Understanding through Discussion


Welcome! You are not logged in. [ Login ]
EvC Forum active members: 63 (9072 total)
71 online now:
PaulK (1 member, 70 visitors)
Newest Member: FossilDiscovery
Post Volume: Total: 893,122 Year: 4,234/6,534 Month: 448/900 Week: 154/150 Day: 0/8 Hour: 0/0


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
Author Topic:   Exposing the evolution theory. Part 2
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 188 of 294 (847913)
01-29-2019 1:49 PM
Reply to: Message 187 by WookieeB
01-29-2019 1:27 PM


Re: Hunt versus Axe
quote:

You are making a very fine distinction here. You're implying that "building a system" is very different from ""how the system gets to be". But you haven't explained how a system gets to be IC or given an example.

The distinction between how the system is built and how it becomes IC doesn’t seem that great to me. And I am explaining how a non-IC system can become IC.

quote:

Yes, really. As I explained before, if you have an IC system and you add a some part that is not essential to the function of the system, it does not become a non-IC system. It is just an IC system with an extra part.

A system with a non-essential part is not IC therefore it is a non-IC system.

quote:

Making IC does not consist of taking a system that is pretty much already IC + a non-essential part, and removing that non-essential part.

Changing a non-IC system into an IC system seems to qualify to me.

quote:

I really don't know what you mean by this. If a system is already functioning, parts already "need" other part(s) to have the system function. I think you might be equivocating on what "need" or "essential" means in this hypothetical system you are referring to. How would you take a part that is already finely tuned to work with other parts, and then change it to necessarily work with other parts without affecting it's first task?

Of course there is nothing wrong with changing functions, so long as it doesn’t break the system. And let me remind you that you were insisting that the parts don’t need to be “finely tuned”, just to work with each other. Indeed, it might change by becoming more finely tuned, or losing a non-essential part of its function, or gaining an additional function.

quote:

I thought you brought this up at the very beginning to talk about how to evolve IC.

And I tell you again that it is only the feature of being IC that is of concern.

quote:

I agree. And how do you get IC systems? The only demonstrated way to get an IC system is via design!

I’m going to repeat a question I asked you before. Do you know when a scientist first suggest that evolution should be expected to produce IC systems ?

quote:

Perhaps, but you haven't really given an example of non-IC either. Other than a couple materials being in proximity to each other by happenstance, I'm having difficulty in even conceiving of a non-IC system that has something akin to "parts".

In other words you can’t conceive of a non-IC system.

quote:

Behe's point in highlighting IC is because life appears full of IC systems, and there is no plausible route to creating IC via Darwinian processes.

No plausible route that Behe could think of. Sadly for him, the flaw was in his thinking.


This message is a reply to:
 Message 187 by WookieeB, posted 01-29-2019 1:27 PM WookieeB has replied

Replies to this message:
 Message 191 by WookieeB, posted 01-29-2019 3:09 PM PaulK has replied

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 190 of 294 (847917)
01-29-2019 2:11 PM
Reply to: Message 189 by WookieeB
01-29-2019 1:57 PM


Re: Hunt versus Axe
quote:

But how well that algorithm can climb the hill does depend on the landscape.

Indeed. But since I was only describing the basic algorithm I still wasn’t making any assumptions about the landscape.

quote:

Yes, but if you do not have much in the way of options to climb, and you are limited by time, then your climb ends up remaining static. When you can work out the probabilities, Darwinian mechanisms have a tremendous problem going anywhere.

Now you are making assumptions.

quote:

Isn't that a bit of an oxymoron? How would proteins exist apart from genes?

No, it’s not. It’s quite possible that RNA-based life was using proteins before DNA-based life existed.

quote:

This may depend on what you mean by overlap, but I don't think this helps as much as you think it would. In the vast expanse of all protein function, there is little overlap to be found.

Of course I was referring to an actual example of an overlap. It’s interesting that Axe’s chosen example isn’t isolated, isn’t it ? If overlap is so rare what is the probability against that ?


This message is a reply to:
 Message 189 by WookieeB, posted 01-29-2019 1:57 PM WookieeB has replied

Replies to this message:
 Message 193 by WookieeB, posted 01-29-2019 5:30 PM PaulK has replied

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 192 of 294 (847931)
01-29-2019 4:18 PM
Reply to: Message 191 by WookieeB
01-29-2019 3:09 PM


Re: Hunt versus Axe
quote:

No, your not really. Because whatever you have up till now been saying is non-IC is actually turning out to be IC with something irrelevant on top.....

