Evolution vs Creationism

[Molal]

I agree with that at least.

I'll bow out of this... I shouldn't have tried to argue anything when I don't have access to all the journals etc I need to be able to cite.

Out of interest... what peer reviewed journals would you accept?

Well, something to support your statistics. I know that you are a statistician and I trust you, I would just like to get my hand on some published research so I can read it and absorb it.

I am a structural geologist, not a mathematician.

Anyway, I hope our paths cross again in the future, say Hi to your wife and look after yourself.

 

[Kitangel]

Well, something to support your statistics. I know that you are a statistician and I trust you, I would just like to get my hand on some published research so I can read it and absorb it.

I am a structural geologist, not a mathematician.

Anyway, I hope our paths cross again in the future, say Hi to your wife and look after yourself.

Hi back

 

[BlueIceDragon]

I was just passed this paper:
http://creationontheweb.com/content/view/5467

which may be of interest relating to cell death and why it is too important to have evolved.

I was looking for a paper that explained my 'information loss theory' (which to me makes sense statistically, but I'd need to do simulations to 'prove it' if I can't find a direct link).

 

[shernren]

Evolution is the most inefficient way of designing something known to man. This is because it is NOT a method of design, but rather a method of enabling adaptation. Adaptation is NOT evolution.

That's a strange statement to say, especially from someone who knows well enough not to use Excel for serious statistics Haven't you heard of genetic algorithms that use evolution to design things?

http://www.talkorigins.org/faqs/genalg/genalg.html

For example, a genetic algorithm developed jointly by engineers from General Electric and Rensselaer Polytechnic Institute produced a high-performance jet engine turbine design that was three times better than a human-designed configuration and 50% better than a configuration designed by an expert system by successfully navigating a solution space containing more than 10^387 possibilities. Conventional methods for designing such turbines are a central part of engineering projects that can take up to five years and cost over $2 billion; the genetic algorithm discovered this solution after two days on a typical engineering desktop workstation (Holland 1992, p.72).

Tornado in a junkyard!

 

[BlueIceDragon]

Well, something to support your statistics. I know that you are a statistician and I trust you, I would just like to get my hand on some published research so I can read it and absorb it.

I am a structural geologist, not a mathematician.

Anyway, I hope our paths cross again in the future, say Hi to your wife and look after yourself.

Will do. And look after yourself too

 

[BlueIceDragon]

That's a strange statement to say, especially from someone who knows well enough not to use Excel for serious statistics Haven't you heard of genetic algorithms that use evolution to design things?

http://www.talkorigins.org/faqs/genalg/genalg.html

For example, a genetic algorithm developed jointly by engineers from General Electric and Rensselaer Polytechnic Institute produced a high-performance jet engine turbine design that was three times better than a human-designed configuration and 50% better than a configuration designed by an expert system by successfully navigating a solution space containing more than 10^387 possibilities. Conventional methods for designing such turbines are a central part of engineering projects that can take up to five years and cost over $2 billion; the genetic algorithm discovered this solution after two days on a typical engineering desktop workstation (Holland 1992, p.72).

Tornado in a junkyard!

Did they start with junk? Did they know what they were after in the first place (an efficient turbine engine)? Not meaning to be completely sarcastic here, but if anything it shows that design was needed in the first place for this to hope to work.

Ah yes, uh, I have - though I hadn't read of this particular solution (thanks for sharing!). Just because it is USED for design doesn't make it an efficient design tool! Something that needs to evaulate 10^387 different combinations in order to find something that 'work's doesn't seem to be efficient to me....

A) You have to know what you want
B) You have to provide the building blocks for the program to solve the problem. If the wrong blocks, or insufficient blocks are provided... the problem will not be solved.

So much care has to be taken in designing the whole genetic algorithm / programming structure and in the computation time to produce results that only "simple" problems can be solved. Sometimes quite unusually, granted, but still "simple" problems. Furthermore, you have to be VERY careful not to solve the wrong problem.

