[Stripe]

Originally Posted by Vendergood

Impossible.

[The Barbarian]

Do you know what 'Shannon information' is? As far as I can tell, it has nothing to do with understanding at all.

Quite so. Shannon information is a measure of uncertainty in a message.

[Stripe]

Originally Posted by chair

Do you know what 'Shannon information' is? As far as I can tell, it has nothing to do with understanding at all.

Told ya, Y.

[Yorzhik]

Originally Posted by Stripe

Told ya, Y.

You did indeed.

[The Barbarian]

I'd be interested in seeing Yorzhik's (or Stipe's) testable definition of information as it applies to populations of organisms.

Not that we're very likely to see it.

For the same reason we aren't likely to see a Yeti.

[DavisBJ]

Let me ask some questions about “information” and evolution. I have seen it said that Shannon information deals with the degradation of a message. Assume the message sender had a clear idea he wanted to communicate, and he correctly composed a succinct message with that idea. For comparison, assume a military commander identified a weakness in the enemy lines, and sent a message to “attack at this location”. Any alteration of that message would mean that commander’s wishes would not be honored. Maybe no attack would happen, maybe the wrong size of force would be employed, maybe the wrong location would be specified.

Because the message was corrupted, in the vast majority of cases a loss of function resulted. But rarely, if the coordinates were garbled, an attack might be mounted on the enemy which achieved beneficial results that the military commander had not forseen. Maybe this mistaken attack thwarted an impending enemy action, or eliminated the enemy command center, or crucially interrupted the flow of supplies to the enemy. Even though in a purely academic sense, the wrong message went out, if in the end result it turned out well, so what? In the biological equivalent, if a mistake is made in DNA that garners some advantage, then so what if information was lost in the “Shannon” sense? Is the message the important thing here, or is the result the message is supposed to facilitate the important thing?

[The Barbarian]

Let me ask some questions about “information” and evolution. I have seen it said that Shannon information deals with the degradation of a message. Assume the message sender had a clear idea he wanted to communicate, and he correctly composed a succinct message with that idea. For comparison, assume a military commander identified a weakness in the enemy lines, and sent a message to “attack at this location”. Any alteration of that message would mean that commander’s wishes would not be honored. Maybe no attack would happen, maybe the wrong size of force would be employed, maybe the wrong location would be specified.

This is not a theoretical problem. Shannon's theory is so useful and powerful, because it shows that one can reduce the uncertainty in the message by redundancy in various ways to eliminate that problem.

Because the message was corrupted, in the vast majority of cases a loss of function resulted. But rarely, if the coordinates were garbled, an attack might be mounted on the enemy which achieved beneficial results that the military commander had not forseen. Maybe this mistaken attack thwarted an impending enemy action, or eliminated the enemy command center, or crucially interrupted the flow of supplies to the enemy. Even though in a purely academic sense, the wrong message went out, if in the end result it turned out well, so what? In the biological equivalent, if a mistake is made in DNA that garners some advantage, then so what if information was lost in the “Shannon” sense?

Technically, any reduction in uncertainty is a loss of information. In evolution, it has to do with the uncertainty of sampling a specific allele in a population. The more alleles, the more information, and the greater the uncertainty of your sample. Would you like to see a simple example of how that is calculated?

Is the message the important thing here, or is the result the message is supposed to facilitate the important thing?

Understand, every favorable mutation is a mistake in transcription, and although most mistakes do little (and a few are harmful), some of them do turn out to be favorable as you suggest. And those tend to persist, because they give the recipient a leg up on surviving long enough to leave offspring.

"Information" is used by creationists as a buzzword to make it seem like a "scientific criticism." But as you see, few of them really understand it.

[Alate_One]

Originally Posted by DavisBJ

Is the message the important thing here, or is the result the message is supposed to facilitate the important thing?

In the biological sense this is all that matters, the outcome. And ultimately its only the functionality of the outcome that matters.

Because we are dealing with chemistry, rather than precise coordinates or even English words, there are many many outcomes that have the same function, and an almost infinite number of other possibilities that have a different function.

Of course you note the creationists here don't actually want to talk biology or DNA itself, they want to talk information theory whether it has any bearing on the biology or not.

[noguru]

Originally Posted by The Barbarian

I'd be interested in seeing Yorzhik's (or Stipe's) testable definition of information as it applies to populations of organisms.

Not that we're very likely to see it.

For the same reason we aren't likely to see a Yeti.

Do you really think either one of them is going to give you a testable definition of anything? They are "Hiding behind the Lord for all the wrong reasons."

[The Barbarian]

Of course. But it's a good idea to keep rubbing their noses in it.

[Stripe]

Originally Posted by The Barbarian

I'd be interested in seeing Yorzhik's (or Stipe's) testable definition of information as it applies to populations of organisms. Not that we're very likely to see it. For the same reason we aren't likely to see a Yeti.

I'd like to see Barbie show some intellectual honesty. Our position has been explained clearly enough. Perhaps he could accurately restate what our definition might be without pejorative or insertion.

[Stripe]

Originally Posted by DavisBJ

Let me ask some questions about “information” and evolution. I have seen it said that Shannon information deals with the degradation of a message. Assume the message sender had a clear idea he wanted to communicate, and he correctly composed a succinct message with that idea.

You are not talking about Shannon information if you're talking about meaning, ideas and understanding.

For comparison, assume a military commander identified a weakness in the enemy lines, and sent a message to “attack at this location”. Any alteration of that message would mean that commander’s wishes would not be honored. Maybe no attack would happen, maybe the wrong size of force would be employed, maybe the wrong location would be specified.

