What exactly is ID?

quote:

---

Yes I do and I have already given those examples several times and demonstrated why.

---

This is completely false. Firstly there are no examples for you to give. Secondly having reviewed your contributions to this thread I can find no post where you make any attempt to demonstrate that anything is CSI in Dembski's sense. I can find a post where you deny knowing of Dembski's definition (despite using Dembsk's term CSI), I can find a post where you use a definition of "information" completely different from Dembski's CSI.So no, I cannot even find a post which contains a failed attempt at such a demonstration - or even indicates that you know what Dembski's definition of CSI actually is.

quote:

---

If you have a rebuttal I am all ears.

---

Here's is my rebuttal. So far as I can tell you gave no examples of Dembski's CSI in living beings, nor did you offer any argument to demonstrate the existence of Dembski;s CSI in living beings.Are you really listening ?

There was a problem, which has been fixed now.

quote:

---

My intent here is not to make any defenses for Dembski's use of the term csi.

---

While Dembski' use arguably DOES need defending (in that it is quite misleading) that was not the point under discussion.Let me remind you. When I pointed out that you had no known examples of Dembski's CSI to use as evidence, you claimed that you had "...already given those examples several times and demonstrated why." You claimed to be ready to listen to rebuttals.Well I rebutted your claim by pointing out that you had done no such thing. And this is your reply ? "My intent here is not to make any defenses for Dembski's use of the term csi."

quote:

---

He doesn't own or have a patten on the term.

---

In fact so far as I can tell he essentially invented the term. While there are ideas using similar phrasing predating Dembski, Dembski is the first to use CSI as a specfic term denoting a specific idea.

quote:

---

I have pointed out that there is observable csi in living organisms and that it can be quantified and measured.

---

Using a meaning of CSI that does NOT rule out natural patterns.Let us sum up the conversation (paraphrased for brevity):You: ...I am going to talk about complex and specific information which does rule out natural patterns.Me: Then you have to use Dembski's CSI, and you have no examplesYou: I AM using Dembski's CSI and I DID provide examples.Me: No you didn'tYou: I'm NOT using Dembski's CSIPerhaps this makes what you are doing more obvious.

I am going to try to phrase this as politely as I can without being reduced to dishonesty.Brad, I did at least point out that Dembski was using different definitions of both "Complex" and "Information". If you wanted more details then this thread, at least would be an appropriate place to ask for them. You did not.In fact you went on to claim that you had provided examples of Dembski's CSI, complete with arguments demonstrating that fact. However, as it turns out not only had you not done so, you do not even know if your examples fitted Dembski's definition nor how to provide the demonstrations that you claimed to have already given. In short you had absolutely no valid reason to think that your claim was true.And if you think that it is rude to be told that you should settle on a single definition for your arguments - well the fact is that it was necessary. We cannot have a rational discussion concerning CSI until we know what you mean by it. I hope that you are not going to say that your arguments rely on switching between different meanings whenever it happens to be convenient, but that was what you were in effect doing. And if it was not deliberate you needed to be told to stop.

quote:

---

Merely meaning that, I do have examples of csi (nucleotide arrangement), and I had already given them as an example.

---

Yet what you wrote - taken in context - did not mean that at all. It meant that you had provided examples of Dembski's CSI. Which is completely untrue.Indeed, if your current response is accurate, in your actual reply you should have denied using Dembski's CSI and offered some argument that eliminated natural patterns from your definition of CSI. Neither of which you did.

quote:

---

Again I don't either align myself with nor do I distance myself from Dembski or any other ID scientist.

---

Dembski is not a scientist.

quote:

---

If my terms bare a striking resemblance to or are taken from something I may or may not have read or heard one of them say, I don't know or really care. I know when I use an English word, I always mean it in the sense that it is most often used in the English speaking population.

---

So far as I am aware the "majority usage" would be Dembski's since Dembski is the one who formulated and popularised the specific term and the acronym. And as I read on it becomes clear that your usage of CSI is an idiosyncratic usage that I have never seen before - it is neither of the two common definitions.

quote:

---

So just for the record let me clarify what I mean when I say Complex Specified Information. I am referring to anything of a highly intricate nature that is arranged in a specific order to serve the sole purpose of relaying instructions to another system, through the activation of one possibility to the exclusion of several others.

---

Then - as I have pointed out- you have no sound basis for ruling out natural patterns. Indeed to make that claim you must beg the question by assuming that evolution is false.

quote:

---

Author mentioned that you were looking for me to give a method by which one can recognize csi as opposed to purely natural phenomenon. Like some kind of meter that we can plug into any situation and see the needle swing left for natural or right for intelligent. First we have to understand that intelligence is the ability to make a choice and that information from that intelligence is the choice of activation of at least one possibility to the exclusion of several others. And that activation must be recognized to mean something by both the transmitter and the receiver. And finally for a third party (humans) to detect that information, we also have to be able to understand the meaning of the conditions that were activated. I mentioned logs on a beach, the other day, arranged to read, "Marooned... please send help." If I were from China and did not speak a word of English, the logs would appear to be arranged in purely a random order to me. So detecting csi is not as simple as having a meter we can hook up, but it is detectable nonetheless. And as I said before, we have detected a high degree of csi in the DNA code of all living organisms.

