’t just ignore the knowledge that we exist. Ignore the knowledge that it’s possible for us to exist. (In the analogy, I’m saying, don’t just consider how many tickets the family members might have bought. Also consider the possibility that, if there were no tampering, they wouldn’t have bought any tickets, or there wouldn't even have been a lottery with any winners. Reality says, they must have bought a ticket because they won. But a court of law wouldn’t assume it. Show us your receipts. Name the store. Have you ever bought tickets before? Could you even afford to buy the tickets?)
My “outside observer” is someone who knows nothing about any specific existing universe, but he’s smart enough to figure out what universe will result from any arbitrary selection of universal constants. We’re imagining that the lottery hasn’t been run yet, the tickets haven’t even gone on sale yet. He’s not starting with the assumption that we’re here. He’s also not starting with the assumption that if I just find the right combination of constants, I know some life-supporting universe must be possible. The only knowledge he has is the theoretical requirements for life and the theoretical constants. That’s the only way we could possibly come up with any meaningful probability of fine-tuning.
I can’t tell you what probability he’d end up with. Yes it is philosophical. You and I would come up with very different expectations, and there’s no way of knowing who’s right. My initial question only involves one issue; does a multiverse raise the odds of our existence from infinitesimal to a sure thing, or, as I see it, from an infinitesimal to a not quite as infinitesimal probability. It can not be proof positive, not based on probabilities, but it could end up being proof beyond a reasonable doubt.
But I can’t resist one last analogy. I’m going to pick a number the way that observer would narrow down the universe selection. I’ll start with the requirements of life. The number must be even. The number must be a perfect square. The number must be one greater than a prime. The number must have the digit sequence 579 in it somewhere. As I list the requirements, the percentage of numbers that simultaneously meet all of them dwindles. If I started out with a specific number in mind, I might eventually get to the point where only one out of a billion billion numbers will work; or maybe only a single number out of infinity. But if my process was truly random, then I couldn’t have a number in mind to start with. And by the time I get to the fifth or tenth or twentieth requirement, I’m likely to find that no number fits them all.
The fact that you could narrow down the possibilities to one in a billion billion without accidentally eliminating all possibilities, I think, very strongly suggests that a number was preselected.
The multiverse only explains our existence if you preselected a workable universe before narrowing things down.
< Message edited by BookerG -- 11/25/2008 8:05:56 PM >
|
|
|
|
|
RE: Constant balance - 12/1/2008 11:13:13 AM
|
|
|
robto
Posts: 103
Joined: 10/10/2006
Status: offline
|
The problem with your lottery analogies is that we all know how a lottery is SUPPOSED to work: some device generates a random number between (say) 1 and 999,999,999, in such a way that each value has an equal probability of being selected. You are ASSUMING that something similar is true of the universal constants of nature. But we don't know that that is true: we don't know if ANY of the constants can vary, or what range they can vary in, or what sort of a probability distribution that variation might have.
Here's a better analogy: One day, a stranger walks up to your door and hands you a suitcase containing a million dollars in cash.
Is this a low probability event or a high probability event? Your immediate reaction is "low probability", because it has never happened to you or anyone else you know before. But in some cases, it might not be low probability at all. Suppose you have a rich distant relative who is an admirer of yours, who left you this money in his/her will. In that case, there is only one person in the world who that money could go to, and it's you. The probability is one that you will get it.
On the other hand, the money might come from an organization that picks someone at random to give a million dollars to. If they pick someone from the whole US, then the probability is very low that it would have been you. But if they pick someone from your town, then the probability is much higher. If they pick someone from your street, the probability is even higher.
My point is that you can't talk about probabilities at all unless you have some idea of the "probability space": the range of options and the likelihood of those options being chosen.
Here's a variation on an example you used before. Suppose I say I'll give you a million dollars if you pick the right number between one and 12. Your first thought is that I'll pick the number randomly, with equal probabilities. Then it doesn't matter what number you choose. But if you know that I'm using dice, then you certainly won't pick one (zero probability of being correct). So you pick seven, since that's the highest probability with two dice, right? Wrong! It turns out that my dice have six dots on each face, so the only outcome possible is 12 - probability one!
