|
Source: (consider it)
|
Thread: The Death of Darwinism
|
ken
Ship's Roundhead
# 2460
|
Posted
quote:
Originally posted by Faithful Sheepdog:
I have been doing some research on other scientific models of evolution that have been proposed apart from neo-Darwinism. These seem to be based on saltation, i.e. explicit leaps, rather than a pure gradualism as per Dawkins and classical Darwinism, or a modified gradualism as per Gould and punctuated equilibrium. I will try to post more on this in due course.
Punk-eek is certainly part of the neo-Darwinist synthesis. A variety of it, not an alternatve to it.
"Saltationism" has meant a number of different things over the years,. In its early 20th-century form, it was used by people who misunderstood Mendel to argue against Darwin. But there is a sort of neo-saltationism around (not that they woudl use the word) which stresses the possible importance of like large-scale mutations such as chromosomal inversions.
It is at least possible that these have sometimes created barriers to breeding between populations living in the same location - one mechanism for the elusive sympatric speciation (Yes, it can happen. No, we have no evidence that it ever does much)
The main opposition to the adaptionist view of evolution has come from proponents of neutral evolution - look for references to Kimura.
But this is really still within the Synthesis, or at most an addition to it, not a disproof of it.
[ 22. June 2004, 18:31: Message edited by: ken ]
--------------------
Ken
L’amor che move il sole e l’altre stelle.
Posts: 39579 | From: London | Registered: Mar 2002
| IP: Logged
|
|
ken
Ship's Roundhead
# 2460
|
Posted
quote:
Originally posted by Faithful Sheepdog:
quote:
It (i.e. the modern theory of the mechanism of evolution) recognizes several mechanisms of evolution in addition to natural selection. One of these, random genetic drift, may be as important as natural selection.
Is random genetic drift a scientific speculation being intensively researched at the moment? Do we not know for certain?
We know for certain that it happens. What we don't know is how important it has been in causing speciation and specialistation.
These things are "natural laws" of the sort that physics investigates. They are descriptions of contingent history.
NB in the case of humans and chimps, genetic drift alone - so-called "neutral evolution" could account for ALL the observed genetic diffrences between us. In fact more than are observed, our genomes are probably more alike than would be expected in the absence of selection.
That doesn't mean that there has been no selection - but it is, to say the least, impraobable that a new mutation will proceed to fixation (i.e. become universal in the population) more quickly in the absence of selection than it woudl hav if there had been selection. So the expected rate of accquisistion of new mutations from neutral drift is a good lower limit on how fast we could expect change to occur under selection.
quote:
The Talk Origins site mentioned that, in both humans and chimpanzees, the same gene has been switched off, so that we cannot manufacture vitamin C ourselves. We need to acquire it in our diet, as do chimps. That could be evidence for common descent, or it could be that both changes occurred independently.
That's just one example of millions of known mutations. You can't tell descent from one or two - but when many seem to vary in a nested way, it's a Big Clue.
There are at the very least hundreds of characters shared by humans and chimps and no other primates. And many more shared by all the apes and no other primates. And others shared by apes, old world monkeys, and no others. And so on.
quote:
One thing I would note is that there is an important difference between increasing the information and changing the information. The genetic information which makes protein B in the improved model does not contain more information than that which makes protein A. It is simply a different but grammatically acceptable variation in the existing gene. A 1.4MB floppy disk can contain all sorts of different information, but not more than 1.4MB.
There are many examples of gene duplication, where one gene has become two, and they have then diverged in function.
For example, human globin genes, the homeobox genes, the ribsomal RNA genes, some spider-web genes, literally hundreds of metabolic genes.
In plants it is quite common for whole chromosomes, or even the whole genome, to duplicate. After a while many of the duplicate genes are randomly knocked out (because missing one copy of multiple genes coding for the same product may not have a deleterious effect), others accquire new function.
--------------------
Ken
L’amor che move il sole e l’altre stelle.
Posts: 39579 | From: London | Registered: Mar 2002
| IP: Logged
|
|
Rex Monday
None but a blockhead
# 2569
|
Posted
quote:
Originally posted by Faithful Sheepdog:
quote:
Rex Monday said:
An interesting aside: NASA is using evolutionary principles to design spacecraft antenna. Seems nobody told these engineers that such things are impossible.
This was an interesting link, but unfortunately it did not give me enough technical information to assess exactly what software algorithms and design techniques are being used.
Fortunately, there is much information available about what an engineer means if he says he's using evolutionary (or genetic - same thing) algorithms.
http://www.geatbx.com/docu/algindex.html may be a good start for you.
quote:
As a user of specialist technical software myself, I think the quoted phrase:
quote:
"The software also may invent designs that no human designer would ever think of,"
needs to be taken with a large pinch of salesman’s salt. I was certainly smiling at that point. Artificial intelligence on this scale is a long way off, if ever.
You precisely miss the point. No intelligence, artificial or otherwise, is required. That's why this approach yields results where some of the other approaches, such as knowledge systems, have failed.
quote:
What the software is probably doing is extensive optimisation of some trial designs following miscellaneous rules and deterministic theories for the subsequent technical performance of a given design.
No it's not. That's not what they said: they said randomness followed by evaluation for fitness. If you don't believe them, then that's up to you -- although arguments from incredulity are wearing so thin they count as nanotech by now -- but perhaps you could consider what it means if it works the way the NASA engineers say it works?
quote:
In my opinion this software is inventing nothing. All technical information, design algorithms and theoretical understandings are programmed in from the start.
Well, I'll take your opinion under advisement, but it would have more weight if you had some facts behind it. The engineers who produced the system are on record as saying something completely different, and their jobs are on the line. They aren't doing this for fun, and presuming that they wish to be taken seriously by their peers then they aren't deliberately (or accidentally) misrepresenting what they're up to.
No 'design algorithms' were programmed in aside from the 'take a device, randomly modify it, assess it, select it if it is better than the rest, repeat' That's how this stuff works, after all, and that's how natural selection works.
And nothing in the system needs 'knowledge' of *how* the finished article works - no 'intelligence' required - just how to change, test and compare.
I don't know if you've ever done any antenna design (oddly enough, I have), but doing it from first principles is bloody hard. It's much easier to take an existing design, modify it and run it through a simulator.
When you've also got to take other factors into account, like weight, size, materials and so on, mechanising the evolutionary approach seems a very sane way to approach the problem.
Tell you what, since you're disinclined to take my or NASA's word for it, why don't you have a look at some of the software freely available on the web that does genetic algorithmic antenna designs? You can examine it and decide for yourself how it works.
http://www.qsl.net/wb6tpu/swindex.html has a link to instygo2.exe, the Yagi Genetic Optimiser 2, of which the designer says: "YGO2 (YAGI GENETIC OPTIMIZER II) designs optimized Yagi antennas using one of the most recent and powerful techniques in applied electro- magnetics: a genetic algorithm. A genetic algorithm mimics natural selection ("survival of the fittest") by creating generations of Yagi designs in which the best designs ("individuals") pass their design parameters ("genes") onto the next generation in such a way that better and better antennas are the result. GA optimizers can produce remarkable results not achievable with any other design technique. Genetic algorithms are superior to deterministic optimzers, and they are easily implemented on a PC."
Unambiguous, eh? It's a bit technical, but "as a user of specialist technical software myself" I don't see you having a problem. I look forward to your analysis as to whether the software is doing what the designer says it's doing.
quote:
What the software can do is to consider millions of possible options in a short space of time. It is not dissimilar to those computers that can play chess and beat a grand master. The computer has to be told the rules beforehand. We can then exploit the massive calculation power that it offers – and beat the grand master with a computer. But who invented chess in the first place?
It's not that dissimilar, certainly. And using these simple rules - nobody would claim a chess computer had true intelligence - the machine can 'outthink' humans. So we have evidence that a mechanistic approach to a complex problem can produce better results than humans: thus, it should be no surprise when the mechanistic process of evolution produces better results than we can achieve.
Who invented the basic rules of chess? I don't know, and neither do you. Where did the basic rules of physics come from? I don't know, and neither do you.
Sorry, what was your point again?
quote:
For a similar computational approach I can recommend the section in The Blind Watchmaker where Richard Dawkins describes his various computer simulations, using the phrase Methinks It Is Like A Weasel, and then his fascinating geometrical shapes called biomorphs. The intelligent application of small incremental steps can certainly construct very complex shapes and systems.
However, Dawkins seems unaware that he is modelling a situation with substantial inbuilt constraints on where it may end up. In effect he assigns a creative intelligence to the power of natural selection – it knows where it is going. I do not consider that his computer models reflect the natural environment at all, which is random and uncontrolled.
I'm looking through my window outside at the natural environment, where a large weather system is sweeping across London. Random? Uncontrolled? The environment is heavily controlled by physical laws, which also produce notable unrandomness at all scales from atom to galaxy.
For a living being, the rules are complex and ruthless: if you can't extract energy from the environment, can't defend yourself from weather, predators or natural hazards, or can't successfully procreate, you're stuffed. Not much random there.
That's exactly what Dawkins was demonstrating: randomness filtered through rules produces change that can be constructive. No sense of direction required, unless you consider 'survival' as a sense of direction. The randomness doesn't apply equally to all parts of the system. Just because there's a random element at one point doesn't mean it's equally important everywhere.
(The trouble with Dawkin's analogy is the same trouble as all analogies of this nature - at some stage, someone will say "Naw, look, there's a human involved in constructing the analogy - therefore there must be an intelligence involved in real thing". Hopeless logic, of course, belying a thundering incomprehension about analogies, but it's amazing how often it happens.)
quote:
Neil
R
--------------------
I am largely against organised religion, which is why I am so fond of the C of E.
Posts: 514 | From: Gin Lane | Registered: Mar 2002
| IP: Logged
|
|
Alan Cresswell
Mad Scientist 先生
# 31
|
Posted
quote:
Originally posted by Faithful Sheepdog:
A 1.4MB floppy disk can contain all sorts of different information, but not more than 1.4MB.