---

This seems to be a concept of information, since neither complexity nor specification play any role in the definition. Indeed, it seems of be a concept of semantic information, assuming intelligence in the receiver and transmitter. Or am I mistaken in thinking that the choice is supposed to be made by the transmitter, and that transmitter and receiver are supposed to understand the meaning of message ? After raising the issue of choice you fail to explain exactly where it fits into the definition.I will also note that this is completely different from the definition of information that you copied from wikipedia.In fact, it seems that you are using Dembski's terminology to make Gitt's arguments - and Gitt simply refers to information, not CSI.So at this point, it is clear that you are NOT using the majority definition of CSI. Your actual definition is unclear, and it is far from clear that CSI as you mean it is present in DNA.

quote:

---

But what we clearly do know, is that the one that it had is now lost.

---

Which, of course, I did not argue against.

quote:

---

You simply said that Dembski didn't follow his own steps. Which it totally false.

---

Simply denying the facts is hardly a road to productive discussion. If I am incorrect and Dembski did calculate the probability of P(D*) you need to demonstrate that.

quote:

---

And than I told you fine, I'll do the new calculation. You just need to tell me what exactly do you want.

---

I already did that. You need to calculate the number of proteins that are no more than 20% different from each of the 50 used in the E Coli flagellum.\

quote:

---

And I did, becasue D* is the specified pattern in question. It's the "bidirectional rotary motor-driven propeller" consisting of 50 proteins.

---

No. D* is the specification considered as an event. So it is ANY "bidirectional rotary motor-driven propeller" however many - or few - proteins are involved.

quote:

---

But that's not a detachable patternt. Becasue 499 coins could be heads, so could 498, 400, 333 etc... You need to get something like a Fibonacci sequence to infer design.

---

Of course, the Caputo case involved nothing more than the Democrats taking first position on the ballot far more often than expected. There was no other pattern to the results.In fact your objection has nothing to do with detachability at all. All you are saying is that D* is much wider than E. But that is the reason WHY we need to use P(D*) - because P(D*) CAN be much higher than P(E) and thus the SPECIFIED information may be only a tiny fraction of the whole.

quote:

---

Well it had to be put in at some point in time, and some point is space. Where that was we do not know. But I'm not claiming that it is being inputed right now. I'm claiming that all desgin we see now is a copy of teh original one.

---

Which rules out your claim:

quote:

---

Why should we not be able to detect design in afully automated process? The information had to be inputed at some point in time.

---

Because Dembski's method relies on detecting the input of information from a designer (as opposed to natural sources). Let me be clear, I am claiming that the operation of a fully automated process cannot be used to reliably detect design because it has a high probability of producing its output. It is certainly possible that the process itself might be designed and if it is, that you might - in principle - be able to use Dembski's method to detect it with reasonable reliability. (While Dembski's method has a number of problems, if the practical difficulties could be overcome, it might be usable as a decent argument for design - when applied correctly).

quote:

---

ARRRGH!!!!
Yes, all of them. If mutations suck so much, than their effective frequency is zero. If it's a bit better than it's 1%, if it's bit more better than it's 2% etc. With the best possible mutation being in 100% of the population.

​

So now tell me, if we know that this applies to frequencies of alleles for hair color, and eye color, do you not agree that this alos applies to frequencies of sickle cell?

---

So what you mean is that the alleles for hair colour and eye colour are neutral and the frequencies aren't changing much through selection or drift, and ask why that isn't the case for sickle-cell. The answer is that sickle-cell trait (heterozygous for sickle-cel) is quite strongly advantageous in malarial areas and mildly deleterious elsewhere, while the homozygous state is strongly deleterious everywhere.

quote:

---

So no matter how much examples I provide you are not convinced. Tell me, why should I bother with you anymore?

---

If you only want to talk to people who are convinced by your claims regardless of how poor your evidence is, then I suggest that you go to a creationist-run forum. I've already told you what sort of evidence you would need to produce. And any reasonable person would agree.

quote:

---

Umm... not. There is noise during selection which maeks your view of evolution primitive and childish. The efficiency of selection is not infinite, it actually sucks very much. There fore, beneficial mutations do not spread, and deleterious do nto get removed all the time.

---

Obviously you have not understood the position I am putting forward at all. I have not said that natural selection is perfect or that there is no noise. What I have said is that beneficial mutations have a greater chance of spreading - and the more beneficial they are the better their chances. And that deleterious mutations have a lower chance of spreading - and the more deleterious they are, the less their chances (in the extreme case of causing sterility or death before reproductive maturity, NO chance). This fact has to be taken into consideration, rather than simply shrugging it off.

quote:

---

Actually, it shows that for the duration of experiment the mutations casued the chains to get shorter and shorter. And those are the ones that spread more and more. The longer ones died out faster. So the population of RNA chains was losing their complexity, and becoming simpler not more complex. So how in the world do you think this process could have ever evolved those chains into people? Genetic entropy at it's best.

