The brown fur. Random changes can only degrade information. There is a cost to the information that outweighs any accidental advantage.
NIH: 100M Years to Change a Binding Site
- Thread starter Jefferson
- Start date
- Status
Specifically identify what is the cost in going to white fur.
In an imaginary animal you just made up? We almost certainly have little idea of how DNA is informed for fur colour in real animals. And there is certainly no way we could identify the disadvantage of a change (even if we knew what it was). And you want me to show you this in an imaginary animal?
Like my code, it does not matter that we cannot understand the message. The fact still remains that random changes scramble information. How is it that data from DNA is immune to this fact?
Tyrathca
New member
Something which is not demonstrated as universal, if it is demonstrated (mathematically) could you show this or direct us to where this is shown?
If I had two sets of DNA but you don't apriori which is the original and which is mutated from it could you determine which is which? Based on you claims that loss of information is universal, and not a mere assumption based on the situation, one would think you could do so. Just quantify the amount of information in each and the one with the most is the original. If you can't do something so simple I fail to see what justification you have for saying random changes lead to loss of objective information.
I'm not sure why you thought that was an assumption of the truth of evolution. The whole premise of evolution is that the similarity to the original DNA is irrelevant, merely if it survives better or not, ergo it does not fit the criteria for an appropriate scenario to apply your (unspecified) aspect of information theory (I'm assuming some relation to Shannon). It only seems to work in situations where you a priori assume that a change from original must be bad (something which there is no reason to do in evolution/biology).
This is unless you can demonstrate (or reference where it is demonstrated) that your stated aspect of information theory is universal to all instances and measurements of information. If this can be demonstrated then it would challenge evolution, but you have not done so.
Could you be more specific about which aspect of information theory you think is being claimed can/can't or should/shouldn't be applied? You're the one saying that information theory disproves evolution after all (without actually applying any of the mathematics btw). I'm just saying your particular application seems to be innapproporiate, unless you can be more detailed/specific.
Where is this demonstrated as anything more than an a priori assumption? And does this apply to all measurements of information or just some?
A forest fire kills all the forest-dwelling boders. Centuries later a snow-living boder has a brown-fur DNA alteration that restores the DNA to what the ancient forest-dwelling population started with. With an abundant untapped food supply in the regrown forest, that new brown-fur strain of boders moves back in.
Was there a loss of information in going from the white-fur DNA to the brown-fur DNA? Was there a cost in that change?
Meiosis only works in controlled parts. Thus it is programmed variation.
That's a great example. You can mix up some of your message, but not all of it. It's the same with the next generation, since only parts are varied, and not all, it shows the variation is programmed.
You should ask a programmer at the college about the advantages of programmed variation. Random in programs is a huge problem that continues to be worked on to this day. Biomimetics, BTW, is a promising path for computer scientists in solving this problem.
Apparently it was rhetorical, seeing as you answered it. How about this question?: can you have control without messages?
I have no comment against this. If I wanted a great example to prove my point, I'd have to quote this! So you prove my point and yet...
Maybe you didn't really mean to say what you just said.
Assuming the white fur is a loss of information, the brown fur would be a gain in information. You realize we are back at a net 0 gain in information, right?
Yes. But with our current knowledge of DNA and a sample of 2, it wouldn't be enough data. However, when we encounter this situation, we have to assume the veracity of information theory. For instance, if someone comes to you with a perpetual motion machine in a black box, you should immediately assume the 2nd law still works and the black box doesn't.
First, it's me that stands squarely on Shannon. Second, it works in every data set we apply it to, including DNA. If it didn't, there would have to be a reason why. Do you have a reason why Shannon doesn't apply to DNA?
No, information theory is established. Evolution, so far, has relied on "we don't know how DNA works therefore information theory doesn't apply". Thus, it is your side that must show evidence for your claim that information theory does not apply to DNA.
Shannon. But if you can get by that one, there's a few more to overcome. With all of evolutionists bluster, I think they know Shannon is unassailable, and will never even try.
No, you are. Here's your quote: "If this can be demonstrated then it would challenge evolution"
My pick for Subscriber POTD :first: Just brilliant.