Last edit: it depends on what you mean by efficient and how you measure efficiency. If you measure it in 'human time to solve the problem (days) then you might be able to consider GA/GP efficient. If you consider efficiency in terms of "number of solutions that need to be evaluated to solve the problem" you get a very different picture. If it takes 10^387 evaluations for a program to solve a turbine jet problem where the program exists and the environment is carefully constructed to make solving the problem possible AND the right building components are provided to enable construction what hope have we?

 

[shernren]

Did they start with junk? Did they know what they were after in the first place (an efficient turbine engine)? Not meaning to be completely sarcastic here, but if anything it shows that design was needed in the first place for this to hope to work.

It's exactly the same as evolution. Evolution imposes fitness functions on variants in a population and the best survive to reproduce. Just the same as the engineers did.

Ah yes, uh, I have - though I hadn't read of this particular solution (thanks for sharing!). Just because it is USED for design doesn't make it an efficient design tool! Something that needs to evaulate 10^387 different combinations in order to find something that 'work's doesn't seem to be efficient to me....

A) You have to know what you want
B) You have to provide the building blocks for the program to solve the problem. If the wrong blocks, or insufficient blocks are provided... the problem will not be solved.

So much care has to be taken in designing the whole genetic algorithm / programming structure and in the computation time to produce results that only "simple" problems can be solved. Sometimes quite unusually, granted, but still "simple" problems. Furthermore, you have to be VERY careful not to solve the wrong problem.

Last edit: it depends on what you mean by efficient and how you measure efficiency. If you measure it in 'human time to solve the problem (days) then you might be able to consider GA/GP efficient. If you consider efficiency in terms of "number of solutions that need to be evaluated to solve the problem" you get a very different picture. If it takes 10^387 evaluations for a program to solve a turbine jet problem where the program exists and the environment is carefully constructed to make solving the problem possible AND the right building components are provided to enable construction what hope have we?

The probability space had size 10^387. We haven't built a computer yet that can perform 10^387 calculations in two days. The entire BOINC system, which distributes work out to lots and lots of home computers that run it in their spare time, averages about a petaflops (floating-point operation per second), which over two days would give a measly 2 x 10^20 calculations. The very fastest supercomputer in the world (which is specialized for molecular dynamics computations and doesn't do anything else) also averages about a petaflop. The fastest PC processors do about 30 gigaflops, which over two days gives about 5 x 10^16 calculations. An evolutionary algorithm can use that little calculations to explore a humongous probability space. Inefficient? Not a chance.

 

[BlueIceDragon]

The probability space had size 10^387. We haven't built a computer yet that can perform 10^387 calculations in two days.

Yeah I should have noticed that. I'll blame it on the fact that it was midnight when I read your post. My apologies.

Compared with the total population possibilities... yeah, okay, it's efficient. Whether it is necessary to evaulate 2x10^20 calculations to find the solution... I don't know.

That's still a heck of a lot of failures...

 

[shernren]

Yeah I should have noticed that. I'll blame it on the fact that it was midnight when I read your post. My apologies.

Compared with the total population possibilities... yeah, okay, it's efficient. Whether it is necessary to evaulate 2x10^20 calculations to find the solution... I don't know.

That's still a heck of a lot of failures...

Compare that with the amount of failures a five-year billion-dollar straight design project must have experienced in its course.

Maybe God knew what He was doing after all when He decided to use evolution, no?

 

[Hnefi]

This is really hard for me to explain without resorting to pictures since it is in pictures that it the easiest to explain.

Still, again I shall use the genetic programming example since it is the easiest to talk about and, in this case, it does directly translate to biology.

Let's take a population of "Super Frogs". Now these super frogs have the potential with all the information in them (characteristics: colour, shape, size, etc) to specialise to be any "kind" of frog. Kind of like the "common ancestor of frogs".