---

No, it is the exact opposite of genetic entropy. Genetic entropy supposedly degrades the fitness of the population until it is forced into extinction. In the Spiegelman monster case, mutations improved the population radically, to the point where the "monster" drove all other populations into extinction.

quote:

---

Choice is required in the origin of the information. However if an intelligent source forms a system that mechanically transmits the information, and a receiver that mechanically receives the information and utilizes it, then choice is not necessary with in the two mechanical systems. That was what I meant when I used the Wikipedia article which said, "Information is any type of pattern that influences the formation or transformation of other patterns.

---

So, just to be clear, are you including a requirement for an intelligent choice in your definition of information ? Because that is nowhere in the Wikipedia definition you quoted.

quote:

---

We can still recognize specific patterns being utilized to create specific conditions, to the exclusion of several others, that are received and utilized for a specific function. All other systems that we have ever observed in the whole of human history, that produce specified information, required an ability to make a choice. Therefore we can logically conclude that a choice was required (and detected) to form the specified information in DNA.

---

However, such an inference is very weak because it pretty much ignores everything else we know about DNA and instead uses a comparison with quite different systems - all of them human created.In fact we know that the DNA "instructions" can and do change with no sign of a designer intervening. Even antibiotic resistance can be seen as a specification, as can the colour change of the Peppered Moth. Yet we see no sign of any intelligent choice in the development of these.

quote:

---

Choice is the main ingredient necessary for intelligence. Intelligence is what is necessary to produce complex specified information.

---

This does not explain what role choice plays in the definition. Are you asserting that your definition of complex specified origination requires that it originates with an intelligent choice ? If not, just where does it fit in ?Until we have settled your definition (which currently looks like Gitt information and not like any common formulation of CSI) we cannot say that DNA contains CSI or that intelligence is required to produce it.

quote:

---

No, I am saying that (a) intelligence is required for the origination of information, and (b) it is not necessary for the continued transmission of information. And that does fit the Wiki def.

---

I really need to get this absolutely clear. When you say that "intelligence is required for the origination of information" is that part of your definition, or is it an assertion ? Because it certainly isn't part of the Wikipedia definition.

quote:

---

The changes we "know" that occur in the DNA are just as you said, changes with no sign of a designer. But they are not changes that can explain its origin as a whole to begin with.

---

However, they do go some way to explaining the information we currently find in the DNA of existing organisms. And we certainly do get new variations carrying useful information (by the Wikipedia definition) that was not previously available. The question is, is this information by your definition and if not, why not ?

quote:

---

Than what's your point? That the enzyme got a new function, but that we do not know about it?

---

My point is that we do not know whether the mutated enzyme had no function or if it did. That is what I said. Perhaps you should try harder to remember just what you are arguing against ?

quote:

---

D* is the "bidirectional rotary motor-driven propeller" which consists of 50 proteins. Which have a probability of forming into that specific pattern of about 1:10^2954.

---

In other words, your real "specification" is the E Coli flagellum plus near variants of it. And as I have pointed out, this is a fabrication. (The probability calculation is wrong, too, but that isn't important now).

quote:

---

Fine. Explain how in detail.

---

Since actually doing the calculation would be a complete waste of time, why exactly do I need to explain how to do it ? After all, I don't care whether you do it or not.

quote:

---

Yes, but what is the complexity of this particular specification? It's 50 proteins. Please tell me how do you intend to calcualte the probability of a specification without knowing the complexity.

---

If D is "bidirectional rotary motor-driven propeller" then it makes no mention of 50 proteins. That is just an unspecified detail read off the event (i.e. fabrication). The "complexity"(or rather the information content) is calculated from the probability (it's -logmax(P(D|H)) over all possible explanations H). That's Dembski's method.

quote:

---

And that is why nobody would care if Democrats won 21 and Republicans 19 time the ballot.

---

Exactly - because Dembski was right and the correct probability is P(D*). Which does not permit adding in details that are not in the specification D.

quote:

---

How does it rule it out.

---

Obviously if the information was only input in the creation of the first E Coli (or it's ancestor) it cannot also be input in the ordinary operation of the growth mechanisms.

quote:

---

But there is no input from natural sources. Because natural sources have no teleology. They are constrained by deterministic natural laws. Therefore, any information they transmit, was originally inputed by an intelligence.

---

Natural sources do not produce Complex Specified Information in Dembski's sense (or rather they almost certainly do not, even if Dembski were entirely right). However we cannot work out if something is Complex Specified Information by Dembski's method without looking into its origin. By Dembski's definition, natural sources can and do generate information, even complex information and specified information (because his measure of information is simply an measure of improbability).So, without applying Dembskis method to this information, we cannot work out whether it is CSI or not. And if it is not CSI - if the specified part of the information (D*) is not complex - then, according to Dembski, it could have a natural source.

quote:

---

Except that that is teh case. Just like blue eyes and blond hair is kept at a certain frequency, so is sickle cell. And every other gene.