If you make a back-up of your 1.4MB disk then you have, I would say, no more information. If, however, on one of those copies you change a file then you now have more information than you did before.
quote:
Artificial intelligence on this scale is a long way off, if ever ... All technical information, design algorithms and theoretical understandings are programmed in from the start.
This is a technique known as a genetic algorithm, so called because of the parallel between it's method of operation and genetic evolution. It is inaccurate to call it artifical intelligence, as it is distinctly anti-intelligent. It is setting a problem (say "what is the optimum dish size and shape to communicate with a probe?") then providing a mimimum of intelligent input (ie: the ideas and experience of communications engineers).
It is a pure brute force computational method that includes some way of coding the antenna shape, a means of modelling it's functionality and a selection filter that is simply "does this configuration perform better than the others?". Then the best models are "bred" and random mutations introduced and the cycle continued.
quote:
In my full size engineering model of The Old Man of Hoy, I was not considering the soil ecosystem, nor whatever flora and fauna may be present. There is certainly information there, but very little pertaining to a loadbearing engineering structure. I am focussing on information relevant to structural strength, such as overall size, cross-sectional shape, rock stratum depth, precise rock type, the location of geological joints, etc. All these features could be modelled, even in concrete.
These factors contain a considerable amount of information. A concrete facsimile may have the same shape, but it will be a considerably different structure with much less informational content. The structure of something like the Old Man of Hoy would depend to a large extent on very small components just like a tower block ... the slight differences in minerals that make some rocks more resistant to erosion than others for example. Even the soils and plants you ignored have helped govern the present shape of the tower in the way they affect erosion.
--------------------
Don't cling to a mistake just because you spent a lot of time making it.
Posts: 32413 | From: East Kilbride (Scotland) or 福島 | Registered: May 2001
| IP: Logged
|
|
Callan
Shipmate
# 525
|
Posted
Originally posted by Serafina in Purgatory:
quote:
It is an entirely plausible and respectable scientific viewpoint to believe that the method of evolution which is purported to have happened from primordial sludge to a mitochondrion is extremely unlikely and not a very logical theory. There are many reasons for this and I'm sure they've been discussed in the dead horses thread - I'll reiterate them there if anyone isn't familiar with them. My belief in ID doesn't come from any theological objection to the origins of life as popularly understood (or from an 'objection to Darwinism'), it comes from an understanding of molecular biology.
Karl Popper's account of scientific knowledge suggests that science progresses through a process of conjecture and refutation. For a belief to be scientific it needs to be empirically falsifiable. (There is a good argument that falsification is a necessary but not sufficient condition for science, but that is not germane to my point.)
Can you suggest ways in which ID could be empirically falsifiable? What experiments could one do? What research programme might find data to refute it? Given that the existence of a deity is a metaphysical hypothesis, is not the basis for ID conveniently placed outside the realm of refutation? And if so, how can it claim to be scientific.
--------------------
How easy it would be to live in England, if only one did not love her. - G.K. Chesterton
Posts: 9757 | From: Citizen of the World | Registered: Jun 2001
| IP: Logged
|
|
Alan Cresswell
Mad Scientist 先生
# 31
|
Posted
quote:
Originally posted by LatePaul (on the same Purgatory thread):
As a Christian I'd've seen Romans 1 and Psalm whatever-it-is-I'll-look-it-up-later* as appeals to ID.
<snip>
My question would be though - if those passages aren't an appeal to ID what are they about?
---
*you know the one about the heavens preaching
The passages are, I assume, Romans 1:20 ("For since the creation of the world His invisible attributes, His eternal power and divine nature, have been clearly seen, being understood through what has been made, so that they are without excuse.") and Psalm 19:1 ("The heavens are telling of the glory of God; And their expanse is declaring the work of His hands.") and surrounding verses.
First, I think it should be understood that these are words written by and for people who already believe in a Creator God. They are an appeal to people who know God exists - "look at the world around you, the heavens above, and from them something of the nature of God is plain". They are not appeals to people who don't believe in God - "look at the world around you, the heavens above, and from them the existance of God is plain".
Second, both are linked in very strongly with the theme of morality and Gods law. Romans is set within a section declaring that Gods wrath against people isn't against the ignorant, but against those who have sufficient evidence to recognise and respond positively to God yet have chosen not to do so. Psalm 19 uses the wisdom, power and perfection shown in Creation to point out the same characteristics of the Law of God.
Neither would have any power to speak to someone who looking at the universe sees nothing but apparent order brought about by random events and the immutable laws of physics. Neither passage, nor any other I can think of, would support the idea that one can prove the existance of God from the existance and nature of the physical universe. There are plenty of passages of Scripture which would support the idea that those with faith in God can look at the physical universe and learn something true about God from that (though, by no means all there is to know about God which is why we need his self-revelation supremely by Incarnation).
--------------------
Don't cling to a mistake just because you spent a lot of time making it.
Posts: 32413 | From: East Kilbride (Scotland) or 福島 | Registered: May 2001
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
quote:
Rex Monday said:
You precisely miss the point. No intelligence, artificial or otherwise, is required. That's why this approach yields results where some of the other approaches, such as knowledge systems, have failed.
I have no doubt that the engineers responsible for the implementation of an “evolutionary” algorithm in their design software are very proud of it. I’m certainly not accusing them of misrepresenting anything. All I’m saying is that enthusiastic self-reports from successful engineering project teams need careful examination, especially if their language uses some unusual engineering terminology borrowed from other scientific disciplines.
If you have experience in antenna design, then you are one step ahead of me. Words like monopole, dipole and multipole have emerged from my memory.
I presume that the “clan” of all possible antenna designs comes in “families” of different patterns (i.e. monopole or dipole or whatever). Within a given pattern (a “family”), a particular antenna design can be characterised by a string of N parameters describing the size, shape, material, functional limits etc. In this respect an antenna design is no different to a design for a girder or a pump, or any other engineering system whose technical content can be reduced to a string of N parameters applicable to a particular pattern.
Here is a description of evolutionary algorithm from the link that you provided, with my emphasis:
quote:
Evolutionary algorithms are stochastic search methods that mimic the metaphor of natural biological evolution. Evolutionary algorithms operate on a population of potential solutions applying the principle of survival of the fittest to produce better and better approximations to a solution. At each generation, a new set of approximations is created by the process of selecting individuals according to their level of fitness in the problem domain and breeding them together using operators borrowed from natural genetics. This process leads to the evolution of populations of individuals that are better suited to their environment than the individuals that they were created from, just as in natural adaptation.
Evolutionary algorithms model natural processes, such as selection, recombination, mutation, migration, locality and neighbourhood. Figure shows the structure of a simple evolutionary algorithm. Evolutionary algorithms work on populations of individuals instead of single solutions. In this way the search is performed in a parallel manner.
So the first point to emphasis is that there is a metaphor being used here. Just because the software says “evolutionary” algorithm does not mean that it necessarily reflects accurately the processes that have actually taken place in biological history. Clearly, however, it has taken its inspiration from a Darwinian model.
The second point is that each design is presumably characterised by a string of N parameters, and only N parameters. N will vary depending on which “family” the proposed design comes from. The software may even be able to consider different design families simultaneously, but that doesn’t affect my overall point.
The start parameters of the first generation are presumably generated randomly by the software rather than being fed to it in the form of a trial design. Subsequent generations can be generated by the software from the existing population in a process involving selection, recombination and mutation. This mimics the effects of breeding.
Each generation is tested against the programmed technical criteria for “fitness”, and the whole process then repeats. In this way the design performance of huge numbers of trial designs can be assessed in an extremely efficient manner. No argument there at all.
But let’s look at the inbuilt constraints on this process.
Firstly this software only applies to an antenna system described by the N parameters needed. The initial parameters for the first generation may indeed be generated at random from within the software, but it is likely that the software has in-built limits for the key parameters involved, based on practicality if nothing else. An antenna that is too large or too small to be built is of no use to us, so the software rules it our from the start.
The technical criteria against which “fitness” is assessed at each iteration are programmed in from the start. In principle the engineers tell the computer what is fit and what is not, through their design theories. This is essential information that must be programmed in, otherwise the computer has no basis on which to say one design is any better than any other. The arithmetical donkey work alone is done by the computer. The engineer is still the master at the level of ideas.
The process of selection, recombination and mutation will follow a programmed route. Again, the software is programmed to allow for a certain range of possibilities which the engineers are calling “random”. Note that the word “random” is being used rather loosely here, much like so-called “random” numbers on my calculator are always between 0 and 1. The software can only generate the next generation within its programmed constraints, and there is no possibility of generating a design outside those limits.
Millions of iterations later we still have an antenna design, or a population of antenna designs, but now they are highly optimised and very useful. However, at no point can the software evolve anything other than an antenna design, which is what it started with. Whatever numbers it throws out will always be an antenna design. Even if it ran for all of eternity, it will not morph into producing parameters for, say, a pump design.
A considerable amount of intelligence was needed on the part of the mathematicians, engineers and programmers to produce this software using “evolutionary” algorithms. I’ve not had personal experience with such algorithms, but I am happy to accept for some technical problems they may indeed represent a vast improvement on earlier approaches.
However, although “evolutionary” algorithms have taken their inspiration from biology, I do not think that they are a model of Darwinian evolution, unless you are prepared to accept that natural selection can operate in a mathematical fashion according to intelligently determined numerical rules with a clear sense of purpose behind them.
quote:
Rex Monday said:
It's not that dissimilar, certainly. And using these simple rules - nobody would claim a chess computer had true intelligence - the machine can 'outthink' humans. So we have evidence that a mechanistic approach to a complex problem can produce better results than humans: thus, it should be no surprise when the mechanistic process of evolution produces better results than we can achieve.