Draw a circle to represent this population. Call it A.
Draw another circle with a dotted line encompassing all of A. We'll call it A'

This outer circle, A', is the potential reach of A from the current population of A. That is, it defines the reach of the possible population with mutations, breeding etc.

Now, logically, as these super frogs reproduce etc, the population becomes specialised to area. Some frogs only survive if they are large, others if they are small, others again if they are green, others if they swim fast. Whatever the reason, different features of the population A are 'bred out' due to being unfit.

Draw these four circles within the A circle - again extend by drawing dotted lines around them.

It is completely normal, and indeed expected, that the area covered by A and A' significantly outweighs the area covered by the small specialised populations. The combinatory possibilities have been reduced. Multiply this by a large number of generations and you have an instant problem. Diversification is supposedly increased - you have a large number of different types of frogs - but in actual fact genetic potential has been lost.

The whole design of genetic programming was built around this logical principle. Otherwise "natural selection" would be a complete waste of time and we certainly would not be able to solve software problems with it.

This IS happening with every population round the world. We are getting greater "specialisation", but at the cost of lower information potential. Kinda like kinetic versus potential energy.

Why do you think mongrels have less disease problems than particular breeds? It is purely because the information loss is not the same. There is less likelihood that the information necessary to combat disease X has been bred out of them.

I'm afraid you are wrong. What you say only applies to certain genetic algorithms that are specifically designed to find a single global optimum to a particular problem. Typical for these setups is to start with a population that is greatly diverse - usually random - and then let them adapt to a global, static fitness function.

But that's not how it is in nature. First of all, the starting population is usually not very genetically diverse at all. We humans are largely genetically identical to each, for example. Secondly, the fitness function is neither static nor global. A cheetah lives in a different environment than a polar bear, and therefore develops a different genome. And since other creatures are part of the environment, the fitness function is never the same for any two populations; that's why we get hunter-prey arms races in nature.

When you set up a genetic algorithm that mimics nature in this regard, you will find that the surface area of A' expands over time.

 

[gluadys]

A fundamental flaw huh?

My understanding of evolution was that it is a hypothesis of how the variety of species we observe today came into existance through breeding (cross-fertilisation)

Well, there is the first fundamental flaw. Evolution does not require cross-fertilization.

The hypothesis is that through the mixing of the genetic information new combinations are produced.

No, this is not the hypothesis of evolution. You are talking about the independent inheritance of genes as discovered by Mendel.

There is a genetic basis to evolution--of course--but you do not seem to be clear on what it is.

What does "distribution of alleles" mean to you?

If evolution somehow changed it's definition over the last 1000 years please do enlighten me.

Yes, the definition of evolution has changed in the last 1000 years. Most evolutionary theories up to the 19th century were "great chain of being" theories which were more about the evolution of the soul than of species.

Then you get Lamark's theory of evolution via the inheritance of acquired characteristics.

Neither of these is applicable to today's definition.

Then in 1859 Darwin publishes his thesis of evolution of species via natural selection. This is still a bedrock of evolution. In the 1930's the neo-Darwinian definition which incorporates the insights of genetics into natural selection was created and that is the one still in use today.

It's A hypothesis of how the species came into existance, but it fails spectacularly:
a) to explain how the information in the DNA came into being in the first place
b) how the interpretor of the DNA came into being (if you look at this closely, you observe a circular argument because you need the interpretor to reproduce DNA yet you need information in the DNA to build the interpretor... go figure).

Neither of these pieces of information is necessary to the theory of evolution. They will be essential to any theory of abiogenesis when one is formed.

c) So many things need to evolve at the same time to be useful. For example, the complexity of the human body in regard to male plus female and the all the interactions which need to occur in exactly the right order for reproduction to happen simply would not happen by chance.

Evolution does not happen by chance. You are confusing one aspect of the mechanism of evolution (mutations) with evolution itself.

What does co-evolution mean to you?