I have done some studying of informational theory. One of its tenets is that you have to start with an intelligence to design the information systems. DNA is loaded with information. One could say that it operates like a programming tape filled with commands that tell the micro factories in the cell what to do in replicating the cells. This information tells the cell if it is the human cell, a bird cell etc or even what the cell's function is. Old computers used to operate with magnetic tape that had commands on it. How the cells take sections of the DNA and use them reminds me of this tape.
Take a gander at this--->http://www.phy.duke.edu/~hsg/176/table-images/bacterial-motor.html
A paper from Duke university that shows a picture of the bacterial motor. Doesn't that look an awful lot like a rotary engine in an airplane? This device is built using DNA information, just like we use blueprints to build our car engines. Did chance create this? I don't think so!
I've given a couple of very simple examples. Did you read them? Can you understand them? Not sure what else I can do. :idunno:
Two sets of DNA? You mean two full genomes from two animals of the same kind?
Sure. You could work out a compression system for the data and apply it to both in the same way and see which was most compressible. The one that could be sent using less bits would have less information because the reduction of uncertainty in the receiver would be less.
But I'm guessing that DNA cannot be much further compressed. :think:
Because it is likely already very compressed, I would doubt that DNA information can be so easily processed as I outlined above. Certainly I would want to see a much larger data set to cancel out any natural variation in the results and realise a good average. If we start to see the average shift, then we're talking!
Because "finding one that survives" has nothing to do with a study of information! You cannot claim that the numbers are not relevant just because you can suggest a situation where a random change can have an effect outside what the data is capable of controlling. If a random change makes a white rabbit and the white rabbits happen to live longer, that doesn't make the random change non-random and able to fit the compression system better. There is always a cost. You're just able to avoid talking about the cost by ignoring the entire genome and by only looking at the expression of one trait.
Which is contrary to the premise of information theory.
How can you discuss this if you do not understand what we are talking about?
Random changes always degrade the information in a transmission. Fact. Evolution has to fit with the facts. Not the other way around.
Or how about you show us a non-evolutionary model where random changes to a signal are good for the process. :idunno:
Name one!
Atheists in this thread have intimated repeatedly that information theory does not apply to evolution. How about you tell us how and quit asking me to do your heavy lifting for you. :up:
I didn't say information disproves evolution. :idunno: And I have provided maths. :idunno:
Information relies upon certainty that what is sent is able to be understood correctly by the receiver. Random changes can only ever be detrimental to that end. My assumption is that information theory holds true for biological information.
Yes.
Exactly!
And what is further ignored is that organisms may not respond solely in the expression of fur colour to random changes in their DNA. Most certainly there are other systems and functions that are also affected by the random changes. Those effects will bring much greater disadvantage than the accidental advantage that is the focus of this example.
A_O, there is natural tension when the two sides talk about X & Y Adam and Eve because they have differing assumptions as to what produces the genetic evidence we see. Here's my understanding, and if you point out any errors, and I can see and accept your correction, I will adopt it.