Which only means that they are non-IC systems.

quote:

No. You're not using the definition correctly. If that part is non-essential than it is not being used to contribute to the essential function (of the IC portion) and it doesn't qualify in the IC definition. I think you are getting hung up in the semantics now.

No, you only wish I was getting hung up in your semantic games.

Remember we are talking about part of the system, just one that is not essential. Which is necessary for the system to be non-IC.

quote:

A "non-essential" part is not "indispensable to maintaining the systems basic function"

There’s a perfect example of not using a definition properly. An IC system cannot have non-essential parts. But we are talking about a non-IC system! Which is not an IC system, and so cannot fit the definition of IC !

quote:

And then back to you needing to explain how an IC system is made? That is the central question. How could evolution do it? So far you only have already assumed some functioning system is already in place with some fluff, and remove the non-important fluff and voila! - an IC system. IC doesn't work that way. You need to explain how to get the stuff without the fluff.

And that is the way - by losing the “fluff”. No I do not have to start with an IC system. If I did you would immediately declare my argument a failure - or you would if you had the sense to realise it.

quote:

Umm, normally it would break a system. In any working system, a part is tuned to have a particular interaction with another part. (Remember Behe's "well-matched" criteria?)

Not if the original function was still performed. Adding an extra function is fine, modifying the function in a way that keeps the essential elements of the original function is fine.

quote:

"Fine tuned", "work with each other", "well-matched" - tere all the same to me

That is really your problem

quote:

So, can you describe how this happens without assuming an IC core pre-exists?

By which you mean “without having a working system” ? Or can you tell me how it is possible to have a working non-IC system without a minimal subset of parts that would adequately perform the necessary functions of the system.

quote:

Can't say with confidence I know the "first". Without using the specific words (irreducible complexity), I do know that Darwin alluded to it when speaking about the eye, and Aristotle (or maybe it was Plato) made statements of a fashion describing it. Why is this important?

I didn’t ask about mentions of irreducible complexity. I mean that idea that evolution would produce irreducible complexity. Seriously, if the theory predicts that there should be some IC systems around Behe’s argument is in trouble.

quote:

No, I can. It's just that many examples are happenstance of proximity and/or rely on rarely random events. Even then, more than two or three parts, and I'm not getting much. The problem comes when ascertaining a system is in place, defining the function, without placing a dependency on whatever is constituting "parts"

Which means that you can’t imagine a non-IC system. Too bad.

quote:

And your plausible route is.....?
Behe’s definition of irreducible complexity was always intended to test a Darwinian explanation where some function is built up gradually over time, a direct evolutionary pathway. You haven't even begun to do that.

The whole point is that it is the indirect paths. Don’t forget that evolution works without foresight. It doesn’t care about constructing systems, all that matters is what works now. One of the problems in Behe’s thinking was that he didn’t see that.


This message is a reply to:
 Message 191 by WookieeB, posted 01-29-2019 3:09 PM WookieeB has replied

Replies to this message:
 Message 195 by WookieeB, posted 01-29-2019 6:31 PM PaulK has taken no action

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 198 of 294 (847984)
01-30-2019 12:21 AM
Reply to: Message 193 by WookieeB
01-29-2019 5:30 PM


Re: Hunt versus Axe
quote:

So is it irrelevant when just explaining a mechanism (regardless of where it is applied). OK.

So now that we have a mechanism, we can apply it to reality and see how well it works....


That is, of course, why Axe’s response to Hunt is so inadequate, on isolation. Hunt looks at reality, Axe simply invents an “analogy” without any comparison to reality.

quote:

No, not really.
Let's take malaria developing a resistance to chloroquine. At minimum, it requires at least two specific point mutations, possibly involving a third. Odds of the resistance developing has been calculated to occur on the rate of 10^20 based on empirical studies.

You do realises that that is the probability of a single - specific - mutation ? Hill climbing doesn’t help there. It isn’t even applicable.

quote:

If you extract these numbers out to eukaryote life, and consider the types of mutations needed (lot more than 2 point mutations) to account for the differences we see, Darwinian processes run into some walls as to what it could do.