Evolution is the most inefficient way of designing something known to man. This is because it is NOT a method of design, but rather a method of enabling adaptation. Adaptation is NOT evolution.

I agree. Adaptation is not evolution. But evolution enables adaptation. IOW there is no adaptation without evolution.

Do you agree?

As for the efficiency of evolution as designer, tell that to the engineers using genetic algorithms to solve complex problems of design.

 

[BlueIceDragon]

Compare that with the amount of failures a five-year billion-dollar straight design project must have experienced in its course.

Maybe God knew what He was doing after all when He decided to use evolution, no?

Big difference.

We're finite limited human beings who don't KNOW the answer to everything.

God is an infinite, omnsicience being - why the heck would HE waste time on an evolutionary project when He can design the thing in picoseconds?

 

[BlueIceDragon]

As for the efficiency of evolution as designer, tell that to the engineers using genetic algorithms to solve complex problems of design.

Yeah well, apparently I know nothing about GA or GP despite getting an A+ for my design of a GP program... *smiles sarcastically* Then again, it was a number of years ago... so maybe the design of the GA/GP paradigm had a punctuated equilibrium jump. Though at the time all the literature was very firm on the limitations of GA and GP... *shrugs* Koza's books are extremely optimistic.

In the mean time, NO ONE has addressed the problem which GA and GP begin with at the start.

1) You need building blocks
2) You need SOMETHING to evaluate your fitness by - I don't care whether it is a defined objective or whether you're blindly stabbing in the dark for something only you can understand once you see it. SOMEONE or SOMETHING needs to either be watching it's progress or monitoring it.
4) Also... someone has to 'start' it... with everything set in place, ready to go, before you CAN start it. I mean, the whole process is incredibily useful if you press "run" when all you've written is main() {}.
5) If no one is watching and/or made it.... well, go solve the problem of abiogensis before trying to solve the problem of how life works. If you don't understand the beginnings how can you understand the now. That's why we *should* study history right?

No comment on the number of times you GA/GP engineers here have designed a GA program and let it run only to find it solved the wrong problem and/or failed for some other reason.

 

[BlueIceDragon]

I'm afraid you are wrong. What you say only applies to certain genetic algorithms that are specifically designed to find a single global optimum to a particular problem. Typical for these setups is to start with a population that is greatly diverse - usually random - and then let them adapt to a global, static fitness function.

But that's not how it is in nature. First of all, the starting population is usually not very genetically diverse at all. We humans are largely genetically identical to each, for example. Secondly, the fitness function is neither static nor global. A cheetah lives in a different environment than a polar bear, and therefore develops a different genome. And since other creatures are part of the environment, the fitness function is never the same for any two populations; that's why we get hunter-prey arms races in nature.

When you set up a genetic algorithm that mimics nature in this regard, you will find that the surface area of A' expands over time.

The fitness function doens't have to be the same for two populations for my argument to work. However, that's not the point of the argument. If you start at a common ancestor move into little subpopulations then keep resampling even with differing population sizes... you still willl keep *roughly* the same characteristics.

And there are global static functions which apply to life... death is a 'global static function'.

Plenty of animal, and especially plant, species live longer than humans... what happened to our 'higher evolution'?

A' expands over time... you're meaning by inserting junk in the "GA DNA string"?? If you are... then I don't think that really counts... because you're not "adding" to the search space... all you're trying to do is to protect what's there.

 

[Molal]

As you are aware, I am not a mathematician, but I am a scientist and it appears to me, BlueIceDragon, that you have the cart before the horse.

The reality is that the evidence for evolution exists and evolution occurs. This is not a function of belief or faith, evolution is. To argue that it cannot through statistics seems pointless. Regardless of what statistical analysis tells us, evolution does occur. If you think that statistical analysis indicates that evolution does not occur, then you must provide evidence for your assertion - I know, I harp on about this, but it is of paramount importance.