---

Even if it is the case that the frequencies are constant (and I'm far from certain of that), it isn't the same sort of selection as applies to sickle-cell. Which means that all you are really doing is pointing out the weakness of drift in large populations. Remember, it was you who introduced sickle-cell as an example of a beneficial mutation (and even claimed that it was spreading), so comparing it to neutral mutations really misses even your own point.

quote:

---

And yes, I will again say that I know that. But, there is so much noise that regardless of beenficial and deleterious mutations, they do not get spread around as much as they would if there was no noise.

---

This is the same noise that - according to you - is having absolutely no effect on the frequency of hair or eye colour. Which means that it isn't going to have much effect against selection either.

quote:

---

So in reality, when we take the effects of noise into considerations, we almost always have genetic drift, as opposed to positive selection working. Meaning, beneficial and deleterious mutations are equally likely to spread through the population.

---

For that to be true, the effect of drift would have to be so strong as to completely overwhelm selection. To be entirely true it would mean that there were no advantageous or deleterious mutations at all. All mutations would be strictly neutral. A completely sterile individual would - on average - have the same number of offspring as the best and most fertile member of the population. Obviously that cannot be the case.So since it can't be entirely true, the question is, how close to the truth is it. What is the evidence ? Is it so weak as to have almost no effect - as you claim, or so strong as to completely overwhelm selection - as you claim.

quote:

---

NO! How many tiems do I have to say that geentic entropy does not always have to equal loss of fitness?

---

The number of times you say it, doesn't matter. Genetic entropy IS a cumulative loss of fitness, leading to mutational meltdown and extinction.

quote:

---

The point is that the chain got shorther. Because of mutations. And the more time went by, it got shorter and shorter. The longer ones died out. And only the shortest were left. Clearly demonstrating that evolution does not work. You can't get a human from RNA chains by this process becasue they are losing not gaining complexity.

---

Aside from the fact that that is going well beyond what the experiment can reliably tell us, it isn't even relevant to the real point. You claimed that it was evidence that genetic entropy increases. And it isn't. The monster beat genetic entropy.

quote:

---

We know it lost one. It's an observable, empirical, fact! And we are going to stick with that. Unless you think we should invent imaginary functions just for the sake of the argument?

---

That is not in contention. The question is whether the mutated version had NO function. And we do not know that.

quote:

---

WHY!? Why the hell is it a fabrication? If it can be described without looking at the event E, than it's not a fabrication. And yes, it can be described in such a way, therefore, it's a specification.

---

Since the probabilty you are calculating relies not only on the number of proteins in the E Coli flagellum, but also their structure it seems very unlikely that it can be separated to the degree required by Dembski's TRACT condition. But please, if you think otherwise then give the specification.

quote:

---

If you do not want to show me what calculation I should do, than don't bother me with it anymore.

---

So long as you give me no reason to bring up your error I am happy to stop talking about it.
On the other hand if, for instance, you try to use the result of your erroneous calculation it is perfectly reasonable for me to point out that it is wrong.
(I will add that anyone with even a basic understanding of probability theory should be capable of working out how to calculate the number without too much difficulty).

quote:

---

Which comes out at 1:10^2954, as I have mentioned for a trillion times already. With simply applying this here equation.
χ = —log2[120^10 ϕ s(T)P(T|H)]

---

As we both know that isn't true. Your calculation is based on using details which are completely absent from the specification "bidirectional rotary motor-driven propeller".

quote:

---

Basicly what you're saying is that we can't know if something is CSI if we do not know how it arises. Why?

---

Basically I am saying that we cannot know if something is CSI unless we use the method for working out whether it is CSI or not. Which requires looking at all the possible explanations of how it arose.

quote:

---

Than what kind of selection is operating on sickle cell? Is it some magical one? Obviously it's not. It's the same as the one that is operating on any other gene frequency. Why would this one be special?

---

I have already explained this. Firstly the sickle-cell trait is beneficial while hair and eye colour are neutral. So, the frequency of hair and eye colour is dominated by drift, while selective effects dominate in the case of sickle-cell. That is one difference. Secondly - as I have pointed out more than once - the sickle cell allele is unusual in that the heterozygous state is quite strongly beneficial in malarial areas, while the homozygous state is strongly deleterious even there. Thus selection acts to hold the frequency in balance.

quote:

---

Well it did spread didn't it? It is irelevant if it is considered a neutral, a deleterious or a beneficial mutation. And it is also irelevant which mechanism has spread it. A certain population in Africa has this mutation fixed in.