Who invented the basic rules of chess? I don't know, and neither do you. Where did the basic rules of physics come from? I don't know, and neither do you.
Sorry, what was your point again?.
At school I was still using a slide-rule and log tables, so I am not denying for a moment that computers have vastly more computational power than humans. In that sense computers can produce better results – but only in accordance with the constraints that we give them. In any case, the first rule of computers is “garbage in equals garbage out”.
My point about chess was that human beings invented chess in the past. The rules are the human-derived constraints that govern the movement of pieces and the flow of the game. That took intelligence, something present day computers certainly do not have. Without us there would be neither chess nor computers.
Neil
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
quote:
Callan said:
Can you suggest ways in which ID could be empirically falsifiable? What experiments could one do? What research programme might find data to refute it? Given that the existence of a deity is a metaphysical hypothesis, is not the basis for ID conveniently placed outside the realm of refutation? And if so, how can it claim to be scientific.
You may want to ask your first question to Darwinists, too. ![[Smile]](smile.gif)
I think that a short answer to your question on ID is the recognition that in certain circumstances a methodological naturalism completely fails to provide the right answer to certain questions, especially ones where we know what the correct answer is.
Here are some rock formations. It is immediately obvious that the most comprehensive understanding of geology, erosion, wind, weather and the physical sciences generally will never correctly explain such formations. We know for a fact that they were deliberately formed by intelligent beings – humans – in honour of other humans.
The same category mistake would be made by those who open a Bible, only to insist that the discussion be limited to the chemistry of printers’ ink and the science of paper-making and book-binding. What they say about ink and paper may well be very true, but it is incapable of giving a full explanation of the Bible, namely that the ancient authors wrote in order to communicate a message of faith to others, and to record it for posterity.
ID is far more than throwing one’s hands up in the air and saying “God did it”. Here is a very short note from William Dembski on some relevant research issues suggested by an ID paradigm.
Neil
[fixed blown code]
[ 23. June 2004, 12:34: Message edited by: Faithful Sheepdog ]
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
Rex Monday
None but a blockhead
# 2569
|
Posted
quote:
Originally posted by Faithful Sheepdog:
quote:
Rex Monday said:
You precisely miss the point. No intelligence, artificial or otherwise, is required. That's why this approach yields results where some of the other approaches, such as knowledge systems, have failed.
<huge restatement of basic evolutionary algorithm ideas snipped>
The technical criteria against which “fitness” is assessed at each iteration are programmed in from the start. In principle the engineers tell the computer what is fit and what is not, through their design theories. This is essential information that must be programmed in, otherwise the computer has no basis on which to say one design is any better than any other. The arithmetical donkey work alone is done by the computer. The engineer is still the master at the level of ideas.
As I said in my reply to your point about Dawkins weasel demonstration (a reply you seem to have missed - don't worry, it's coming up again later), this 'essential information' does not need to be intelligently provided. In the natural world, the 'information' is the environment in which a particular species lives. In the antenna design software, the test is 'how well does the antenna perform' (and, presumably in the NASA example, 'how light/small/spaceworthy is it?' - fitness can be a very complex thing). That one is set by engineers and another by nature does not mean that nature is the product of engineering. That is a basic logical fallacy.
What the evolutionary algorithms show is that the basic idea behind evolution is sound.
quote:
The process of selection, recombination and mutation will follow a programmed route. Again, the software is programmed to allow for a certain range of possibilities which the engineers are calling “random”. Note that the word “random” is being used rather loosely here, much like so-called “random” numbers on my calculator are always between 0 and 1. The software can only generate the next generation within its programmed constraints, and there is no possibility of generating a design outside those limits.
There's also no possibility of natural selection on Earth producing a design outside its limits. Will we ever see a methane-breathing giant balloon animal capable of living in the upper atmosphere of Jupiter evolve on Earth? No, any more than Jovian natural selection will produce a bunny rabbit.
So I don't understand what your point is here: you appear to be saying that evolution can't operate outside its own constraints. No, it can't, and that'e entirely consistent with evolutionary theory.
Were animals to appear that *did* break evolutionary constraints, that would be an indication that evolutionary thought is wrong. IDers claim that this has happened: unfortunately, they cannot demonstrate it. Evolutionary thought has proved more than powerful enough to answer all their objections (and design antennae on its tea break).
quote:
Millions of iterations later we still have an antenna design, or a population of antenna designs, but now they are highly optimised and very useful. However, at no point can the software evolve anything other than an antenna design, which is what it started with. Whatever numbers it throws out will always be an antenna design. Even if it ran for all of eternity, it will not morph into producing parameters for, say, a pump design.
You are correct: it won't (or at least it is very improbable). Again, you're stating basic evolutionary theory. The system will not select for things which are not conducive to fitness. However, if you asked it to select for antennae that were also pumps, then it might - and probably would, given enough time. That's setting the environment to select for a different class of fitness.
It's very simple. The antenna design software uses evolutionary ideas to select for a better antenna. Natural selection uses evolutionary ideas to select for species better able to survive in their environments. Is it surprising that the output of the latter is more diverse than the former?
quote:
<snip>
However, although “evolutionary” algorithms have taken their inspiration from biology, I do not think that they are a model of Darwinian evolution, unless you are prepared to accept that natural selection can operate in a mathematical fashion according to intelligently determined numerical rules with a clear sense of purpose behind them.
The people who build the software consciously model that software on Darwinian evolution, so your claim that it is not so modelled is prima face false.
By your logic, as soon as you model a thing you're proving that the thing is of intelligent origin. That is prima face absurd.
Of course they use their intelligence and a sense of purpose in building a model, that's what building models (and intelligence) is all about. How else would you do it? But it is a model - and the thing that is modelled does not automatically acquire the attributes of its modellers. (It is hard indeed to conceive of the world where this would be the case: you appear to be drifting into sympathetic magic and voodoo, rather than engineering.)
Unfortunately, you appear to have missed what I said about Dawkins and the weasel, when I replied to this point the first time you brought up.
As you missed it, here it is again.
quote:
quote:
For a similar computational approach I can recommend the section in The Blind Watchmaker where Richard Dawkins describes his various computer simulations, using the phrase Methinks It Is Like A Weasel, and then his fascinating geometrical shapes called biomorphs. The intelligent application of small incremental steps can certainly construct very complex shapes and systems.
However, Dawkins seems unaware that he is modelling a situation with substantial inbuilt constraints on where it may end up. In effect he assigns a creative intelligence to the power of natural selection – it knows where it is going. I do not consider that his computer models reflect the natural environment at all, which is random and uncontrolled.
I'm looking through my window outside at the natural environment, where a large weather system is sweeping across London. Random? Uncontrolled? The environment is heavily controlled by physical laws, which also produce notable unrandomness at all scales from atom to galaxy.
For a living being, the rules are complex and ruthless: if you can't extract energy from the environment, can't defend yourself from weather, predators or natural hazards, or can't successfully procreate, you're stuffed. Not much random there.
That's exactly what Dawkins was demonstrating: randomness filtered through rules produces change that can be constructive. No sense of direction required, unless you consider 'survival' as a sense of direction. The randomness doesn't apply equally to all parts of the system. Just because there's a random element at one point doesn't mean it's equally important everywhere.
In other words, rules -- even complex rules -- don't need to be created or enforced by intelligence.
quote:
quote:
Rex Monday said:
It's not that dissimilar, certainly. And using these simple rules - nobody would claim a chess computer had true intelligence - the machine can 'outthink' humans. So we have evidence that a mechanistic approach to a complex problem can produce better results than humans: thus, it should be no surprise when the mechanistic process of evolution produces better results than we can achieve.
Who invented the basic rules of chess? I don't know, and neither do you. Where did the basic rules of physics come from? I don't know, and neither do you.
Sorry, what was your point again?.
At school I was still using a slide-rule and log tables, so I am not denying for a moment that computers have vastly more computational power than humans. In that sense computers can produce better results – but only in accordance with the constraints that we give them. In any case, the first rule of computers is “garbage in equals garbage out”.
As the evolutionary algorithm engineers have shown a computer can operate within those constraints and be genuinely creative.
quote:
My point about chess was that human
beings invented chess in the past. The rules are the human-derived constraints that govern the movement of pieces and the flow of the game. That took intelligence, something present day computers certainly do not have. Without us there would be neither chess nor computers.
So your point is: humans came up with chess and now they're building chess computers.
You got me there.
What has that got to do with evolution?
quote:
Neil
R
--------------------
I am largely against organised religion, which is why I am so fond of the C of E.
Posts: 514 | From: Gin Lane | Registered: Mar 2002
| IP: Logged
|
|
Serafina
Apprentice
# 7185
|
Posted
quote:
Originally posted by Callan:
Originally posted by Serafina in Purgatory:
quote:
It is an entirely plausible and respectable scientific viewpoint to believe that the method of evolution which is purported to have happened from primordial sludge to a mitochondrion is extremely unlikely and not a very logical theory. There are many reasons for this and I'm sure they've been discussed in the dead horses thread - I'll reiterate them there if anyone isn't familiar with them. My belief in ID doesn't come from any theological objection to the origins of life as popularly understood (or from an 'objection to Darwinism'), it comes from an understanding of molecular biology.
Karl Popper's account of scientific knowledge suggests that science progresses through a process of conjecture and refutation. For a belief to be scientific it needs to be empirically falsifiable. (There is a good argument that falsification is a necessary but not sufficient condition for science, but that is not germane to my point.)
Can you suggest ways in which ID could be empirically falsifiable? What experiments could one do? What research programme might find data to refute it? Given that the existence of a deity is a metaphysical hypothesis, is not the basis for ID conveniently placed outside the realm of refutation? And if so, how can it claim to be scientific.
Well, whether Karl Popper's theory applies to unrepeatable events is a whole other thread.