And so, I think your argument can be boiled down to the fact that you don't want it to be true because it contradicts your biblical interpretation - it creates cognitive dissonance.

 

[BlueIceDragon]

Well, there is the first fundamental flaw. Evolution does not require cross-fertilization.

If you're going to quote me, please finish my sentences...

I SAID breeding and mutations. The cross-fertilisation was in brackets for a reason - namely that breeding permits cross-fertilisation. Anyhow, I thoroughly disagree with your statement as it stands.

Evolution does require cross fertilisation - not because it needs it to exist, but it needs it to be effective. Mutations are incredibily dangerous factors if there isn't enough junk in the code to make sure that the probability of a mutation being dangerous is very small. This is why any GA/GP program with a high mutation application invariably develops a large amount of junk. This proportion of junk can be directly related as a function to the porportion of mutation application versus 'cross-over' (aka breeding) application.

Obviously, true breeding in nature is more complicated than that because the breeding operation itself "knows" about the status quo in the DNA and actually tries to reproduce that. It seeks to avoid mutations rather than to produce them. Mutations are accidental - not part of the design. This makes sense of course... if the design was for mutations to be avoided.

No, this is not the hypothesis of evolution. You are talking about the independent inheritance of genes as discovered by Mendel.

Right. Since evolution isn't about producing new combinations of 'things' ie: one more rib, one more limb, a few more eyes... what IS it about?

There is a genetic basis to evolution--of course--but you do not seem to be clear on what it is.

Do you care to enlighten me then?

What does "distribution of alleles" mean to you?

The natural variation in the features common to a characteristic (eg: eye colour, skin colour, etc). Changing these do not produce new 'sub-species' but information CAN be removed from the population unless reintroduced by accidental mutation.

Okay... so I used a bad example of colour being a feature in my frog example.

Yes, the definition of evolution has changed in the last 1000 years. Most evolutionary theories up to the 19th century were "great chain of being" theories which were more about the evolution of the soul than of species.

Then you get Lamark's theory of evolution via the inheritance of acquired characteristics.

Neither of these is applicable to today's definition.

Then in 1859 Darwin publishes his thesis of evolution of species via natural selection. This is still a bedrock of evolution. In the 1930's the neo-Darwinian definition which incorporates the insights of genetics into natural selection was created and that is the one still in use today.

Fine.... that wasn't quite what I was meaning... but okay. Thanks.

I understood that the theory of evolution was still in a process of equilibrium...

Neither of these pieces of information is necessary to the theory of evolution. They will be essential to any theory of abiogenesis when one is formed.

Yup I'm sure that rocks do leap up and start dancing somewhere...

Evolution does not happen by chance. You are confusing one aspect of the mechanism of evolution (mutations) with evolution itself.

*Bursts out laughing*

Thank you!!! You've made my day. What exactly is the 'non-chance' bit of evolution since this contradicts every part of the theory of evolution I've read so far?

If I'm confusing the mechanics with evolution, and evolution needs the mechanics to exist I'm guessing pure chaos is actually purely deterministic and therefore not chaotic at all...

What does co-evolution mean to you?

Two things 'adapting' together because their fitness function indirectly influences each other rather than direct competition because of conflicting fitness functions...

I agree. Adaptation is not evolution. But evolution enables adaptation. IOW there is no adaptation without evolution.

You've got it the wrong way round. Evolution does not enable adaptation. Adaptation might enable evolution but I shall just end with various quotes and then leave this thread completely... and I do mean it this time lol.

Some quotes. I had to quote this one since I found it.... For all the theistic evolutionists

‘Oh but of course the story of Adam and Eve was only ever symbolic, wasn’t it? Symbolic?! Jesus had himself tortured and executed for a symbolic sin by a non-existent individual. Nobody not brought up in the faith could reach any verdict other than barking mad!’