Again, the point is that we don't know which scenario we are in with respect to the universe. We don't know what values are possible for the constants. Even if we knew what values were possible, we wouldn't know what probabilities to assign to them.
The same is true of your unbiased observer, unless he has access to multiple universes from which he can draw inferences about the ranges of possible values and their probabilities. The difference between us and him is that we KNOW that there is one possible set of values: the set we observe. And we KNOW that that set is amenable to life. But that's all we know: we don't know if those values are inherently high probability or low probability, or even if "probability" makes sense with regard to those values.
One more way to look at it: suppose we were having this discussion 100 years ago. You might point out that it's crucial to life that oxygen's atomic weight is 16 times that of hydrogen. What if oxygen were only 12 times hydrogen, or 25? In such a universe, life would be impossible!
That argument might have seemed reasonable back then. But today, we know that hydrogen has one proton, and oxygen has 8 protons and 8 neutrons. We can't change their atomic weights without adding/subtracting protons (which would make it a different element) or adding/subtracting neutrons. But adding or subtracting neutrons gives us unstable isotopes of oxygen. O12 and O25 are highly unstable and decay rapidly, so there's no way to imagine a universe in which oxygen is 25 times heavier than hydrogen. The masses are not independent: if you change the mass of the proton, you change BOTH the mass of hydrogen AND that of oxygen.
Things that seem like "fundamental" constants today might be in a similar bind. If we knew more, we might realize that, say, the gravitational constant and the speed of light cannot be varied independently. Or maybe they can't be varied at all. We just don't know. You can't base a probability argument on complete ignorance. It's necessary to know SOMETHING in order to make any kind of estimate of the probabilities.
_____________________________
The Theory of Almost Everything: Everything you always wanted to know about elementary particles but were afraid to ask
|
|
|
|
|
RE: Constant balance - 12/1/2008 5:45:00 PM
|
|
|
BookerG
Posts: 91
Status: offline
|
First, I’ve said from post one that I’m not trying to calculate the probabilities. Can’t be done. But maybe it’s possible to categorize the probability in rough terms as either probable, reasonably unsurprising, improbable, or incredibly, mind-bogglingly unlikely.
Analogy. I ask two people to pick a number. What are the odds they will pick the same number. I can’t tell you the probability. I didn’t give them a range. I have no idea what distribution curves either person will use. Either they’ll pick the same number or they won’t. Two possible outcomes, with no knowledge of either one’s likelihood. That should mean the odds are 50/50. Instead, I’d categorize it as improbable, and the only reason this example doesn’t fall into the mind-bogglingly unlikely category is that they are likely to have similar thought processes about what range they should use.
But when applied to the fine-tuned universe, the selection methods should be completely independent (the life-requirements selections should be completely independent of the universe’s method of choosing its constants). It’s like one person picking his number by dividing his social security number by his zipcode, and the other person multiplying his telephone number by his birthday. The fact that zipcodes are not distributed evenly, that some are impossible and some are small towns while others are megalopolises, doesn’t really have any effect at all on the probability that the two numbers will match, because any theorized distribution curve, or limited range, or even a single possible value, is just as likely to hurt as to help (actually it's more likely to hurt). What is the probability of a match? I don’t have a clue, but it sure ain’t 50/50.
You keep trying to up the odds by lowering the range of possibilities. To me, that seems to betray the same train of thought that made multiversers think they’ve solved the fine-tuning problem. It only helps if you are of the mindset that naturalism is the only possible answer, and therefore the fact that we know that our universe exists means that we know that it must fall somewhere in the range of universes that naturalism could produce by a random selection of constants. The more universes there are, or the narrower the range of possible universe, the more likely it becomes that you would end up with ours. My outside observer was intended to help you think outside the box. What if we weren’t inevitable? What if no randomly selected set of constants would result in a life-supporting universe? The universe doesn’t care if we exist, so why should it provide any range of possible universes that includes one with life? Neither a multiverse nor a smaller range of constants would raise the odds of life, unless we already won an incredible lottery just by the existence of the potential of a fine-tuned life-producing universe.