But, since you asked, then, yes, if you really wanted to and threw a lot of money at the project, there are experiments which could seek to falsify the idea of irreducible complexity. An example is the attempt to synthesise amino acids by running electrical currents through a mixture of gases of the composition thought to be present at around the time when life is said to have first formed.
A more likely way that it could be falsified would be if a living organism less complicated than a mitochondrion were found. But since that wouldn't be a designed and controlled experiment I guess you wouldn't consider that science.
I've answered from your perspective, so now will you tell me what's wrong with irreducible complexity using my definition of scienctific method? My definition being to look at all the available evidence and decide logically what the most likely explanation is.
Posts: 24 | From: UK | Registered: May 2004
| IP: Logged
|
|
Toby
Shipmate
# 3522
|
Posted
quote:
Originally posted by Serafina:
I've answered from your perspective, so now will you tell me what's wrong with irreducible complexity using my definition of scienctific method? My definition being to look at all the available evidence and decide logically what the most likely explanation is.
'Irreducible complexity' is not so much a 'most likely explanation' for observed phenomena - it is a criticism of another theory. And as a criticism, it can be in turn criticised through observing ways in which 'irreducibly complex' things might, in fact, be reducible.
For example, the biochemical structures labelled 'irreducibly complex' by some people could have been formed through 'scaffolding' - structures which no longer exist but once aided in putting together the various parts of the protein.
Lets face it, intelligent design is not really a fully grown theory of the development of the diversity of life - it is an ideologically motivated, desparate attempt to criticise another established theory that works very well.
Rejecting the modern evolutionary synthesis, which is a comprehensive, lucid, elegant and adaptable theory which explains a huge amount about genetics, biogeography, diversity and natural history on the basis of some spurious and (frankly) subjective idea like 'irreducible complexity' is ridiculous. Intelligent design theory is a speculative idea that does not have the tiniest fraction of the explanatory power of the modern synthesis of evolutionary theory.
[ 23. June 2004, 20:34: Message edited by: Toby ]
--------------------
'Civilization is only savagery silver-gilt'
Allan Quartermain
Posts: 99 | From: Aotearoa/New Zealand | Registered: Nov 2002
| IP: Logged
|
|
Serafina
Apprentice
# 7185
|
Posted
quote:
Originally posted by Toby:
[Irreducible complexity] can be in turn criticised through observing ways in which 'irreducibly complex' things might, in fact, be reducible.
For example, the biochemical structures labelled 'irreducibly complex' by some people could have been formed through 'scaffolding' - structures which no longer exist but once aided in putting together the various parts of the protein.
Well it's still something of a mystery how protein tertiary and quaternary structure is determined, so I'll reserve judgement on that point. But I've yet to hear a decent explanation for how half a Krebs cycle would have enough of a selective advantage to stick around until the other half evolved, or two thirds of glycolysis.
quote:
Lets face it, intelligent design is not really a fully grown theory of the development of the diversity of life - it is an ideologically motivated, desparate attempt to criticise another established theory that works very well.
No, it isn't a theory which explains everything. It just suggests that (assuming God was involved at some point) God took an active role in the origins of life, rather than the 'clockwork toy' model where He sets everything up for the big bang then sits back and passively observes.
--------------------
Awaiting inspiration
Posts: 24 | From: UK | Registered: May 2004
| IP: Logged
|
|
Alan Cresswell
Mad Scientist 先生
# 31
|
Posted
quote:
Originally posted by Serafina:
No, it isn't a theory which explains everything. It just suggests that (assuming God was involved at some point) God took an active role in the origins of life, rather than the 'clockwork toy' model where He sets everything up for the big bang then sits back and passively observes.
Well, that's OK then ... because I believe in a God who takes an active role in the origin and evolution of life and doesn't sit back and passively observe his clockwork toy (that belief would be deism not theism). The thing is, I see no reason to assume that a scientific understanding of the universe would leave explicit evidence of his action - which is the assumption behind ID as I understand it.
--------------------
Don't cling to a mistake just because you spent a lot of time making it.
Posts: 32413 | From: East Kilbride (Scotland) or 福島 | Registered: May 2001
| IP: Logged
|
|
mr cheesy
Shipmate
# 3330
|
Posted
Re the issue of the evolutionary software...
I confess to knowing little about the mechanics of the programming, but AFAIUI, the algorithm randomly mutates then the result is tested to see how well it works (presumably against some set values).
That being the case (please correct my faulty understanding someone who knows), surely the only thing it shows is that given the right circumstances, mutation can cause evolutionary change for the better.
Now this gets to the crux of the matter. Dawkins says evolution occurs because he can imagine a circumstance where it might beneficial change leading to speciation. This program shows that happens, given certain conditions.
The debate, therefore, should therefore be whether these ideas and algorithms can accurately mimic the conditions in the natural environment.
I would argue that they do not. Comparing a simple computer program, which mutates fairly randomly against a given (and small) set of criteria is totally different to the natural environment, where organisms are under a miriad of selective pressures, which might be constantly changing. Creating something artificially is not the same as proving it happens in nature, especially when - as it appears in this case - known features are not directly being modelled.
C
--------------------
arse
Posts: 10697 | Registered: Sep 2002
| IP: Logged
|
|
Toby
Shipmate
# 3522
|
Posted
quote:
Originally posted by Serafina:
But I've yet to hear a decent explanation for how half a Krebs cycle would have enough of a selective advantage to stick around until the other half evolved, or two thirds of glycolysis.
The Krebs cycle is found only in mammals, if I recall (and it is possible that my recall on this point is imperfect - my limited and fairly unpleasant experience of doing a biochemistry paper was some years ago). In that case it evolved in a context of an already incredibly sophisticated biological system, and it may have (again, I am no expert) been a development of another biochemical pathway. The biochemical pathway that evolved into the Krebs cycle (assuming for the moment that there was one, and that it happened without divine intervention) may have had a different function, or else may have been a longer and more clumsy process that was refined (ie the 'scaffolding' I mentioned before)down to the relative elegance of the modern mammalian process.
And the example of the Krebs cycle is really nit-picking and misleading in a larger context. There are millions of things that evolution cannot explain because (a) The tools that biologists have to peer back into the past are imperfect and often misleading - like historians, biologists see through a glass darkly; and (b) There are not millions and millions of biologists working on every single biochemical and ecological problem, nor are scientists backed by infinite resources.
Again, I am no biochemist. What I do know with more familiarity is that on an ecological, population genetics level evolutionary theory is an extremely persuasive and powerful tool for both prediction and explanation. On a population level, evolution fits sooo well...
Of course, this does not really disprove what you are saying about God being involved at a biochemical level. I suppose that the intelligent design argument does acknowledge the larger population level evolutionary processes (I certainly hope so). But I cannot see how taking individual problems of biology that have not been comprehensively investigated and saying that these 'prove' that there was theistic intervention in the development of life can be a final conclusion, or a final rejection of evolutionary theory.
So much remains uninvestigated, unproven and unknown. Intelligent design is based on too many assumptions that rely on the certainty of ambiguous and unknowable evidence.
--------------------
'Civilization is only savagery silver-gilt'
Allan Quartermain
Posts: 99 | From: Aotearoa/New Zealand | Registered: Nov 2002
| IP: Logged
|
|
mr cheesy
Shipmate
# 3330
|
Posted
So what is the difference between believing in a God-of-the-gaps and evolution-of-the-gaps? We have reached a strange point in the argument if we are appealing to an intermediate biochemical pathway that we cannot even imagine.
![[Confused]](confused.gif)
C
--------------------
arse
Posts: 10697 | Registered: Sep 2002
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
quote:
Cheesy* said:
The debate, therefore, should therefore be whether these ideas and algorithms can accurately mimic the conditions in the natural environment.
Absolutely agreed. I have been arguing that these software algorithms are built on a metaphor derived from one biological hypothesis. It is a mistake to push the metaphor too far. It is also a mistake to say that, just because the software works well and produces optimised antenna designs efficiently, actual historical biological processes must have followed the same route.
As Ken pointed out earlier on the thread, natural selection can be modelled very simply. One genetic variation outbreeds other genetic variations, and becomes dominant in a population. In principle we simply count the relative number of grandchildren.
“Fitness” in this understanding is a numerical criterion, which can be modelled as a simple binary parameter for the parent generation: “yes” (better than average breeding success) versus “no” (not better than average breeding success). All the myriad environmental influences in the natural world boil down to this one parameter. All theoretical eggs are in the one basket. “Natural selection” alone is credited with a huge amount of creative power.
In NASA’s evolutionary algorithm, I think it highly unlikely that the state of optimisation of an antenna design can be represented in such a simple binary fashion as “yes” or “no”. The selection criteria are likely to be far more sophisticated than a single binary parameter.
In mathematical terms, the evolutionary algorithm has been formulated so that the process is continually driven forward in some predetermined way to a point of convergence. By design the software will not allow the process to oscillate about a mean (stability), or degenerate into chaos (divergence).
quote:
Rex Monday said:
That's exactly what Dawkins was demonstrating: randomness filtered through rules produces change that can be constructive. No sense of direction required, unless you consider 'survival' as a sense of direction. The randomness doesn't apply equally to all parts of the system. Just because there's a random element at one point doesn't mean it's equally important everywhere.
In Dawkins’ Methinks it is like a weasel simulation, he ascribes in advance binary fitness values to the various letters of the phrase. For the initial letter, the letter M is declared “fit”, and all other letters of the alphabet declared “unfit”. He decides the same for all other positions in the phrase. When the correct letter for the given position appears in his simulation, he selects it and then maintains it as he runs further simulations.
In other words, Dawkins’ simulation knows where it wants to go right at the start. By maintaining the correct letters as they appear, Dawkins also ensures that his simulation can only change consistently in one direction - it can never be stable or cyclical. Not surprisingly, his simulation converges very quickly to the desired goal of a meaningful phrase. It is most definitely not a model of an unintelligent and undirected process (“natural selection”).