- Richard Dawkins,The root of all evil? (broadcast on Channel 4, 16 January 2006)

‘The interpretation of evolution is in a state of upheaval: the rapid advancement of Molecular Biology has led into question many of the tenets of Darwinism and neo-Darwinism which, although valuable approaches at the time they were formulated, never fulfilled the criteria demanded by real scientific theories… In the author’s opinion, no real theory of evolution can be formulated at present.’

Evolution Without Selection: FORM AND FUNCTION BY AUTOEVOLUTION by A. LIMA-DE-FARIA
http://creationontheweb.com/content/view/1160/105/

"Evolution has been observed. It's just that it hasn't been observed while it's happening. ... It is rather like a detective coming on a murder after the scene… the detective hasn't actually seen the murder take place, of course. But what you do see is a massive clue ... Huge quantities of circumstantial evidence. It might as well be spelled out in words of English."

- Richard Dawkins

The first sentences are correct...

As for the detective coming on the murder scene, it is ironic that the detective in this case views the circumstantial evidence and refuses to consider any alternatives other than ideas the detective already has. This detective would look at a well laid out crime scene destined to point to a particular antagonist and would refuse to consider that the crime scene had been set up. A rather pathetic detective wouldn't you think??

Anyway, I'm off now... have fun guys.

 

[gluadys]

In the mean time, NO ONE has addressed the problem which GA and GP begin with at the start.

Because they are not applicable to evolution. Humans have to write the programs for GA/GP. But God has already written the program for evolution.

1) You need building blocks

Evolution begins with the building blocks already in place: the genetic information in living/reproducing organisms.

2) You need SOMETHING to evaluate your fitness by - I don't care whether it is a defined objective or whether you're blindly stabbing in the dark for something only you can understand once you see it. SOMEONE or SOMETHING needs to either be watching it's progress or monitoring it.

Evolution occurs with a built-in fitness evaluator: the current local environment of the species.

4) Also... someone has to 'start' it... with everything set in place, ready to go, before you CAN start it. I mean, the whole process is incredibily useful if you press "run" when all you've written is main() {}.

Reproduction is the biological "run" button.

5) If no one is watching and/or made it.... well, go solve the problem of abiogensis before trying to solve the problem of how life works.

Exactly. The questions you are raising are questions that pertain to abiogenesis, not evolution.

Evolution studies how living things change across generations. We can study that without even asking how life came to be.

Of course, we are curious about how life came to be and scientists are working on that, but it is a study independent of evolution.

If you don't understand the beginnings how can you understand the now.

By studying the now. Do you need to study ancient chariot-making to understand automobiles?

 

[BlueIceDragon]

As you are aware, I am not a mathematician, but I am a scientist and it appears to me, BlueIceDragon, that you have the cart before the horse.

The reality is that the evidence for evolution exists and evolution occurs. This is not a function of belief or faith, evolution is. To argue that it cannot through statistics seems pointless. Regardless of what statistical analysis tells us, evolution does occur. If you think that statistical analysis indicates that evolution does not occur, then you must provide evidence for your assertion - I know, I harp on about this, but it is of paramount importance.

And so, I think your argument can be boiled down to the fact that you don't want it to be true because it contradicts your biblical interpretation - it creates cognitive dissonance.

Your definition of evolution and mine might be different then. I'm happy for "evolution" (loosely defined) to occur IF and only IF God started with a bunch of 'common ancestors' which he made fully complete, perfectly defined, from scratch as defined in Genesis 1 and death did not exist till Chapter 3 of Genesis. Evolution therein is a natural adaptation of these 'common ancestors' into the variety of critturs we see today.

David out.

 

[Hnefi]

Yeah well, apparently I know nothing about GA or GP despite getting an A+ for my design of a GP program... *smiles sarcastically* Then again, it was a number of years ago... so maybe the design of the GA/GP paradigm had a punctuated equilibrium jump. Though at the time all the literature was very firm on the limitations of GA and GP... *shrugs* Koza's books are extremely optimistic.