Whereas, I’m of the mindset that maybe the lottery was rigged. Maybe there’s a God who not only chose the right set of constants to fine-tune the universe, but he also predetermined the whole concept of constants in such a way that a right set was possible. Maybe a fine-tuning of reality was necessary before a fine-tuning of constants within reality had any hope of producing life.
So I really don’t care at all about the probability of our universe, compared to the probability of all the other possible universes. I care about the probability of a reality in which life is one of the possibilities compared to the probability of realities in which even an infinite number of universes would not result in one that has life. Now, it’s really impossible to calculate those odds. But if there were a way, it would start, not with the range or distribution of gravitational values or proton weights, but with all of the requirements of the complexity of life and specifically intelligent life, and whether or not it’s likely or it should even be possible to fulfill all of those requirements simultaneously.
One last thought that was sparked by your suggestion of linking the constants. I believe scientists have wrestled for decades with the odd fact that gravity is so much weaker than the other forces, something like 35 orders of magnitude weaker. Maybe science will find a reason. Maybe they already have. But whatever that reason is, it wasn't naturalistically established in order to make life possible. I'd say we won our first lottery right there. If a random, uncaring quality of reality itself conspired to make gravity so weak, another random, uncaring quality of reality could just as easily have made one of the other forces 35 orders of magnitude stronger than expected, and no adjusting of the constants by random selection would overcome that handicap.
|
|
|
|
|
RE: Constant balance - 12/2/2008 4:09:18 PM
|
|
|
robto
Posts: 103
Joined: 10/10/2006
Status: offline
|
quote:
ORIGINAL: BookerG
First, I’ve said from post one that I’m not trying to calculate the probabilities. Can’t be done.
Agreed.
quote:
But maybe it’s possible to categorize the probability in rough terms as either probable, reasonably unsurprising, improbable, or incredibly, mind-bogglingly unlikely.
Analogy. I ask two people to pick a number. What are the odds they will pick the same number. I can’t tell you the probability. I didn’t give them a range. I have no idea what distribution curves either person will use. Either they’ll pick the same number or they won’t. Two possible outcomes, with no knowledge of either one’s likelihood. That should mean the odds are 50/50. Instead, I’d categorize it as improbable, and the only reason this example doesn’t fall into the mind-bogglingly unlikely category is that they are likely to have similar thought processes about what range they should use.
But when applied to the fine-tuned universe, the selection methods should be completely independent (the life-requirements selections should be completely independent of the universe’s method of choosing its constants). It’s like one person picking his number by dividing his social security number by his zipcode, and the other person multiplying his telephone number by his birthday. The fact that zipcodes are not distributed evenly, that some are impossible and some are small towns while others are megalopolises, doesn’t really have any effect at all on the probability that the two numbers will match, because any theorized distribution curve, or limited range, or even a single possible value, is just as likely to hurt as to help (actually it's more likely to hurt). What is the probability of a match? I don’t have a clue, but it sure ain’t 50/50.
Having just declared calculating probabilities to be impossible, you go on to try to do it anyway! The problem with all these analogies is, we don't know if the origin of the universe is "like that" at all. No argument from analogy can work when we have no idea of what sort of process we are trying to analogize.
quote:
You keep trying to up the odds by lowering the range of possibilities. To me, that seems to betray the same train of thought that made multiversers think they’ve solved the fine-tuning problem. It only helps if you are of the mindset that naturalism is the only possible answer, and therefore the fact that we know that our universe exists means that we know that it must fall somewhere in the range of universes that naturalism could produce by a random selection of constants. The more universes there are, or the narrower the range of possible universe, the more likely it becomes that you would end up with ours. My outside observer was intended to help you think outside the box. What if we weren’t inevitable? What if no randomly selected set of constants would result in a life-supporting universe? The universe doesn’t care if we exist, so why should it provide any range of possible universes that includes one with life? Neither a multiverse nor a smaller range of constants would raise the odds of life, unless we already won an incredible lottery just by the existence of the potential of a fine-tuned life-producing universe.