The equivalent for NASA would be to begin their antenna design optimisation software knowing precisely what optimised design parameters they were aiming for. But if they knew that, why would they need to run the software in the first place?
Neil
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
Serafina
Apprentice
# 7185
|
Posted
quote:
Originally posted by Toby:
The Krebs cycle is found only in mammals, if I recall (and it is possible that my recall on this point is imperfect - my limited and fairly unpleasant experience of doing a biochemistry paper was some years ago). In that case it evolved in a context of an already incredibly sophisticated biological system, and it may have (again, I am no expert) been a development of another biochemical pathway. The biochemical pathway that evolved into the Krebs cycle (assuming for the moment that there was one, and that it happened without divine intervention) may have had a different function, or else may have been a longer and more clumsy process that was refined (ie the 'scaffolding' I mentioned before)down to the relative elegance of the modern mammalian process.
No, the Krebs cycle occurs in all eukaryotes (i.e. pretty much everything except bacteria). And it's hard to see how it could have evolved 'in the context of an already sophisticated system' because the sophisticated system probably couldn't have evolved without it; the Krebs cycle is not an optional extra refinement, it an 'absence = dead'.
quote:
And the example of the Krebs cycle is really nit-picking and misleading in a larger context.
Well, I only picked it as an example. I can give you hundreds of alternative examples if you prefer. The same argument of irreducible complexity applies to all of them
quote:
On a population level, evolution fits sooo well...
Yes, it does
quote:
this does not really disprove what you are saying about God being involved at a biochemical level.
No, it doesn't
quote:
But I cannot see how taking individual problems of biology that have not been comprehensively investigated (1) and saying that these 'prove' (2) that there was theistic intervention in the development of life can be a final conclusion, or a final rejection of evolutionary theory (3).
(1) I beg to differ. Do a google search on a metabolic pathway of your choice and see just how much research has been done.
(2) I don't know who you're arguing with here, but I've never said that ID has been proved. IMO it's the explanation which fits best with the available evidence. That's very different from claiming proof.
(3) Again, I've never said that.
quote:
Intelligent design is based on too many assumptions that rely on the certainty of ambiguous and unknowable evidence.
Oh, what a contrast to evolution!
--------------------
Awaiting inspiration
Posts: 24 | From: UK | Registered: May 2004
| IP: Logged
|
|
Rex Monday
None but a blockhead
# 2569
|
Posted
quote:
Originally posted by Faithful Sheepdog:
quote:
Rex Monday said:
That's exactly what Dawkins was demonstrating: randomness filtered through rules produces change that can be constructive. No sense of direction required, unless you consider 'survival' as a sense of direction. The randomness doesn't apply equally to all parts of the system. Just because there's a random element at one point doesn't mean it's equally important everywhere.
In Dawkins’ Methinks it is like a weasel simulation, he ascribes in advance binary fitness values to the various letters of the phrase. For the initial letter, the letter M is declared “fit”, and all other letters of the alphabet declared “unfit”. He decides the same for all other positions in the phrase. When the correct letter for the given position appears in his simulation, he selects it and then maintains it as he runs further simulations.
In other words, Dawkins’ simulation knows where it wants to go right at the start. By maintaining the correct letters as they appear, Dawkins also ensures that his simulation can only change consistently in one direction - it can never be stable or cyclical. Not surprisingly, his simulation converges very quickly to the desired goal of a meaningful phrase. It is most definitely not a model of an unintelligent and undirected process (“natural selection”).
It's a model of selection, not evolution as a whole, showing the difference between cumulative and single-step selection (an important point for the IDers). Read it in context! (something that seems to be an issue for many creationists).
quote:
The equivalent for NASA would be to begin their antenna design optimisation software knowing precisely what optimised design parameters they were aiming for. But if they knew that, why would they need to run the software in the first place?
The NASA software and the Dawkins' demo do different tasks, and cannot be directly compared. I think that's reasonably plain.
Do you agree with the rest of the points I made, then?
quote:
Neil
R
(edited for formatting)
[ 24. June 2004, 12:03: Message edited by: Rex Monday ]
--------------------
I am largely against organised religion, which is why I am so fond of the C of E.
Posts: 514 | From: Gin Lane | Registered: Mar 2002
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
quote:
Rex Monday said:
Do you agree with the rest of the points I made, then?
That would definitely be an unwise assumption. ![[Razz]](tongue.gif)
I chose to focus my reply on one part of your post linked to Cheesy’s comments. But I will comment on another part of it now:
quote:
The people who build the software consciously model that software on Darwinian evolution, so your claim that it is not so modelled is prima face false.
I don’t think I’ve made this claim at all. The developers of evolutionary algorithms undoubtedly took their inspirational ideas from Darwinism, but NASA probably had to modify them somewhat to be applicable to antenna design. See my comment above on NASA’s selection criteria.
quote:
By your logic, as soon as you model a thing you're proving that the thing is of intelligent origin. That is prima face absurd.
Again, you’re misreading me. The key element in any computer simulation is how authentically it reflects the true physical processes, whether in the behaviour of antennas or in the biological world. In their software NASA can define the behaviour of their antennas with a certain numerical precision. The environment which is responsible for natural selection cannot easily be so defined.
Serious question - has anyone produced a numerical model of the natural environment? It strikes me as an exceptionally difficult task due to the large number of random variables involved.
quote:
Of course they use their intelligence and a sense of purpose in building a model, that's what building models (and intelligence) is all about. How else would you do it? But it is a model - and the thing that is modelled does not automatically acquire the attributes of its modellers. (It is hard indeed to conceive of the world where this would be the case: you appear to be drifting into sympathetic magic and voodoo, rather than engineering.)
No, you misunderstand me completely. I’m asking for the biologists’ computer models to truly reflect what is being claimed for the theoretical side of Darwinian evolution. From what I know about natural selection, I also want the computer models to reflect observational reality. Finch beaks in the Galapagos have been observed to cycle, rather than to grow ever longer. The models need to incorporate this possibility.
Dawkins is more outspoken than most in flatly rejecting any metaphysics in his presentation of the theory. For him nature is unintelligent, undirected and described completely by the laws of science. His biomorphs are great fun, but he needs to build his computer models with much more care if he wants to claim the high scientific ground without an inadvertent concession to metaphysics at the same time.
Neil
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
Alan Cresswell
Mad Scientist 先生
# 31
|
Posted
quote:
Originally posted by Faithful Sheepdog:
In their software NASA can define the behaviour of their antennas with a certain numerical precision. The environment which is responsible for natural selection cannot easily be so defined.
True enough, the natural environment is a very complex situation that would take an awful lot of effort to even begin to approximate in a simulation (even limiting things to the subset of the environment that is the climate is a really substantial problem. In addition to it being very complex to merely describe, it is also dynamic and will change in response to changes in the organisms that live there. Such feedback would make such a model, even for a simplified situation, very difficult - but would account for such things as the cycling of beak lengths of finches as what may be the best solution for succesful reproduction now may be less favourable in a few hundred years.
Posts: 32413 | From: East Kilbride (Scotland) or 福島 | Registered: May 2001
| IP: Logged
|
|
HenryT
Canadian Anglican
# 3722
|
Posted
quote:
Originally posted by Callan:
Can you suggest ways in which ID could be empirically falsifiable? What experiments could one do? What research programme might find data to refute it? ...
And likewise the neo-Darwinian hypotheses.
I suspect that the issue is undecideable, because neither is subject to full-scale falsifiability. However, the subparts of the neo-Darwinian model seem to me to have been tested fairly hard and survived.
To continue Callan's point
quote:
Given that the existence of a deity is a metaphysical hypothesis, is not the basis for ID conveniently placed outside the realm of refutation? And if so, how can it claim to be scientific.
Specifically, then, the distinction, as I understand it, between a Darwinian view and ID is the claim of non-random "goal seeking" in a very subtle way.
This leads to the strong Anthropic Principle (I recommend a Google search here) which essentially is a meta-physical non-falsifiable claim that the Universe exists to bring forth life and intelligence.
My own personal belief is in just that, a careful adjustment of the initial conditions to produce a "friendly" universe. A summary of the macro-scale life-friendly design is quote:
the seemingly arbitrary and unrelated constants in physics have one strange thing in common--these are precisely the values you need if you want to have a universe capable of producing life.
But, the initial conditions having been laid, I don't see the need for God to directly intervene to produce the featherless biped with toenails. Which is not to say that I'm a Deist, with a non-personal and uninvolved God. Just one who appreciates subtlety.
--------------------
"Perhaps an invincible attachment to the dearest rights of man may, in these refined, enlightened days, be deemed old-fashioned" P. Henry, 1788
Posts: 7231 | From: Ottawa, Ontario, Canada | Registered: Dec 2002
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
On the website to which Henry Troup linked above, I came across a reference to Haldane’s Dilemma, named after the evolutionary geneticist JBS Haldane. I had never heard of it. This is how the website summarises it:
quote:
Haldane's Dilemma is glaringly plain. Take the population we discussed above. In ten million years, it could substitute 1,667 beneficial nucleotides. That is less than 50 millionths of one percent of the genome. (And that is *before* we make deductions. For example, Gould says species typically spend *at least* 90% of their time in stasis, where little or no evolution occurs. There are other deductions we'll discuss later, but together they reduce the figure far below 1,667.) Is that enough to explain the origin of upright posture, speech, language, and appreciation of music, to name just a few of our uniquely human capacities? Is 1,667 beneficial nucleotides enough to make a sapien out of a simian?
So, has anyone ever heard of Haldane’s Dilemma? Have any new solutions to the dilemma been proposed? Does the website present it correctly? Is it a “major scandal”?
Neil
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
Glenn Oldham
Shipmate
# 47
|
Posted
quote:
Originally posted by Faithful Sheepdog:
On the website to which Henry Troup linked above, I came across a reference to Haldane’s Dilemma,
See http://www.talkorigins.org/indexcc/CB/CB121.html for a swift response.