---

It is not fixed in ANY human population (unless you mean some small and soon to be extinct group). As I have already explained. it can't be, because the homozygous state is strongly deleterious and for it to be fixed the entire population must be homozygous for sickle-cell.And while sickle-cell did spread in the past, it is no longer spreading - which was your claim.

quote:

---

So all this traits get evaluated before an individual gets selected for. It's obvious that just because a particual individual has a beneficial mutation, that doesn't mean that such an individual will on average get selected for by natural selection. Becasue that is only one trait, there are 6 more others to evaluate before it gets selected.

---

This only shows that you badly fail to understand my argument. My argument is not based on individual cases at all. Instead it is based on the long term aggregate effect over all the population.

quote:

---

How could he have beat the genetic entropy if he WAS CONSTANTLY DECREASING IN SIZE!!!!!!?????!?!!?!?!?!

---

The monster beat genetic entropy BY decreasing in size. By decreasing in size it increased it's fitness so that it could outcompete all the other variants.

quote:

---

Would you consider any mutation which increased binding affinity to anything to therefore represent an increase in information/CSI? Even if the subsequent binding served no functional purpose for the mutated protein?

---

I thought that I would let Smooth Operator reply before giving the real answer.Strictly speaking the specified information measure refers to the specification, rather than the actual event - in this case the ribosome. In fact it has to, for the obvious reason that only that information is specified. I can think of only one sensible way to get a measure of the specified information for the event, and that is to use the highest value for all the valid specifications that include the ribosome.An increase in specificity would allow us to draw a tighter specification (i.e. if the specification includes a minimum level of specificity) which would be expected to have a higher amount of specified information (it cannot be lower, since the new specification is contained within the older one). Thus it is possible for the specified information measure of the ribosome to increase - but it may not - and if the change has any other effects the specified information level might even decrease.(Technically we could do much the same with a specificity decrease, but unless very low specificity values are rare it's unlikely to be significant)I'll also comment on this:

quote:

---

You seem to be saying that as soon as the full streptomycin biosynthesis pathway had evolved the specified information content of the ribosomal genes jumped despite there being no change in its sequence...

---

In fact this does seem to be a flaw in Dembski's methodology. Once streptomycin came into existence, specificity to streptomycin became a valid specification even though - in my assessment - it would not have been before. And that would make that specificity into specified information. Which, as you correctly point out, raises the possibility of a false positive.

quote:

---

No, that's not the question. That was not the point of the experiment. The point was to show how much mutational load can a enzyme take before it loses it's function. The one function that was known to exist was measured. And it went away after some time. So now we know how much changes can there be on average before a certain function is lost.

---

I am afraid that you are incorrect again. You claimed that it had lost all function. I claimed that you did not know that. And as we have seen I was right, although it took an amazingly long series of posts for you to realise it.

quote:

---

The only relevant question is. Does the flagellum describe this pattern: "bidirectional rotary motor-driven propeller"? The answer is - yes. So since we know of other objects that exhibit this pattern, we conclude that it's a specification, and not a fabrication.

---

Amazingly enough your clumsy phrasing has made you say something that is about right, although probably unintentionally. For your calculation to be correct, the E Coli flagellum (plus any allowed variants) must include all "bidirectional rotary motor-driven propellers". And in that sense it would describe it. However, we are certainly in no position to say that that is true, and in fact I know that it is not true.You are also correct to say that "bidirectional rotary motor-driven propellers" is a valid specification, because there are other things that fit the definition. Unfortunately this means that all those things must be included in D*. And since your calculation uses specifics of the E Coli flagellum that do not necessarily apply to all "bidirectional rotary motor-driven propellers" you are calculating the probability for a different "specification" - and in fact that "specification" is a fabrication.

quote:

---

I'm willing to do the "correct" calculation. I'm simply waiting for you to tell me what to do.

---

I've told you what to do. Calculate the number of sequences that are no more than 20% different. If you can't even work out how to do that then you don't know even basic probability theory. I might be willing to help you a little more, but to do that sensibly you will need to explain how you did the calculation in the first place. Something you were unwilling to do before.

quote:

---

Let's see what Dembski has to say about that...

​

quote:

---

For a less artificial example of specificational resources in action, imagine a dictionary of 100,000 (= 105) basic concepts. There are then 105 1-level concepts, 1010 2-level concepts, 1015 3- level concepts, and so on. If bidirectional, rotary, motor-driven, and propeller are basic concepts, then the molecular machine known as the bacterial flagellum can be characterized as a 4-level concept of the form bidirectional rotary motor-driven propeller. Now, there are approximately N = 1020 concepts of level 4 or less, which therefore constitute the specificational resources relevant to characterizing the bacterial flagellum.

---

http://www.designinference.com/.../2005.06.Specification.pdf

​

Let me explain what he is saying here breafly.

​

The ϕs(T) part of the equation describes the specificational resources. That is, the amount of all possible specifications relevant the the specification T that is exibited by the even E. Imagine now a descriptive language D* for which we will say will be English language consisting of aprox. 100.000 words. That's 10^5 basic concepts.