Whereas, I’m of the mindset that maybe the lottery was rigged. Maybe there’s a God who not only chose the right set of constants to fine-tune the universe, but he also predetermined the whole concept of constants in such a way that a right set was possible. Maybe a fine-tuning of reality was necessary before a fine-tuning of constants within reality had any hope of producing life.
So I really don’t care at all about the probability of our universe, compared to the probability of all the other possible universes. I care about the probability of a reality in which life is one of the possibilities compared to the probability of realities in which even an infinite number of universes would not result in one that has life. Now, it’s really impossible to calculate those odds. But if there were a way, it would start, not with the range or distribution of gravitational values or proton weights, but with all of the requirements of the complexity of life and specifically intelligent life, and whether or not it’s likely or it should even be possible to fulfill all of those requirements simultaneously.
I agree that there are two issues here that need to be separated. What I find odd in your approach is that you have ruled out any explanation that avoids the issue of life in the universe, but you also have ruled out any explanation that invokes the fact of life in the universe (anthropic principle)! In effect, you have ruled out any sort of naturalistic explanation right from the start.
Now, I have agreed already that naturalistic explanations of the multiverse type are of little use. But I don't see how you can philosophically justify ruling out one type of explanation (naturalistic) while ruling in another type (theistic). It seems to me that the only honest approach is to admit that neither explanation is satisfactory.
_____________________________
The Theory of Almost Everything: Everything you always wanted to know about elementary particles but were afraid to ask
|
|
|
|
|
RE: Constant balance - 12/2/2008 6:05:41 PM
|
|
|
BookerG
Posts: 91
Status: offline
|
There’s an old saying, “Every analogy limps.” The point of using analogies is to take a complicated issue and distill one single thread, to solve a simple problem that helps us refine our thinking on the complex issue. So the universe is not like a lottery or a firing squad. But they help me explain my train of thought on one concept at a time to make an intractable problem tractable.
The universe is not like two people independently picking a number. But that example can explain why I think that even something that resists probability calculation can be still be called improbable. And why, when the lottery is won repeatedly by relatives of the lottery maintenance man, I think it's valid to consider a high probability of tampering, and not just say they won, therefore nothing improbable happened, a la the anthropic principle.
Whatever it is that makes gravity so much weaker than the other forces must do so, under naturalism, completely independently of a need to satisfy the requirements of life. Yet it satisfies the requirements of life. Serendipity. Improbable.
Not impossible. I have not ruled out naturalism (at least not for the sake of this debate). But I am claiming that naturalism is stacking the deck in its own favor when it pretends that the anthropic principle either makes the fine-tuning issue a meaningless philosophical triviality, or else that you can legitimately discuss the probabilities in a way that sets aside our knowledge that the probability that “this universe exists” is one (which is exactly what the multiverse implies when it is presented as a solution to the fine-tuning problem) while not setting aside our knowledge that the probability that “a life-supporting universe is possible” is one. And even if I can’t calculate the probabilities, such a broadening of the horizon of possibility lowers the odds that we are here just by lucky coincidence considerably, in the same way that picking a number between one and a billion is much harder than a number between one and ten.
Let’s just suppose that the electromagnetic force must, in every possible universe, be exactly ten to the thirty-seventh power stronger than gravity. Your argument so far, and the multiverse’s, would say, in that case our existence has just become far more probable, because we just need the right random setting of one force instead of two. I want to know why it’s not still just as improbable as ever, because if that number was 36 or 38, then life would be ruled out. You’ve just shifted the improbability into a different sphere so you can ignore it.
If you say this universe exists therefore this universe must exist, that just says I refuse to even think about the possibility of purposeful design.
The only things I've ruled out are that intelligent life is no more special than door knobs, and that we exist therefore there must a priori be a naturalistic explanation for our existence.