Note that in item 2 the list of 4 points are those that Remine fails to take into account. (The previous sentence might mislead one to believe that they are being listed as wrong assumptions that Remine has made.)
--------------------
This entire doctrine is worthless except as a subject of dispute. (G. C. Lichtenberg 1742-1799 Aphorism 60 in notebook J of The Waste Books)
Posts: 910 | From: London, England | Registered: May 2001
| IP: Logged
|
|
Glenn Oldham
Shipmate
# 47
|
Posted
quote:
Originally posted by Serafina:
But I've yet to hear a decent explanation for how half a Krebs cycle would have enough of a selective advantage to stick around until the other half evolved, or two thirds of glycolysis.
This was the subject of the following paper in the Journal Of Molecular Evolution
J Mol Evol. 1996 Sep;43(3):293-303.
The puzzle of the Krebs citric acid cycle: assembling the pieces of chemically feasible reactions, and opportunism in the design of metabolic pathways during evolution.
Melendez-Hevia E, Waddell TG, Cascante M.
The abstract of which reads:
quote:
The evolutionary origin of the Krebs citric acid cycle has been for a long time a model case in the understanding of the origin and evolution of metabolic pathways: How can the emergence of such a complex pathway be explained? A number of speculative studies have been carried out that have reached the conclusion that the Krebs cycle evolved from pathways for amino acid biosynthesis, but many important questions remain open: Why and how did the full pathway emerge from there? Are other alternative routes for the same purpose possible? Are they better or worse? Have they had any opportunity to be developed in cellular metabolism evolution? We have analyzed the Krebs cycle as a problem of chemical design to oxidize acetate yielding reduction equivalents to the respiratory chain to make ATP. Our analysis demonstrates that although there are several different chemical solutions to this problem, the design of this metabolic pathway as it occurs in living cells is the best chemical solution: It has the least possible number of steps and it also has the greatest ATP yielding. Study of the evolutionary possibilities of each one-taking the available material to build new pathways-demonstrates that the emergence of the Krebs cycle has been a typical case of opportunism in molecular evolution. Our analysis proves, therefore, that the role of opportunism in evolution has converted a problem of several possible chemical solutions into a single-solution problem, with the actual Krebs cycle demonstrated to be the best possible chemical design. Our results also allow us to derive the rules under which metabolic pathways emerged during the origin of life.
the same abstract on pub med
This article is referred to in Kenneth Miller's crtique of Behe in his book Finding Darwin's God
--------------------
This entire doctrine is worthless except as a subject of dispute. (G. C. Lichtenberg 1742-1799 Aphorism 60 in notebook J of The Waste Books)
Posts: 910 | From: London, England | Registered: May 2001
| IP: Logged
|
|
Toby
Shipmate
# 3522
|
Posted
quote:
Originally posted by Serafina:
No, the Krebs cycle occurs in all eukaryotes (i.e. pretty much everything except bacteria). And it's hard to see how it could have evolved 'in the context of an already sophisticated system' because the sophisticated system probably couldn't have evolved without it; the Krebs cycle is not an optional extra refinement, it an 'absence = dead'.
....
(3) Again, I've never said that.
Serfina, I am sorry that I represented what you are saying imperfectly; I have spent too many hours with people who would deny the epistemological foundations of that which I spent some long (sometimes unpleasant) hours studying. Thus I was prejudiced in my regard for what you were saying. Again, I apologise. My misconception about the Krebs cycle originated from the knowledge that 1080 poison, used in NZ against mammalian pests, affects only mammals and acts on the cycle. It must be specific to some mammalian aspect of the cycle rather than to the cycle itself as I assumed.
Saying 'pretty much everything except bacteria' (and, I would add, archaea) is misleading in that bacteria 'rule' the living world in terms of diversity, numbers and (possibly) biomass. Simplicity seems to work well in the natural world. But that is not really entirely relevant to what we are talking about.
Evolution as a self-sustaining process was mooted under a framework that sought validation from the big processes, the things that I know more about in ecology etc. From that point, scientists have focused more on smaller-scale things as the technologies for investigating biochemistry, molecular genetics etc. have improved. Evolution has been validated on a macro level, but there just is not enough technology or a sufficient resolution of investigation to explain every biochemical pathway - yet anyway. Glenn Oldham has pointed to one attempt at fitting the Krebs cycle within an evolutionary framework. I do not doubt that this research is highly speculative - I doubt that it is entirely persuasive.
My problem with notions of biochemical irreducible complexity is that they attack the area where we know the least about what really goes on, where things are most complex and impenetrable. While I have chosen a different eventual course of study other than biology, and thus am not so immersed in biology or passionate about the subject to be anything near an expert, my experience of studying it indicates that on a macro-level, where things can be easily modelled and understood, evolution fits well. So the cynic within me speculates that those like Behe who assail some aspects of evolutionary theory from a biochemical level find a wealth of material there precisely because we really don't know all that much about what goes on, and the incredible complexity of these systems defies easy interpretation or analysis.
--------------------
'Civilization is only savagery silver-gilt'
Allan Quartermain
Posts: 99 | From: Aotearoa/New Zealand | Registered: Nov 2002
| IP: Logged
|
|
ken
Ship's Roundhead
# 2460
|
Posted
quote:
Originally posted by Serafina:
No, the Krebs cycle occurs in all eukaryotes (i.e. pretty much everything except bacteria).
Actually it, or other cycles that are very similar to it, occur in many bacteria. And in almost all eukaryotes - there are some amitochondriate eukaryotes which carry out no respiration.
quote:
And it's hard to see how it could have evolved 'in the context of an already sophisticated system' because the sophisticated system probably couldn't have evolved without it; the Krebs cycle is not an optional extra refinement, it an 'absence = dead'.
Not at all. Its just one of many metabolic pathways available to bacxteria, many of which get on quite well without it. Almost certainly evolved by shunting together two ends of a degradative heterotrophic pathway. It's easy to think up Just So Stories to explain it, though hard to find any evidence fro which of them might be true.
And there are plenty of circumstances under which parts of it would be useful. There is no "irreducible complexity" there at all.
It's slightly harder to think up explanations for the electron-transfer pathways of photosynthesis... but I bet it could be done.
--------------------
Ken
L’amor che move il sole e l’altre stelle.
Posts: 39579 | From: London | Registered: Mar 2002
| IP: Logged
|
|
ken
Ship's Roundhead
# 2460
|
Posted
quote:
Originally posted by Faithful Sheepdog:
So, has anyone ever heard of Haldane?s Dilemma? Have any new solutions to the dilemma been proposed? Does the website present it correctly? Is it a ?major scandal??
No its not. It comes out in the wash. Its based on silly ideas of "mutation load" and "selection load" which are hangovers from pre-Darwininian essentialist thinking and simply aren't a problem in real life.
--------------------
Ken
L’amor che move il sole e l’altre stelle.
Posts: 39579 | From: London | Registered: Mar 2002
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
Here's some more on evolutionary algorithms as applied to electronic circuits:
quote:
In recent years, several research groups have demonstrated the potential for artificial evolution to design electronic circuits automatically. An evolutionary algorithm is implemented, usually as a computer program, which mimics some aspects of Darwinian evolution: small random variations (mutations) are repeatedly made to one or more candidate circuit designs. Natural selection is replaced by a fitness measure of the degree to which a circuit meets the target engineering specification of behavior, size, power-consumption, and so on. Mutations resulting in a poorer fitness measurement are rejected, whereas those producing an improvement (or at least no deterioration) are allowed to persist and be built upon by further rounds of variation and selection. Commencing with either randomly generated circuit designs, results of previous evolutionary experiments, or an initial hand-designed attempt, repeated cycles of automated evolution can eventually produce circuits which satisfy the specification. Many elaborations to this basic scheme exist, such as allowing recombination of features from two or more individuals in an evolving population of variants.
The whole article is rather technical, but I can recommend the discussion about half way down the page on the relative merits of constrained evolution versus unconstrained evolution.
Neil
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
Glenn Oldham
Shipmate
# 47
|
Posted
There are very many features of the natural world for which the best explanation, by far, is descent with modification from a common ancestor. I am not aware of the Intelligent Design movement having anything that is positive and detailed to put in the place of that explanation. All I am aware of being offered is vague words about design. But that simply leaves unexplained huge amounts of detail that Darwinian evolution copes with superbly.
The kinds of features that do not sit at all well with intelligent design can be picked out by questions like these:
Why is that “the hand of a man, formed for grasping, that of a mole for digging, …, the paddle of the porpoise, and the wing of the bat, should all be constructed on the same pattern, and should include similar bones, in the same relative positions? “
Why is it that “the hind-feet of the kangaroo, which are so well fitted for bounding over the open plains, those of the climbing, leaf-eating koala, equally well fitted for grasping the branches of trees,- those of the ground-dwelling, insect or root-eating, bandicoots,- and those of some other Australian marsupials,- should all be constructed on the same extraordinary type, namely with the bones of the second and third digits extremely slender and enveloped within the same skin, so that they appear like a single toe furnished with two claws. Notwithstanding this similarity of pattern, it is obvious that the hind feet of these several animals are used for as widely different purposes as it is possible to conceive.”
Why is the designer so fond of the pentadactyl limb even keeping the bones for two separate fingers when they are joined in the same digit? The pentadactyl limb surely cannot be the best design for all these uses. Nelson and Wells comment that “the suboptimality of the pentadactyl limb pattern has never been empirically demonstrated” - a comment that beggars belief. Apparently the pentadactyl limb is the best design for every one of these uses. It clearly is not in the case of the horse, which has but one toe, – but then why does the horse possess the vestiges of the ones it does not use?