Just because you got a good grade in a course doesn't mean you can't do mistakes. You are not the only one who is or has been studying these things at the university level.

In the mean time, NO ONE has addressed the problem which GA and GP begin with at the start.

1) You need building blocks

That has nothing to do with evolution. It is a separate - but obviously related - field of research, usually referred to as abiogenesis. Now, if you wonder whether we have any answer to the question of how life came to be, AFAIK there are no complete ones yet. You'll just have to wait. But again, that has nothing to do with evolution.

2) You need SOMETHING to evaluate your fitness by - I don't care whether it is a defined objective or whether you're blindly stabbing in the dark for something only you can understand once you see it. SOMEONE or SOMETHING needs to either be watching it's progress or monitoring it.

The evaluation function for evolution is very simple - reproductive success. The more an individual is able to produce viable offspring, the more fit it is. This fitness function is obviously natural and requires no monitoring, adjusting or control of any kind.

4) Also... someone has to 'start' it... with everything set in place, ready to go, before you CAN start it. I mean, the whole process is incredibily useful if you press "run" when all you've written is main() {}.

Arguing by analogy can be useful, but don't push it too far. GA mimics evolution, but programming does not mimic abiogenesis. There are few or no similarities between coding a program and self-replicating molecules arising in a lifeless environment. The question of origins cannot be answered or described by GA, only what happened after life appeared.

5) If no one is watching and/or made it.... well, go solve the problem of abiogensis before trying to solve the problem of how life works. If you don't understand the beginnings how can you understand the now. That's why we *should* study history right?

Nonsense. Generally speaking, the question of origins is usually one of the least useful concepts in any theory, and also the hardest question to answer. We have many theories that are of great use to us eventhough we have no idea how they came to be. Germ theory, for example, or atomic theory, or the theories of relativity. None of these address origins, but all of them are very useful to us. We don't exactly know how or why electrons first appeared, but that doesn't make the electronics in your computer any less useful.

It might not even be possible to answer the ultimate question of origins. By your argument, we should therefore reject ALL knowledge since we cannot understand the beginnings of our existence.

No comment on the number of times you GA/GP engineers here have designed a GA program and let it run only to find it solved the wrong problem and/or failed for some other reason.

What does that have to do with anything?

The fitness function doens't have to be the same for two populations for my argument to work. However, that's not the point of the argument. If you start at a common ancestor move into little subpopulations then keep resampling even with differing population sizes... you still willl keep *roughly* the same characteristics.

No, you probably won't, not unless the problem you are trying to solve only has one solution. If you make the fitness function for one population depend on what happens in another (as is the case in nature), you are almost guaranteed to get divergent characteristics.

And there are global static functions which apply to life... death is a 'global static function'.

Death is not the fitness function in evolution, so I don't see your point.

Plenty of animal, and especially plant, species live longer than humans... what happened to our 'higher evolution'?

"Higher evolution"? I have never heard of such a term. Did you just invent it?

No offense, but I'm getting the impression that you are just throwing a bunch of ad-hoc arguments to see what sticks. It would help if you thought your position through a bit more carefully to avoid all these red herrings and non-sequiteurs.

A' expands over time... you're meaning by inserting junk in the "GA DNA string"?? If you are... then I don't think that really counts... because you're not "adding" to the search space... all you're trying to do is to protect what's there.

What? I'm saying that if you have several separated populations that are evolving according to a genetic algorithm and the fitness function for each population is sufficiently unique, the characteristics of the populations will diverge - even if they were identical at the start. When they diverge, they will traverse different parts and increasingly large subspaces of the solution space, which I stated in your terms as "A' expands over time".

 

[gluadys]

If you're going to quote me, please finish my sentences...

I SAID breeding and mutations. The cross-fertilisation was in brackets for a reason - namely that breeding permits cross-fertilisation.

We will get to mutations later. Unless I misunderstand you, cross-fertilization refers to bringing together two individuals with differing sets of genetic variations.