“I don't see how you can philosophically justify ruling out one type of explanation (naturalistic) while ruling in another type (theistic). It seems to me that the only honest approach is to admit that neither explanation is satisfactory.”
Are you admitting that naturalism is unsatisfactory? Has there been anything in this long discussion that says theism has trouble explaining the fine-tuning? If all you’re saying is that neither side can satisfactorily disprove the other, granted. But I sure wish scientists weren’t so quick to assume naturalism is the only game in town, not just from the point of view of a methodological philosophy, but as fact and the only valid description of reality.
|
|
|
|
|
RE: Constant balance - 12/6/2008 6:34:17 PM
|
|
|
robto
Posts: 103
Joined: 10/10/2006
Status: offline
|
"I want to know why it’s not still just as improbable as ever..."
There is a mathematical reason behind my argument. The chance of hitting a 1 centimeter target out of a 1 meter range is 1/100, assuming the probability is spread equally over the 1 meter. If there are two variables, then we have to hit a 1 square centimeter target out of a range of 1 square meter = (100 centimeters)x (100 centimeters) = 10,000 square centimeters. So the chance is now 1/10,000.
My real point was not this mathematical argument, though, but just to point out that with so many unknowns that can conceivably affect the probabilities there is just no way to even make a reasonable guess about the likelihood involved. You can declare "I have a hunch that the likelihood is really small," but that's not very persuasive.
Let me turn the tables for a minute: here's the anti-fine-tuning argument.
The universe is almost entirely devoid of life. By far, most of the universe consists of empty space, with maybe some tenuous clouds of hydrogen and helium drifting about, completely incapable of supporting life. Most of what remains, apart from empty space, is stars. These are far too hot to support any kind of organized life. Look at our solar system. 99.9% of its mass is in the sun, which, as we just said, can't support life. Of the remaining 0.1%, by far the majority consists of gas giants, planets with no hard surface and as far as we know incapable of supporting life. Certainly they can't support life as we know it. In fact, only the earth, a tiny fraction of a percent of the mass of the solar system, is capable of supporting life.
Clearly, the creator of the universe did not design it to be hospitable for life. We can easily imagine other universes packed edge to edge with life in an infinite variety. In fact, our universe seems fine tuned to EXCLUDE life to an almost unimaginable extent.
I would say that this conclusion, that the creator intended to make the universe as inhospitable to life as possible, is just as reasonable as your conclusion that the universe is fine-tuned for life.
"Are you admitting that naturalism is unsatisfactory?"
I am admitting that there is no satisfactory naturalistic explanation for the origin of the universe (i.e. the pre-big bang state), or for the particular values of universal constants that we observe.
"Has there been anything in this long discussion that says theism has trouble explaining the fine-tuning?"
No, in fact, theism has no trouble explaining any possible set of observations at all. One need only say "God wanted it that way" and the discussion is over. Needless to say, I don't find such "explanations" at all satisfactory.
But what does theism explain, exactly? It doesn't explain why we have this particular set of universal constants, rather than some other set that allow for life to exist. There is still an infinite number of possible values, even if the range of values is small. So, the probability of this particular set of universal constants is still infinitesimal. All theism can say is that the constants must be such that allow for life. But the anthropic principle says the same thing, without the need to postulate a personal creator. So theism is no better off than the anthropic principle in explanatory power.
I would like to know if there are any predictions made by theism that have been tested and found experimentally correct.
I have to add that I have been enjoying our exchange immensely. It is such a refreshing change from the usual display of vitriol in S&O. (And I hope that last remark doesn't discourage you from participating on this board more often!)
_____________________________
The Theory of Almost Everything: Everything you always wanted to know about elementary particles but were afraid to ask
|
|
|
|
 New Messages |
 No New Messages |
 Hot Topic w/ New Messages |
 Hot Topic w/o New Messages |
 Locked w/ New Messages |
 Locked w/o New Messages |
|
 Post New Thread
Reply to Message
Post New Poll
Submit Vote
Delete My Own Post
Delete My Own Thread
Rate Posts |
|
|
Forum Tools
Printable Version