Bob Enyart on Mitochondrial Eve
Actual measured mutation rates as reported by Ann Gibbon in Science magazine indicate that, if these rates have been constant, "mitochondrial Eve… would be a mere 6000 years old." From an anthropology professor's popular article, "Analyses of the mitochondrial DNA of living humans from around the globe have shown that all are ultimately descended (if we trace exclusively through female links) from a common ancestress..." The same result would occur if in fact we have all descended from an original, created biblical Eve. However scientists quickly point out that their analysis doesn't require a biblical Eve. For example, you and all your full siblings have your maternal grandmother's mtDNA and yet you are all also descended from another woman from her generation, your paternal grandmother. Yet this finding does falsify two evolutionary expectations, the first from an old minority view held by evolutionists like the discoverer of "Peking Man," that humans evolved from parallel hominid groups. Secondarily, just as the discovery of soft-tissue from a T-rex falsified the evolutionary expectation that we would never find original biological material from dinosaur fossils, the recent age of mitochondrial Eve falsifies the mainstream Darwinist expectation that she would have been much older. That expectation is falsified whether we use the 6,000 year date which is based on exclusively human DNA and documented mutation rates, or even when evolutionists stretch that date by one or two orders of magnitude as they do by including chimp DNA in their data set. Either way, this finding falsifies the evolutionary expectation that such an Eve would have lived much earlier. Likewise, scientists have discovered a recent Y-chromosomal Adam (yet were shocked by the 30% difference when compared to humans, of the chimp's Y chromosome). As Dr. Walt Brown summarizes all this, "How likely is it that other men lived a few thousand years ago but left no continuous male descendants, and other women lived 6,000 years ago but left no continuous female descendants, and we end up today with a world population of almost 7 billion people?" Extraordinarily unlikely. So just as most astronomers came to admit, uneasily, that the universe had a beginning (but still they reject Genesis by holding to an increasingly untenable Big Bang theory), so too evolutionists are acknowledging much of what the biblical creation model predicts about the human genome, while not realizing that the historic events recorded in Genesis help wonderfully to account for their data. For the Bible says that God recently created Adam and his wife, and that "Eve... was the mother of all."
Of course our conclusions differ, but are there any factual errors here A_O? I'm interested in your criticism. Thanks.
-Bob Enyart
That’s what I would say. But in Stripe’s reply (after yours) he said “yes” to my asking if the change from white fur back to brown fur was a loss of information, but then he turns right around and agrees with you that it is a gain in information.
Previously Stripe has said “Random changes can only degrade information.” (post 241), “My assumption is that it is true that random changes are detrimental to information.” (post 238), “And by definition a random change comes at a cost to information. Every time.” (post 232), “Both changes (assuming that random mutations were responsible for them - … resulted in a degradation of the information.” (post 228).
So now in a single post Stripe responds to me that white fur to brown is a loss of information (consistent with his “always a loss” claim), but then his next word which answers you he contradicts what he just said to me.
Stripe???
Alate_One
Well-known member
For the Mitochondrial DNA only, this would be true although not required. You seem to like talking about DNA but you consistently ignore clear data that contradicts your position. Lets look at the nuclear DNA, as I asked you to do several posts ago.
Complete Journal article
Here's the biologos summary that I posted earlier.
It is physically impossible for the number of alleles of the MHC (among other genes) to have arisen from 2 people 6000 years ago. If you don't have the background to understand that and would like to know why that is, I will explain it to you on request.
| Genetic variation at most loci examined in human populations indicates that the (effective) population size has been - l0^4 for the past 1 Myr and that individuals have been genetically united rather tightly. Also suggested is that the population size has never dropped to a few individuals, even in a single generation. These impose important requirements for the hypotheses for the origin of modem humans: a relatively large population size and frequent migration if populations were geographically subdivided. Any hypothesis that assumes a small number of founding individuals throughout the late Pleistocene can be rejected. Extraordinary polymorphism at some loci of the major histocompatibility complex (Mhc) rules out the past action of severe bottlenecks, or the so-called founder principle, which invokes only a small number of founding individuals when a new species emerges. This conclusion may be extended to the 35Myr-old history of the human lineage, because some polymorphism at Mhc loci seems to have lasted that long. Furthermore, although the population structure prior to the late Pleistocene is less clear, owing to the insensitivity of Mhc alleles, even to low levels of migration, the nature of Mhc polymorphism suggests that the effective size of populations leading to humans was as large as 10 5. Hence, the effective population size of humans might have become somewhat smaller in most of the late Pleistocene. The reduction could be due either to the then adverse environment in the Old World and/or to the increased migration rate. It is also argued that population explosion fostered by the agriculture revolution has had significant effects on incorporating new alleles into human populations. |
Complete Journal article
Here's the biologos summary that I posted earlier.
It is physically impossible for the number of alleles of the MHC (among other genes) to have arisen from 2 people 6000 years ago. If you don't have the background to understand that and would like to know why that is, I will explain it to you on request.
And again you're going into your fallacious logic that "scientists are surprised" means that scientists cannot be trusted to figure out anything at all. You spend half of this little section essentially saying this over and over again.