​

Using the definition of CKS information theory The flagellum describes the pattern "bidirectional rotary motor-driven propeller". This patternt consists of 4 basic concepts, that is, 4 words. The total complexity of this specification is thus 10^5 10^5 10^5 10^5 which equals 10^20.

​

This is the complexity of the specification as calculated by Dembski himslef. After that we multiply it by 10^120, and with the probability of the event, and than we get the final number...

---

I am afraid that you have completely misunderstood what Dembski is saying here. Dembski is attempting to deal with the issue that there are many more possible targets that evolution might have hit. He does so by estimating the number of concepts at the same level and including it as a correcting factor (this is why it reduces the specified information content).Clearly adding in extra details - even if they are legitimate - would mean using a higher level concept, and thus this factor would be increased.

quote:

---

I agree. The problem I have with your interpretation is that you confuse growth mechanisms with high probability events.

---

Again, you are making a mistake. My point was that the growth mechanisms have a high probability of producing a flagellum. Of course we can say that bacterial reproduction includes producing the physical growth mechanisms with high probability. But again, all this means is that the design proponent has to do is to go back to the origin, which is where you say that the actual design implementation occurred.

quote:

---

But this is not a difference. Remember. Mutations are neutral/beneficial/deleterious depending on the environment. Therefore, in some environment, blue eyes could be beneficial, or deleterious. So no, there is no difference, the same natural selection is operating on all traits.

​

In all cases it depends on the environment and the mutations if they will be removed or spread and at which extent. There is nothing special in this case.

---

Of course the point rests not on what MIGHT be the case in some hypothetical environment but on what actually is the case in the environments which actually exist. And in that environment, the allele for blue eyes is classified as neutral and the sickle cell allele is classed as beneficial overall (in malarial areas), but is very strongly deleterious in individuals who are homozygous for the allele. And that is unusual.

quote:

---

Neitehr are a lot of other traits. So what? Are blue eyes fiex in human population? No obviously they are not. Only a small minority has them. Does that mean they are deleterious? No obviously not.

---

In other words it doesn't matter if what you say is true or not ? If you really don't care about then say so. You claimed that the sickle-cell allele was spreading. You claimed that it had been fixed in the population. Neither is true.

quote:

---

That's my argument also. Iw as just making an example on an individual. Are you telling me that other traits that get selected besides genes are not working on the level of population? Are you telling me they do not interfeer with genetic selection on the level of population over a long periods of time? Of course they do! And that is why natural selection is inefficient.

---

That isn't quite true. Your argument is intended to establish what you think will happen to large populations over long periods of time but it doesn't take either point into account. If the other factors are hereditary then they may be included with genes. If they are not then - on the scales we are thinking of - they may be largely ignored unless there is good reason to think that they strongly correlate with the occurrence of a particular allele.You see, the larger the population (in the statistical sense as well as the biological) the more the chance effects of noise tend to average out. This is why genetic drift is weaker in larger populations. So it is not enough to talk about individual cases or noise. What you need to show is that there is a systematic bias - and that that bias undermines selection to the point that mutational meltdown is inevitable.

quote:

---

No, it's the genetic entropy that is causing it's shortness.

---

In other words you claim that an increase in fitness is caused by an inevitable decline in fitness. That doesn't make a lot of sense. I say that a sustained increase in fitness shows that the decline is not inevitable, which makes a lot more sense.

quote:

---

And the most important question is, how is that chain supposed to evolve into something more complex if it is constantly decreasing? It's not! And that's why evolution doesn't work. If it did, the chain would be getting longer. But it's not getting longer.

---

It may be an important question, but it is not relevant to either the genetic entropy discussion or the discussion of Dembski's method. I don't intend to add to the problems of this thread by adding a third subject to the discussion. If you want to talk about it then I suggest that you start a new thread.Edited by PaulK, : Cleaned up a few typos

quote:

---

That is because it had ONE known function. We will not try to imagine new functions if we are not sure the enzyme has it, now will we? That's unproductive.

---

Well we're not doing that.

quote:

---

We know it had ONE function, and it lost that one function, therefore, it lost all functions.

---

That's illogical. Why can't you just admit that we don't know ?

quote:

---

But that's besides the point. Even if it had more function, which I wouldn't be surprised it did, that still doesn't refute my point that it lost a known function. Which, I repeat was the point of the experiment. The experiment was to see how many mutations can an enzyme take befroe it loses it's ONE known function.

---

And nobody has challenged that. So all it means is that you have put a lot of effort into arguing against a fact that you say doesn't matter. Well, why bother ? Why not just accept it and move on ?

quote:

---

Why would I have to include ALL existing "bidirectional rotary motor-driven propellers"? That makes no sense. The reason it doesn't make any sense is becasue some patterns that might fit that description could very well be under 400 bits of information. Therefore making our calculation useless in the first place. Why? Becasue we are than not talking about somplex specified information, but simply about specified information, which Dembski said, chance can generate.