Because the message was corrupted, in the vast majority of cases a loss of function resulted. But rarely, if the coordinates were garbled, an attack might be mounted on the enemy which achieved beneficial results that the military commander had not forseen. Maybe this mistaken attack thwarted an impending enemy action, or eliminated the enemy command center, or crucially interrupted the flow of supplies to the enemy. Even though in a purely academic sense, the wrong message went out, if in the end result it turned out well, so what? In the biological equivalent, if a mistake is made in DNA that garners some advantage, then so what if information was lost in the “Shannon” sense? Is the message the important thing here, or is the result the message is supposed to facilitate the important thing?

If information is lost in the Shannon sense it is very possible that the meaning and non-Shannon information was enhanced. But it would require an intelligent agent to effect that change.

Shannon has nothing to do with meaning or intelligence.

Originally Posted by The Barbarian

Understand, every favorable mutation is a mistake in transcription, and although most mistakes do little (and a few are harmful), some of them do turn out to be favorable as you suggest. And those tend to persist, because they give the recipient a leg up on surviving long enough to leave offspring.

Statistically speaking, random changes must be 99.999...% harmful. The ones that do nothing are most likely harmful without immediately noticeable consequences. If a random change is not harmful (immediately or otherwise) then that is good evidence the change was intended and not a mistake.

Originally Posted by Alate_One

In the biological sense this is all that matters, the outcome. And ultimately its only the functionality of the outcome that matters.

And the analogy DavisBJ gave was a beautiful description of how impossible it would be to advance (a war or an organism) by selecting the best of the "mistakes".

Because we are dealing with chemistry, rather than precise coordinates or even English words, there are many many outcomes that have the same function, and an almost infinite number of other possibilities that have a different function.

You can say the same thing in English many different ways.

Of course you note the creationists here don't actually want to talk biology or DNA itself, they want to talk information theory whether it has any bearing on the biology or not.

Of course you'll note Alate is lying, again.

[Alate_One]

Originally Posted by Stripe

And the analogy DavisBJ gave was a beautiful description of how impossible it would be to advance (a war or an organism) by selecting the best of the "mistakes".

You keep asserting this after I provide evidence to the contrary. Are you going to address the instance of the light and dark colored mice or not? Or are you going to keep pretending it doesn't exist or that it isn't important or isn't information?

Are you going to tell me that the breeding of dogs isn't "advance" as far as a shepherd or a pet owner is concerned?

You can say the same thing in English many different ways.

EXACTLY the same way? Without different nuances or connotations being involved? There are millions of ways of encoding the exact same protein (an average few thousand amino acid one) in DNA. How many ways can you say "the quick brown fox jumped over the lazy dog"?

Of course you'll note Alate is lying, again.

You always accuse me of lying when you have no answer. You refuse to talk about actual DNA when I bring it up, you simply make your assertions over and over again. Assertions are not evidence Stripe, or did you forget that again?

Statistically speaking, random changes must be 99.999...% harmful. The ones that do nothing are most likely harmful without immediately noticeable consequences. If a random change is not harmful (immediately or otherwise) then that is good evidence the change was intended and not a mistake.

This is absolutely not true in the case of DNA. Did you forget the random changes you made to the gene a in a post while back that made no change at all at the amino acid level?

You have no idea what you're talking about, Stripe.

[DavisBJ]

Originally Posted by Stripe

You are not talking about Shannon information if you're talking about meaning, ideas and understanding.

Evolution diversifies life by introducing new variants into the gene pool. If the process I described is not “Shannon”, then I don’t see what Shannon information has to do with evolution. It sounds like a nice theoretical, or even practical, way of talking about the fidelity of a message, but as I was asking, isn’t the final outcome the important thing, even if it involves a corrupted message?

If information is lost in the Shannon sense it is very possible that the meaning and non-Shannon information was enhanced.

If we are going to discuss “non-Shannon information”, then we need a clear mutually agreed-to understanding of what that is so we will know exactly how to recognize it when we see it. In my scenario, I am not sure any kind of “information” was “enhanced”. I posited that the outcome of the process was favorably enhanced, but I don’t see that as “information”.

But it would require an intelligent agent to effect that change.

Didn’t require that in my scenario.

Statistically speaking, random changes must be 99.999...% harmful.

In DNA, I think your percentages are drastically skewed. It is my understanding that only a few percent of the DNA actually “encodes”. Much of non-coding DNA can (and does) get changed without adverse effects. Sure helps in the crime labs, and paternity testing.

The ones that do nothing are most likely harmful without immediately noticeable consequences.

If you can show me some studies in micro-biology that support your claim, I will be glad to look at them. Sans that, I will take your statement as more of what you want to be the case than anything established in science.

If a random change is not harmful (immediately or otherwise) then that is good evidence the change was intended and not a mistake.

If it was an “intended” change, then it was not random.

And the analogy DavisBJ gave was a beautiful description of how impossible it would be to advance (a war or an organism) by selecting the best of the "mistakes".

How did you get that? If a military commander achieves an unexpected advantage via a mistake, don’t you think he might consider keeping his new-found tactic as a resource? Why wouldn't an organism that benefited from a DNA "information" accident be more likely to survive and pass its new advantage on to its offspring?

Of course you'll note Alate is lying, again.

Unless you can show me that this academic “information” thing is more important than the end results, I think he has carried his point

[The Barbarian]

I'd like to see Barbie show some intellectual honesty.

I don't think that's going to cover for your behavior, Stipe. You won't post a testable definition of "information" for biological systems because you have no idea what it is, and because if you commit yourself, you suspect that we'll nail you with it.

Our position has been explained clearly enough.

You've made it up as you've gone along.

Perhaps he could accurately restate what our definition might be without pejorative or insertion.

But you'll never do that, for reasons that are quite clear.