Except that hill-climbing does help accumulate useful mutations. More, cases where one very specific mutation is absolutely needed are rare.

quote:

RNA or DNA matters very little. You would still need a translation system. RNA nucleotides do not form proteins directly. Also, where does the initial RNA information come from?

This is a change of subject, to one where very little is known for sure. I will point out that the translation of DNA to protein goes through RNA. Whether the original replicating RNA formed naturally or evolved from a simpler predecessor is unknown, but both are possible.


This message is a reply to:
 Message 193 by WookieeB, posted 01-29-2019 5:30 PM WookieeB has replied

Replies to this message:
 Message 199 by NosyNed, posted 01-30-2019 10:33 AM PaulK has replied
 Message 204 by WookieeB, posted 01-30-2019 2:03 PM PaulK has replied

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 200 of 294 (847993)
01-30-2019 10:58 AM
Reply to: Message 199 by NosyNed
01-30-2019 10:33 AM


Re: Parasite Numbers
It’s better than you think, Ned.

Behe originally argued that two mutations were needed to get any resistance. When it turned out that only one was needed to get some resistance and the second improved it, the odds of evolution succeeding in getting both went way up. And that is because evolution is basically a hill-climbing search (to the extent it is a search).


This message is a reply to:
 Message 199 by NosyNed, posted 01-30-2019 10:33 AM NosyNed has taken no action

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 206 of 294 (848014)
01-30-2019 2:45 PM
Reply to: Message 204 by WookieeB
01-30-2019 2:03 PM


Re: Hunt versus Axe
quote:

You do realize that Hunt's 'look at reality' only applied to TEM-1 relating to DD-Peptidases

I don’t because Hunt’s statement was not restricted to that and the rest of it is consistent with what I’ve seen in other sources. Axe didn’t dispute it either.

quote:

You certainly are not suggesting that all proteins are grouped together, not in isolation, are you?

I am not suggesting that there is one single group and I wouldn’t expect there to be. However I am sure that there is a lot of overlap.

Consider the existence of gene families

quote:

Of course it applies.

Not to a single mutation, as should be obvious. The advantage is in accumulating mutations.

quote:

A mutation that involves (at minimum) 2-3 specific point mutations is the easiest path for malaria to take to survive against chloroquine.

There is one (neutral) mutation that seems to be required, but there are several known routes to achieving resistance in combination with it. Summers et al (2014)

quote:

I agree completely. But accumulating doesn't matter.

It certainly does. That is why evolution works better than a random search.

quote:

The question is how likely will it move up? If the upward-opportunity doesn't present itself, it has nothing to accumulate. That's the random nature of Darwinian processes. NS cannot encourage a directional move. It can only capitalize AFTER the move is made.

The problem for you is that hill-climbing algorithms can and do work. So long as one of the perturbations (mutations in evolution) finds a higher point it can move on. There is no need for the perturbations to automatically find a higher point and it is not a problem if many do not.

quote:

Cases where one very specific mutation is absolutely needed are rare. Yes. The inverse of that means that more that one specific mutation being needed is common. That makes the case much worse for Darwinian processes.

That is a failure of logic on your part since the inverse includes the case where many mutations will do, and that makes it better for Darwinian processes.

quote:

But RNA to protein needs a translator as well - the ribosome (which happens to be made of proteins as well. Chicken and egg problem). RNA may have formed by your two options, or probably not at all (more probable), but even if I granted you that, you still have to account for all the rest that makes proteins.

No, I don’t. I can live with unknowns. So long as it is the case that we do not expect all proteins to be related your point is answered.


This message is a reply to:
 Message 204 by WookieeB, posted 01-30-2019 2:03 PM WookieeB has replied

Replies to this message:
 Message 207 by WookieeB, posted 01-31-2019 12:04 PM PaulK has replied

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 208 of 294 (848073)
01-31-2019 12:30 PM
Reply to: Message 207 by WookieeB
01-31-2019 12:04 PM


Re: Hunt versus Axe
quote:

How does it accumulate mutations?

That is what the movement in sequence space equates to. Or if you want details it’s natural selection keeping successful variants (in the simplified version).

quote:

Please, can you explain what is being searched for in an 'evolutionary' search? And how is that different from a random search?