Now that's just rude, BJ. I did not agree that it was a gain in information after I said there could never be a gain. That would be silly. I agreed with Y. in his analysis that what you were describing was a net gain of zero. And I added to his analysis an "At best"!
Your example is a net gain of zero at best. And according to information theory, every stage will be a net loss. And my post was pretty clear that this is what I was saying.
Now, are you going to continue to only pick at these tiny issues or are you going to answer the substance of what has been said? Why can information theory not be applied to data from DNA?
This is classic, coming from the undisputed master of the mockery symbol at TOL.
In my scenario there was a period of time in which there were no living brown furred boders. The DNA in the white furred boders has no “memory” that it originally came from the forest animals. If a biologist first came upon boders when only white-furred ones existed, he might sequence their DNA. A couple decades later, he finds there are now a few brown-furred boders. If he applied information theory as you are trying to, what would be his conclusion about this DNA change? With no idea there had previously been brown-furred boders, would he see this change to brown fur as an increase in information?
How do you reconcile “every stage will be a net loss” with your agreement with Yorzhik that the stage going back from white fur to brown is a gain in information?
I realize you have this “forest” you are anxious for us to study. But I am standing in the middle of a bunch of tall things, (I think they are called trees) asking about them. I was told once, perhaps incorrectly, that there is some connection between the trees I am asking about and forests.
Classic, and correct. You're continuing to completely disregard the substance of what I say in favour of repeating what you say.
No. He would only ever see a decrease in information. Unless he is not actually applying information theory and only looking at fur colours and survivability through an evolutionary world view.
Neither Y. nor I ever said there was a gain in information.
I have not contested, that from the viewpoint of the brown fur, a DNA change to select white fur might be considered a loss of information. But if a brown-fur boder finds himself forced to live in snow, a miraculous transformation of all his DNA to make white fur would be beneficial. So if you restrict information theory to analyzing the integrity of a single specified message (how to make brown fur), you are innately starting with a less-than optimum message from the viewpoint of a show-bound boder.
Yorzhik, it seemed pretty clear that you and I were in agreement that if the step from brown fur to white was a loss of the original information, then the step from white to brown must necessarily have been a gain in information which compensated for the earlier loss. Correct?
Stripe, is it your contention that information theory says there is inevitably a degradation (a cost?) when DNA mutates – meaning no mutated offspring can ever be as capable of survival as the parents?
p.s. Yorzhik, tell Stripe whether or not Walt Brown is correct on frozen Mammoths not having old stuff buried under them.
1. You cannot restrict DNA to only it's influence on one trait.
2. Accidental benefits may happen, but you continue to ignore information theory in order to maintain the front for evolution.
Yes.
Survivability is an irrelevant measure.
I am pretty certain that some places in DNA do affect just one trait. But that is a red herring. Is information theory dependent on each “trait” being non-trivially connected to other traits?
First I am trying to see if you understand information theory enough to apply it one simple case. It’s not looking good so far.
I assumed the “cost” you claim is always incurred meant the genome was somehow degraded. I was using survivability as an evolutionarily crucial aspect of that degradation.
But anyway, if you say that going from brown fur to white involves a loss of information, then assume the white boders were the original ones. When a branch of them developed the brown coats, then that must have been a gain in information (if brown to white is automatically a loss of information). But especially considering that in that scenario there had never been any prior brown boders, then this DNA mutation must have been an increase in information (under your rules).
No. But your example requires suppression of inter-relatedness in order to hide the possibility of changes being always bad. In every other example random changes diminish the information of a signal. Why is DNA any different?
...said the man who claimed to be unfamiliar with the subject. :chuckle:
Information theory dictates this will be so.
You cannot assume evolutionary theory as true and then use evolutionary theory to test your claim that evolutionary theory is true.
That's just dumb.
Brown to white and white to brown is irrelevant.
Your rules. Evolutionary rules. Under information theory, random changes are always bad for information. Now, since you seem to be "conversant with the exact definitions and rules applicable to information theory" enough to test me on the subject, perhaps you can answer the question. Can you tell us why information theory does not apply to the data from DNA.
- Status