---

So what you are saying is that it makes no sense to follow the method because you think that it might give a result you don't want.You have to include ALL possible "bidirectional rotary motor-driven propellers" because all of them fit the specification "bidirectional rotary motor-driven propellers". Because for a specification D, D* is EVERY possible thing that is delimited by D. That is how it is defined. That is the point of using a specification, to find the specified information. The specification D tells us ONLY that the event is in D*. So the probability of THAT happening is the probability we use for the specified information.

quote:

---

Therefore, you are wrong. Weare only supposed to include the complexity of the event E, which in this case is the 50 proteins of the flagellum.

---

As I have already shown it is the probability of D*, not E, that matters - according to Dembski himself. The Design Inference p165.

quote:

---

Just imagine a "bidirectional rotary motor-driven propeller" that has a complexity of 1.000.000.000 bits. You do understand that it has a chance vastly less probable to form by chance than a one consisting of 1.000 bits?

---

If most of that is unspecified information, it can be ignored. That's the Design Inference - low probability, unspecified results are attributed to chance.

quote:

---

Like I said, I'm waiting for you to tell me how to do the calculation. I have no reason to stall.

---

As I said, if you want me to help, you need to give me the details of your original calculation - I need them to help you. Obviously your failure to do so is a mere oversight - since you have no reason to stall. (But don't worry - you still get to do the work).

quote:

---

Like I said "bidirectional rotary motor-driven propeller" is a 4-concept pattern. Every pattern has 10^5 possible patterns. Therefore, the full complexity is 10^20. Now how do we use this? Well simple. We have to show that this pattern is still too improbable to hit by chance. Therefore, we multiply it by 10^120, than with teh complexity of the certain pattern, and that if this number is less than 1/2. Which is true, becasue Dembski says it here.

---

As I said, Dembski is using a correction factor, based on the number of concepts in the specification. More concepts means using a higher number. (And you will note that your 10^120 figure doesn't show up in either quote. It isn't in the immediate context either.)

quote:

---

Impossible and useless. We can not go back in time. That's obvious. And it's also useless, because natural causes can only transmit CSI, and not add more of it. So whatever CSI we see in nature, we know that that same amount, or more, was inputted in the first place. Therefore, it's fine to calculate it right there.

---

I agree that Dembski's method is nearly useless, and practically impossible to apply. That's why nobody bothers to use it. The problem is that what you are calculating isn't the specified information. It's just a useless and irrelevant number.The fact is that an algorithm or mechanism to generate a result, may be "simpler" (by Dembski's misleading usage) than the result would be if it were generated by pure chance alone.This is why we don't attribute pulsar signals to design. If we ignored the existence of the pulsar, and only considered pure chance as an explanation for the regular signal pulses we would have to conclude that the regularity of the signal was very unlikely and - inevitably - the sequence of pulses would soon qualify as CSI. If we followed your reasoning we would then conclude that the pulsar had the same "CSI" you calculated and conclude that the signal and the pulsar were designed. But we don't do that. We follow the thinking that I've outlined. We start by working out if the existence of pulsars is likely - and when we decide they are, then that is all we need to attribute the signal to natural causes. The pulsar which produces the signal is "simpler" than the signal would be - if you ignore the existence of the pulsar in calculating the probability of the signal.

quote:

---

No it's not unusual. That's how natural selection works! It is going to select anything, even those mutations that degrade biological functions if they increase reproductive fitness. That's just how things are.

---

If it isn't unusual to have a mutation that is strongly beneficial to individuals heterozygous for the allele and very strongly deleterious to those that are homozygous why not produce a few common examples ? Remember to provide evidence that this really is the case.

quote:

---

No they will not? Why should they be correlated with genetic traits to get passed on?

---

I don't know. But since your argument doesn't work unless they are, it is only fair to give you the opportunity to make a case for it.

quote:

---

Therefore, we see here that a non-genetic trait interfeered with positive seletion, and a genetically less fitter individual got selected for by natural selection. This happens all the time in all populations on average with all individuals. Sometimes more, sometimes less ofcourse.

---

And without the correlation - without a systematic bias - it will go the other way and reinforce selection just as often. That is the point. In a large population, noise averages out. That's statistics for you.

quote:

---

I'm sorry but no. This is the ultimate question and the ultimate point of Sanfords book. The point is that if the genome keeps deteriorating, and we saw from Spiegelman's experiment that it does just that, that logically could not have evolved in the first place.

---

Except that it didn't deteriorate. It optimised itself by losing a lot of unnecessary junk. Instead of going into extinction, it was such a great success that it drove it's rivals into extinction. It beat genetic entropy.

quote:

---

Than what are we doing? I know I'm not. But it seems you are by trying to say that the enzyme didn't lose all it's functions. We onlyknow of the one it had and it lost it.

---

I am saying that we DON'T KNOW if it lost all function. How hard is that to understand ?

quote:

---

No, I was arguing from the start that this experiment shows how many mutations does it take for an average enzyme to lose it's known function.