“What is being searched for” is not closely related to the search algorithm. Evolution search’s for improvements in fitness - in the current environment. I’ve already explained how hill climbing differs from random search.

quote:

No, I have no problem with the hill-climbing algorithm. It's not an issue of whether it can move up. It is an issue of whether it will move up. If you have a 1 in 10 chance to move up (where all the other 9 chances is not moving up), then it will only move up 1 in 10 tries. Now, if you are only allowed 2 tries, it is more likely to miss than to move up. No hill is climbed unless you happen to hit that '1' successful option.

And that leads us to the point where you would need numbers to say anything useful. Numbers you don’t have,

quote:

Oh, so your hinging your statement on the "absolutely needed" part? You've got to be kidding me. Intuition, testing, and all experience indicates that mutating something is FAR, FAR more likely to break the thing being mutated than it is to help it. Even neutral changes would outnumber beneficial changes by a large factor. So mutations will heavily be tilted more to to break than to benefit.

No, I’m hinging my point on there typically being a number of mutations that would do rather than one particular one. Which is what I said in the first place.

quote:

What? Protein relation is irrelevant at this point.

No it is the point you are trying to divert from. If we don’t expect all proteins to be related then we should also expect there to be groups of related proteins that may well be isolated from other groups - at least so far as evolution is concerned.

I don’t need to explain the rest because THAT is irrelevant. We don’t know how it happened but by the evidence we have, it seems more likely that we somehow got from RNA life to proteins and DNA. And I don’t expect to know without a lot more research - if it is even knowable this late in the history of life. It is a complex problem and evidence is hard to come by.


This message is a reply to:
 Message 207 by WookieeB, posted 01-31-2019 12:04 PM WookieeB has replied

Replies to this message:
 Message 209 by WookieeB, posted 02-01-2019 3:57 PM PaulK has replied

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 210 of 294 (848204)
02-01-2019 4:43 PM
Reply to: Message 209 by WookieeB
02-01-2019 3:57 PM


quote:

So NS keeps a successful variant. Why? We would have to assume whatever you are constituting is a "variant" improves fitness. But NS cannot act on anything until the variant appears. So what is the probability of a variant appearing? NS cannot affect that probability.

Since improvement in fitness is success (and performing a useful function or doing it better will often improve fitness) natural selection will automatically tend to keep successful variants.

And the probability of variants in sequence appearing is the probability of a mutation occurring which I am sure you can look up. (And in the long term evolution may have “tuned” that probability, so it may not be completely true that natural selection can’t affect it)

quote:

Pardon??? Can you read that statement again and tell me you're serious?

It’s true. After all you imagine a random search rather than a hill-climbing search for evolution but still searching for the same things.

quote:

No, you haven't explained the difference because I think you are confused over how a hill-climb actually searches.

Nope, I already have explained. But here it is again. A random search simply picks points in search space at random. A hill climbing search explores points around the a location in search space until it finds a “higher” point, whereupon it moves to that point and starts exploring around there.

quote:

I didn't really notice it before, but you are not using consistent terms here when comparing both. When you mention hill-climbing, you said it "perturbs a parameter". No, it doesn't. In the context of what you are describing (the search itself), NS cannot do anything to a parameter, it cannot 'make' (perturb, influence, nudge, etc) a higher point into being. All it can do is wait for a search to reveal a higher point (and not necessarily the highest point to be found) and the lock that one in if found.

So I gave an explanation of a hill climbing algorithm that is clearly distinct from random search. And you can’t see how that relates to evolution since I used general terminology and you can’t see how it relates to evolution. Even though it is pretty simple. The fact that you keep wrongly equating evolution to Natural Selection isn’t helping you either.

To put it simply mutation varies the sequence of a gene - that is perturbing a parameter.

quote:

The search process itself is RANDOM.

No. The perturbations are random. The search process is more than that,

quote:

The search process itself is mutations occurring. NS cannot affect what/where/how mutations occur. It is a wholly random process, thus the search itself is a RANDOM SEARCH

In a random search the next sequence tried could be anything. Mutations do NOT completely randomise the genome at every generation. But that is what you would need for it to be a random search.

quote:

"A random search simply chooses completely random points until it hits the target" is true, which is what is going on with your hill search when it actually is searching. "...with no feedback at all." is a bit misleading, since a random search is looking for a target too (same as NS), and that is the feedback it gets.

The random search does not use feedback in any way to help it choose the next point. A hill-climbing search does, as I have explained.

quote:

If that number can be taken in any way as an example of how Darwinian processes work in general (and there is no reason why it cannot), then it doesn't look good for the Darwinian paradigm to explain much.