Why is it that “the immensely long spiral proboscis of a sphinxmoth, the curious folded one of a bee or bug, and the great jaws of a beetle?- yet all these organs, serving for such widely different purposes, are formed by infinitely numerous modifications of an upper lip, mandibles, and two pairs of maxillae. The same law governs the construction of the mouths and limbs of crustaceans. So it is with the flowers of plants. “
“Why should one crustacean, which has an extremely complex mouth formed of many parts, consequently always have fewer legs; or conversely, those with many legs have simpler mouths? “ i.e. it seems clear that the missing legs have been used for the construction of the additional mouth parts. Does the designer have some hang up about giving crustaceans more complex mouths so he penalises them for it by reducing the number of legs?
“Why should the sepals, petals, stamens, and pistils, in each flower, though fitted for such distinct purposes, be all constructed on the same pattern?”
How do we explain the fact of “the presence of teeth in foetal whales, which when grown up have not a tooth in their heads; or the teeth, which never cut through the gums, in the upper jaws of unborn calves?” Anteaters have teeth as foetuses which they lose before birth and have no teeth thereafter – how is that a case of design?
“How … can we explain these several facts in embryology,- namely, …
- the various parts in the same individual embryo, which ultimately become very unlike and serve for diverse purposes, being at an early period of growth alike;
- the common, but not invariable, resemblance between the embryos or larvae of the most distinct species in the same class;
- the embryo often retaining, whilst within the egg or womb, structures which are of no service to it, either at that or at a later period of life; …?”
Why do some beetles have wings which are “mere rudiments of membrane, which not rarely lie under wing-covers firmly soldered together” and are thus unusable?
Why do “rudimentary organs sometimes retain their potentiality: … thus the tadpole of the common salamander or water-newt, as Mr. G. H. Lewes remarks, "has gills, and passes its existence in the water; but the Salamandra atra, which lives high up among the mountains, brings forth its young full-formed. This animal never lives in the water. Yet if we open a gravid female, we find tadpoles inside her with exquisitely feathered gills; and when placed in water they swim about like the tadpoles of the water-newt. Obviously this aquatic organisation has no reference to the future life of the animal, nor has it any adaptation to its embryonic condition; it has solely reference to ancestral adaptations, it repeats a phase in the development of its progenitors." “
Why is it that “the boa constrictor has rudiments of hind-limbs and of a pelvis, and if it be said that these bones have been retained "to complete the scheme of nature," why, as Professor Weismann asks, have they not been retained by other snakes, which do not possess even a vestige of these same bones? “
Why is the designer so found of making the larygeal nerve go around a structure near the heart that he does it in all vertebrates including even the giraffe where his penchant for this feature makes the nerve go all the way down the neck and all the way back up again making it 5 metres longer than it needs to be?
All of these features fit superbly with evolution by descent with modification from common ancestors. Pre-existing structures are modified in different ways (or fall into disuse) whilst often retaining a marked resemblance with what went before. None of these questions suggest intelligent design as an answer.
And I haven’t even got on to how Darwin explains the geographical distribution of species just as elegantly. And to add to all that we have more and more molecular information that matches the idea of evolution by descent with modification.
(All unattributed quotes are from Darwin’s Origin of Species. )
--------------------
This entire doctrine is worthless except as a subject of dispute. (G. C. Lichtenberg 1742-1799 Aphorism 60 in notebook J of The Waste Books)
Posts: 910 | From: London, England | Registered: May 2001
| IP: Logged
|
|
Justinian
Shipmate
# 5357
|
Posted
Faithful Sheepdog has a point: Dawkins is a weasel. A better illustration of the principle Dawkins was trying to illustrate would be the kids game "Mastermind" played using a simple iterative algorithm.
What's your point about Open v Closed evolution, FS?
Glenn: God's lazy and likes repeating himself? God started the first life and the rest diverged from there, with there being barriers that can't be crossed without outside help?
--------------------
My real name consists of just four letters, but in billions of combinations.
Eudaimonaic Laughter - my blog.
Posts: 3926 | From: The Sea Coast of Bohemia | Registered: Dec 2003
| IP: Logged
|
|
Glenn Oldham
Shipmate
# 47
|
Posted
quote:
Originally posted by Justinian:
Faithful Sheepdog has a point: Dawkins is a weasel. A better illustration of the principle Dawkins was trying to illustrate would be the kids game "Mastermind" played using a simple iterative algorithm.
Anti-darwinists often come up with absurd arguments about randomness in relation to natural selection. Darwinians then offer analogies that are simply trying to show that random variation followed by non-random selection results in the accumulation of useful information, hence the process of natural selection is not random. But all such analogies suffer the same fate in the hands of anti-darwinians: they hunt out every way in which the analogy does not match the details of natural selection and they completely miss the point about randomness plus selection equals non randomness. Dawkins useful Weasel analogy suffers the same fate and is turned by them into something he had no intention of it being: i.e a detailed model of all aspects of natural selection.
quote:
Originally posted by Justinian:
Glenn: God's lazy and likes repeating himself? God started the first life and the rest diverged from there, with there being barriers that can't be crossed without outside help?
Well it's a theory, and, of course since God can do anything and we can't be absolutely sure what he will and will not choose to do there are umpteen zillion possible theories. But the key point is, I think that the idea of common descent is inescapable. Why do alll mammals have 7 neck vertebrae regardless of the use the animal makes of it (rigid like many whales or bendier of twistable in others or very long in giraffes)? The fact that there are 7 cannot be related to design in every case. Why are whales genetically closer to hippos than to other mammals? If God did provide some necessary extra information to evolve the whale from something earlier (crossing the boundary in your terms) we still have the whale still inherits features from that earlier form and so Darwins is right about descent with modification. Is this what Intelligent Designers believe? Who knows? If they don't have at least some commonn descent in they are faced with a large body of similarities which they cannot make sense of.
--------------------
This entire doctrine is worthless except as a subject of dispute. (G. C. Lichtenberg 1742-1799 Aphorism 60 in notebook J of The Waste Books)
Posts: 910 | From: London, England | Registered: May 2001
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
quote:
Justinian said:
What's your point about Open v Closed evolution, FS?
Under Darwinian evolution the only constraints (i.e. the boundary conditions) are:
1) a self-replicating life-form and
2) natural selection of those mutations that provide a breeding advantage.
In Darwinian evolutionary theory for life-forms there is no possibility of changing these constraints – they are the only ones that the theory allows (for simplicity leaving behind the whole issue of genetic drift referred to above).
However, since evolutionary algorithms are only sophisticated numerical models, they have more flexibility in how they model starting conditions and selection constraints. The choice is between a more constrained process (simply called “constrained”) and a less constrained process (called “unconstrained”).
A constrained evolution attempts to keep designs within previously known parameters that have been found to work. In other words the numerical iteration process is unlikely to throw up any real surprises. The trial designs may be quite conventional, and the evolutionary process simply optimises standard technical approaches to the problem. So it’s a safe and reliable process, although somewhat unimaginative.
With unconstrained evolution, the software is programmed to operate “well outside the box”. The initial trial designs may be very unconventional, and bear little or no relation to earlier technical approaches. By exploring many previously unexplored solutions, the algorithm may evolve a feasible designs that is genuinely novel. On the other hand, it may generate an impractical design that is unbuildable, or simply fail to converge at all. So there is more risk in the unconstrained approach.
Such thinking may be very relevant to complex electronic micro-circuits, but since this is outside my own specialist field, I can’t really comment in detail any further. The overall point was that evolutionary algorithms have the freedom to vary the evolutionary process through different starting points and subsequent selection criteria.
quote:
Glenn Oldham said:
All of these features fit superbly with evolution by descent with modification from common ancestors. Pre-existing structures are modified in different ways (or fall into disuse) whilst often retaining a marked resemblance with what went before. None of these questions suggest intelligent design as an answer.
Glenn – an interesting post on homology. I’ve printed it out and will give some thought to a possible reply.
Neil
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
Glenn Oldham
Shipmate
# 47
|
Posted
quote:
Originally posted by Faithful Sheepdog:
quote:
Justinian said:
What's your point about Open v Closed evolution, FS?
Under Darwinian evolution the only constraints (i.e. the boundary conditions) are:
1) a self-replicating life-form and
2) natural selection of those mutations that provide a breeding advantage.
In Darwinian evolutionary theory for life-forms there is no possibility of changing these constraints – they are the only ones that the theory allows (for simplicity leaving behind the whole issue of genetic drift referred to above).
...
Neil, I am still not clear what point you are making in your posts on this nor whether your last post means that you see these algorithims as analogous to natural selection or not.
However, it is important to note that it is potentially misleading to say that "the only constraints" for darwinian evolution are "1) a self-replicating life-form and 2) natural selection of those mutations that provide a breeding advantage." This is because selection occurs by means of a whole range of actual and particular constraints such a food supply, competition with other organisms, parasites, sexual competition, environmental conditions, and so on and so forth. These will vary from species to species and may also vary over time. The reproductive success of organisms will tend to depend on how well adapted they are to deal with these constraints. It is these constraints that are paralleled by the selection for better antennae in the antenna design program.
The other set of constraints operating are the ways in which the laws of physics and chemistry limit the ways in which variation can arise. This parallels the way the antenna program has limits on how the next generation of antennae are produced by alterations to the previous generation.
I think Rex has said much of this already. Forgive me if its already clear to you.
Glenn
--------------------
This entire doctrine is worthless except as a subject of dispute. (G. C. Lichtenberg 1742-1799 Aphorism 60 in notebook J of The Waste Books)
Posts: 910 | From: London, England | Registered: May 2001
| IP: Logged
|
|
Alan Cresswell
Mad Scientist 先生
# 31
|
Posted
quote:
Originally posted by Glenn Oldham:
However, it is important to note that it is potentially misleading to say that "the only constraints" for darwinian evolution are "1) a self-replicating life-form and 2) natural selection of those mutations that provide a breeding advantage." This is because selection occurs by means of a whole range of actual and particular constraints such a food supply, competition with other organisms, parasites, sexual competition, environmental conditions, and so on and so forth. These will vary from species to species and may also vary over time.