---

Hello! Nobody disagreed with that ! This argument is over whether we KNOW that it lost ALL function. And it seems that you concede that we don't, but you go on arguing and arguing about nothing.

quote:

---

They are all included in the number 10^20.

---

Wrong. The number 10^20 is the estimated number of 4-level concepts. It is Dembski's attempt to compensate for the fact that "bidirectional rotary motor-driven propeller" is chosen in hindsight.Here's the quote again: As you can see, it has nothing to do with considering other ways of forming a "bidirectional rotary motor-driven propeller".

quote:

---

If you think about the other number, the one that describes the event E, in this case 50 proteins, than no. It makes no sense to include anything else. Becasue as I said earlier, some pattern that describes a "bidirectional rotary motor-driven propeller" might be less than 400 bits, and some other more than 1.000.000 bits. So it's obvious that they are different cases of design, or non-design because they a different probability of arising by chance.

---

I have thought about it and it makes no sense to do anything else. As Dembski points out it is P(D*) that needs to be low to conclude design. And if you think about that (and remember the examples discussed before) that makes perfect sense - it makes no sense to use the probability of the exact event - and that is exactly what Dembski does in his handling of the Caputo case. Since D is "bidirectional rotary motor-driven propeller", we need the probability of getting a "bidirectional rotary motor-driven propeller".

quote:

---

D* is 10^20.

---

D* is D considered as an event. In this case that would be "a bidirectional rotary motor-driven propeller". 10^20 is Dembski's estimate of the number of 4-level concepts.

quote:

---

But what if it's not? What if it is a GIANT flagellum that consists of 1.000.000 bits? What than?

---

In that case it won't contribute much to P(D*). That would be obvious to anyone with a basic understanding of the mathematics. Very unlikely events that fall within the specification won't contribute much to the probability of meeting the specification. And 1,000,000 bits is a probability of 2^-1000,000 - very, very unlikely.

quote:

---

I simply took away 20% from 10˘2954.

---

I think that you mean that you took 20% away from the exponent. Because taking 20% away from 10^2954 gets you 8*10^2953. Yes, I already worked out that that was how you applied the 20%. And if you remember I showed exactly why that was wrong.
That's not the calculation I need to know. I need to know how you got the figure of 10^2954 in the first place.

quote:

---

Do I have to quote every single letter? The figure 10^120 is mentioned after the quote. If youa ctually read the whole paper you wouldn't be even mentioning that.

---

Actually you should quote the relevant stuff. The figure 10^120 is just another way of presenting Dembski's universal probability bound (in this case reduced to 400 bits).

quote:

---

Of course not. Pulsars do not correspond to any independently given pattern.

---

The pulsar signal does. It's a regular series of pulses. It's certainly not random. A lighthouse produces much the same sort of signal.

quote:

---

I'm doen providing anythign for you anymore. My point stands, that mutation is, like almost any other, beneficial in some environments, and deleterious in others. That's how natural selection works.

---

In other words you choose to dodge the issue. I said that the sickle-cell allele was unusual as a beneficial mutation because it is only beneficial in the heterozygous state, while being deleterious in the homozygous state. If you want to argue otherwise you have to produce evidence that that is not unusual. Arguing that it is normal in some other respects simply ignores my point.

quote:

---

Obviously you don't know. They are not correlated, they interfere with the geentic selection. Non-genetic traits also get evaluated by natural selection. So since the individual gets evaluated overall, than the genetic traits are just a portion of what gets evaluated. And therefore, positive selection for beneficial mutations and agains deleterious mutations suffers from noise.

---

As I said, it is your argument that would benefit from the existence of the correlation. If we are simply dealing with "noise" then the larger the population, the less effect it has. This is why genetic entropy is only a real problem for small populations.

quote:

---

ut not 100%. Genetic traits are just a small portion of what gets evaluated. You would ahve to have an infinitely large population size for natural selection to be 100% efficient. But you don't have that. Therefore, mutations accumulate.

---

Natural selection doesn't have to be anything like 100% efficient to prevent genetic entropy. All it has to do is to maintain a balance where the average fitness is held at a high enough level to maintain the population. If the population is adequately fit, and the rate at deleterious alleles are lost is the same as the rate at which they enter the population then they aren't accumulating. If you want to prove that genetic entropy is inevitable on theoretical grounds you are going to have to crunch the numbers and find out where that balance point is.

quote:

---

Stop dodging my question. How is this RNA chain, or any other supposed to evolve into a human in 3.6 billion years if it keeps getting shorter!?

---

I already told you that I do not intend to add another separate topic to this discussion. You've seen Admin ask for focus, and suggest narrowing posts down to a single topic. We're already discussing two topics. I decline to add a third.

The ID movement will tolerate a few token non-creationists for PR value. They haven't thrown Behe out, despite his change of position, and I don't think that Berlinski is a creationist. On the other hand, I did run across the story of an ID supporter who says that he was asked to leave (by Philip Johnson, no less) - because he wasn't a creationist.