I think you need more than one example to justify using it as a general probability.

quote:

I agree. And if that is the case, it doesn't provide much confidence in evolution to explain much

Nor does it dent the confidence in evolution explaining a great deal of what is observed in biology. Evolution has always required some initial state which it cannot explain - and the earliest life has left so little evidence that any explanations involving it will tend to be highly speculative at best. So really the point doesn’t change a thing.


This message is a reply to:
 Message 209 by WookieeB, posted 02-01-2019 3:57 PM WookieeB has replied

Replies to this message:
 Message 213 by WookieeB, posted 02-01-2019 6:54 PM PaulK has replied

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 220 of 294 (848289)
02-02-2019 3:43 PM
Reply to: Message 213 by WookieeB
02-01-2019 6:54 PM


quote:

Ya, so what. Mutations happen. NS in a search cannot affect the rate, location, or type of mutation that occurs. It can only just view the mutation and do something AFTER a beneficial mutation occurs (assuming there is a beneficial mutation to be had - not a at a local peak or flat plateau)

As I said before, the mutations are just the perturbations, it’s the overall movement through sequence space that matters, and Natural Selection is a major factor there.

quote:

If not, you tell me what a random search is looking for. It has to have a target, all searches do.

Evolution does not have a target as such. There is no end state, other than extinction - which evolution works against.

quote:

How does it "explore"? It looks in a random manner around it, just like in a random search.
Random search is same as your hill-climbing search.

The descriptions are very similar, although that just means that random search isn’t so bad. At least that version of random search.

quote:

No! That is not what a random search is. What you just said is like morphing the landscape, then picking a random direction and moving that way.

I’m certainly not talking about morphing anything I’d consider the landscape. Evolution works on the whole genome, so selecting a random point in the space would be selecting a random point for the whole genome.

quote:

I do agree that Darwinian processes can explain some things in biology. But it cannot explain everything. Things categorized as microevolution - sure. But for the origins of life and macroevolution, which includes new proteins and body plans, IC structures, and all the new information that is required for them - Darwinism is not up to the task.

It certainly manages new proteins and IC structures. I see no reason why it can’t manage body plans. Granted the theory could do with improvements, such as integrating better with developmental biology, but there doesn’t seem to be any reason to abandon it, rather than augmenting it, for anything in its scope.


This message is a reply to:
 Message 213 by WookieeB, posted 02-01-2019 6:54 PM WookieeB has replied

Replies to this message:
 Message 222 by WookieeB, posted 02-04-2019 12:15 PM PaulK has replied

  
PaulK
Member
Posts: 17167
Joined: 01-10-2003
Member Rating: 3.6


Message 223 of 294 (848408)
02-04-2019 12:48 PM
Reply to: Message 222 by WookieeB
02-04-2019 12:15 PM


quote:

No objection here. Random search behaves exactly the same. The real number possibilities is where the question comes in. If NS only allows moves up, what happens if no up perturbation is presented to it?

When that happens - and it does - there’s no movement.

quote:

Don't try to change the frame of reference on me. We've been couching this subject in the terms of 'Search'. In a search, there is ALWAYS a target.

We’re only talking about search insofar as it relates to evolution. And evolution doesn’t have a target. If you want to talk about aspects of search that are irrelevant to evolution then you are trying to change the subject.

quote:

So...then your hill-climb is the same as a random search. You haven't distinguished a difference.

And “random search” - in that sense - isn’t very random and should be quite effective.

quote:

Those are your statements. If you are suggesting that a random search randomizes the entire genome, you are incorrect.

A purely random search would. And would be ineffective because it has to search large parts of sequence space and can’t even capitalise on near-misses.

quote:

Interesting phrase - manage. What does that mean?

It should be obvious. Evolution produced examples of those things.

quote:

As for creating those things, evolution cannot account for them.

I’ve yet to see a good reason to think that if can’t.

quote:

As for body plans, since it hasn't really been identified where the information for them resides, you cannot infer why evolution WOULD be able to manage them.

More importantly you cannot infer that it can’t.


This message is a reply to:
 Message 222 by WookieeB, posted 02-04-2019 12:15 PM WookieeB has taken no action

  
Newer Topic | Older Topic
Jump to:


Copyright 2001-2018 by EvC Forum, All Rights Reserved

™ Version 4.1
Innovative software from Qwixotic © 2022