This is NOT necessary to evolution.

You can begin with a totally homogenous population, maintain scrupulous control to prevent cross-breeding with a variant population, and you will still get evolution.

Now, we can talk about mutations.

Mutations are incredibily dangerous factors if there isn't enough junk in the code to make sure that the probability of a mutation being dangerous is very small.

No disagreement there.

Mutations are accidental - not part of the design. This makes sense of course... if the design was for mutations to be avoided.

As far as this goes, it is correct. Nevertheless mutations occur. From a theistic point of view (even an intelligent design point of view) the creator apparently designed life to include the possibility of mutations.

Whether the creator decided to create some particular mutations is a different question. Some believe so.

Right. Since evolution isn't about producing new combinations of 'things' ie: one more rib, one more limb, a few more eyes... what IS it about?

No, it is not about that. Evolution as a biological process was not designed to produce these things. It is not a teleological process in which these things are goals. They are, however, consequences of the process.

Do you care to enlighten me then?

If you are willing to be enlightened.

The natural variation in the features common to a characteristic (eg: eye colour, skin colour, etc).

Not quite. Characteristic features are a matter of gene expression. Variation in these features implies variation in gene expression.

Would you care to define "allele" in this context?

Also, what does distribution of alleles refer to?

Changing these do not produce new 'sub-species' but information CAN be removed from the population unless reintroduced by accidental mutation.

So you agree that mutations can introduce/re-introduce information into the gene pool of a species?

What exactly is the 'non-chance' bit of evolution since this contradicts every part of the theory of evolution I've read so far?

Surely you have never read anything on evolution that fails to mention natural selection?

Mutation introduces randomness; natural selection applies a non-random filter. Evolution cannot occur without both mutation and natural selection. Therefore the randomness which is a characteristic of mutation, cannot be applied to the whole process of evolution. That is mistaking a part for the whole.

Two things 'adapting' together because their fitness function indirectly influences each other rather than direct competition because of conflicting fitness functions...

So why do you have a problem with the co-evolution of genital organs distinguished by gender?

You've got it the wrong way round. Evolution does not enable adaptation.

Does this mean that you are retracting this statement:

This is because it [evolution] is NOT a method of design, but rather a method of enabling adaptation.

Adaptation might enable evolution ...

I have posed this question many times to those who claim adaptation does not imply evolution. No one has successfully answered it. Perhaps you can be the first.

Describe how adaptation occurs, with particular attention to the role of genetic inheritance.

See if you can do so without invoking the process of evolution.

Jesus had himself tortured and executed for a symbolic sin by a non-existent individual.

- Richard Dawkins,The root of all evil? (broadcast on Channel 4, 16 January 2006)

Dawkins forgets that a symbol is a symbol of something real. Understanding the Genesis story as symbolic does not mean rejecting the reality of what it symbolizes. Sin is real. Jesus did not die to destroy an illusion.

‘The interpretation of evolution is in a state of upheaval:...In the author’s opinion, no real theory of evolution can be formulated at present.’

Evolution Without Selection: FORM AND FUNCTION BY AUTOEVOLUTION by A. LIMA-DE-FARIA
http://creationontheweb.com/content/view/1160/105/

Sounds like an interesting book. But note that it is the interpretation of evolution that is in upheaval, not the question of whether or not evolution happens. The theory of how evolution happens may indeed be in for a significant shake-up. But I expect when you read the book (You are planning to read the book, aren't you--not just quote-mine from a publisher's blurb) you will find no questioning of the fact that evolution happens. Just a lot of questioning about how we can explain the process.

As for the detective coming on the murder scene, it is ironic that the detective in this case views the circumstantial evidence and refuses to consider any alternatives other than ideas the detective already has.

Well, that is not what the scientific detective does. You might review the basics of scientific method relative to formation of hypotheses, prediction, testing and falsification.