Not to forget that different populations of the same species in different geographical locations will experience slightly different constraints that, given some form of isolation between the populations, could result in one population developing slightly differently to the other leading to potential differentiation of the populations into distinct species.
--------------------
Don't cling to a mistake just because you spent a lot of time making it.
Posts: 32413 | From: East Kilbride (Scotland) or 福島 | Registered: May 2001
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
quote:
Glenn Oldham said:
Neil, I am still not clear what point you are making in your posts on this nor whether your last post means that you see these algorithms as analogous to natural selection or not.
The short answer is that I do see these algorithms as reflecting some features of Darwinian evolution, but as I understand it they are not directly analogous to natural selection, especially the unconstrained algorithms.
quote:
Glenn Oldham said:
However, it is important to note that it is potentially misleading to say that "the only constraints" for darwinian evolution are "1) a self-replicating life-form and 2) natural selection of those mutations that provide a breeding advantage." This is because selection occurs by means of a whole range of actual and particular constraints such a food supply, competition with other organisms, parasites, sexual competition, environmental conditions, and so on and so forth. These will vary from species to species and may also vary over time. The reproductive success of organisms will tend to depend on how well adapted they are to deal with these constraints. It is these constraints that are paralleled by the selection for better antennae in the antenna design program.
I agree with you that natural selection is made up of a myriad of influences all operating simultaneously. However, unless I have misunderstood, it all boils down to one simple point: natural selection of a favourable mutation must mean more grandchildren in relative terms.
One of the big differences between evolutionary algorithms for antennas and Darwinian evolution is that the former can specify in a simple deterministic fashion the various selection constraints. In antenna design we can write deterministic equations and set limits for size, weight, material content, frequency range, energy required etc. Within these limits we know from prior experience that an optimised design can be determined and constructed.
I think it would be very difficult to model numerically the actual selection constraints on a living organism, such as “a food supply, competition with other organisms, parasites, sexual competition, environmental conditions, and so on and so forth”. If this can be done numerically at all, I suspect it would have to be based on some kind of probabilistic approach rather than a deterministic one.
There are plenty of places in engineering where a probabilistic approach is used, such as in the calculation of wave loading or turbulent wind loading, but in these cases we can make measurements first and determine the statistical parameters needed by a probabilistic approach.
How one would obtain similarly accurate data in biology I have no idea. Hence the whole complex process of natural selection is actually simplified dramatically by the Darwinian theory to a binary state – either the mutation outbreeds the competition and flourishes, or it doesn’t.
quote:
Glenn Oldham said:
The other set of constraints operating are the ways in which the laws of physics and chemistry limit the ways in which variation can arise. This parallels the way the antenna program has limits on how the next generation of antennae are produced by alterations to the previous generation.
The unconstrained evolutionary algorithm that I described earlier allows us to start with a far-from-optimised design that may be barely functional, or even non-functional. Iterations may also pass through non-functional states in the search to find a fully optimised design. Since a string of numbers never actually “dies”, the computer can carry on with the iteration in the hope of finding some parametric combination that is “alive” – i.e. technically functional and fully optimised.
In the biological world this would be equivalent to a dead initial life-form or to evolutionary changes leading into extinction, something that cannot be permitted at all for a successful species. So the unconstrained evolutionary algorithm departs completely from biological reality at this point – but it may still generate novel and highly optimised antenna designs or whatever.
To my mind Darwinism is more akin to the constrained evolutionary algorithm. An evolving life-form must acquire beneficial mutations in the population and remain functional at all times. This suggests that a more accurate model of the biological process is a highly constrained numerical algorithm, nearly optimised and functional at all stages of the iterative process.
I suspect that a more constrained algorithm is always going to be much less inventive than a less constrained one – i.e. a constrained algorithm may generate improvements on existing designs, but it won’t generate wholly new designs that haven’t already been thought of. In biological terms we may generate new varieties and even the occasional sub-species, but no more than that.
Speaking more generally, I don’t deny that up to a point natural selection in biology is an observed fact. However, in Darwinian evolution it is the only selection constraint allowed to operate, and is held responsible for all physiological and biochemical changes as micro-organisms evolved into mankind. Under natural selection there is assumed to be no limit on the amount of evolution that can take place. I think it remains to be demonstrated that natural selection alone has this creative power.
Neil
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
The sceptical Atheist
Shipmate
# 379
|
Posted
Hope you don't mind me putting in my thoughts, and if I say anything too stoopid tell me to clear off.
I suspect that a more constrained algorithm is always going to be much less inventive than a less constrained one – i.e. a constrained algorithm may generate improvements on existing designs, but it won’t generate wholly new designs that haven’t already been thought of. In biological terms we may generate new varieties and even the occasional sub-species, but no more than that.
Remember, we are looking back with 20/20 hindsight (well, more like 5/5 hindsight considering the fossil record).
Imagine going back to when mammal-like reptiles were alive. The very first mammal would have just been another variety or sub-species at first. It is only now that we see that this sub-species radiated to the wide variety of mammals we see today.
Each big change starts with a sub-species.
Look at the things we consider 'new' in mammals such as fur, lactation, live bearing and in each case it can be seen to be just a modification of a reptilian feature.
Speaking more generally, I don’t deny that up to a point natural selection in biology is an observed fact. However, in Darwinian evolution it is the only selection constraint allowed to operate, and is held responsible for all physiological and biochemical changes as micro-organisms evolved into mankind. Under natural selection there is assumed to be no limit on the amount of evolution that can take place. I think it remains to be demonstrated that natural selection alone has this creative power.
Let us take the cell -> human very quickly.
First multicelled - just a variation.
Then a simple notochord - just a variation
Then cartilidge (sp?)- just a variation
Then a backbone - just a variaton
Then jaws - just a variation
Then bones - just a variation
Then limbs - just a variation
Then lungs - just a variation
Of course the best example of this is the mammalian ear.
The first jawed fishes used the first gill arch of the embryo of the jawless fish and adapted it into a jaw.
Reptiles still use the same gill arch to form the jaw.
Mammals developed their three boned ears from the jaw of reptiles. Our ears develoop from the first gill arch!
The rest is history, but it is easy to see that taking things in context and not jumping steps then each species is more "advanced" and will be able to survive in its environment and yet has what look superficially like new organs.
--------------------
"Faith in God and seventy-five cents will get you a cup of coffee."
[Wayne Aiken]
Posts: 293 | From: Staffordshire | Registered: Jun 2001
| IP: Logged
|
|
Glenn Oldham
Shipmate
# 47
|
Posted
Neil,
You say that:
quote:
Originally posted by Faithful Sheepdog:
To my mind Darwinism is more akin to the constrained evolutionary algorithm. An evolving life-form must acquire beneficial mutations in the population and remain functional at all times. This suggests that a more accurate model of the biological process is a highly constrained numerical algorithm, nearly optimised and functional at all stages of the iterative process.
I suspect that a more constrained algorithm is always going to be much less inventive than a less constrained one – i.e. a constrained algorithm may generate improvements on existing designs
which, (apart from the odd quibble I have with it) that I have is correct. But then you say that:
quote:
... but it won’t generate wholly new designs that haven’t already been thought of. In biological terms we may generate new varieties and even the occasional sub-species, but no more than that.
But on what grounds do you say that? That is very puzzling. The continual gradual modification of existing structures will produce new structures that may eventually be so altered that they are so different from their original structure that the original is uninferable from them.
And even so - and this relates to my post about homologies - look around at the natural world and we see hundreds and thousands of similarities in structure that suggest the alteration of one to the other in ways that go well beyond the species level. All mammals with 7 neck vertebrae, crustaceans with complicated mouths always having fewer legs; whales having no back legs but having a pelvis nevertheless; the tube from the lungs of the whale passing underneath and in front of the brain (as it does in all mammals) rather than going straight up to the top of the head by going above and behind it which would be a good design option). All of these look like the product of an evolutionary process that has indeed been constrained by the requirement that existing structures are altered. And they all extend well beyond the species level. Add the fossil record to this and one has yet more evidence.
--------------------
This entire doctrine is worthless except as a subject of dispute. (G. C. Lichtenberg 1742-1799 Aphorism 60 in notebook J of The Waste Books)
Posts: 910 | From: London, England | Registered: May 2001
| IP: Logged
|
|
Glenn Oldham
Shipmate
# 47
|
Posted
Oh yes, and what I just said is much the same point as the one that skeptical atheist has just made too.
--------------------
This entire doctrine is worthless except as a subject of dispute. (G. C. Lichtenberg 1742-1799 Aphorism 60 in notebook J of The Waste Books)
Posts: 910 | From: London, England | Registered: May 2001
| IP: Logged
|
|
ken
Ship's Roundhead
# 2460
|
Posted
quote:
Originally posted by The sceptical Atheist:
First multicelled - just a variation.
Then a simple notochord - just a variation
Then cartilidge (sp?)- just a variation
Then a backbone - just a variaton
Then jaws - just a variation
Then bones - just a variation
Then limbs - just a variation
Then lungs - just a variation
Actually lungs come before limbs ![[Biased]](wink.gif)
--------------------
Ken
L’amor che move il sole e l’altre stelle.
Posts: 39579 | From: London | Registered: Mar 2002
| IP: Logged
|
|
sharkshooter
Not your average shark
# 1589
|
Posted
I have been trying to understand the main argument here. I think it can be summarized:
"Similarity in design means evolution occurred."
Is that correct?
My house is of similar design to my neighbour's. However, mine has one more bedroom and a den. Does that mean mine evolved from his (since the extra rooms make it a "better" house)? That his evolved from mine (as the other bedroom and den were no longer required)? Or, that they have a common house as an ancestor? Perhas they have the same designer?
--------------------
Let the words of my mouth, and the meditation of my heart, be acceptable in thy sight, O LORD, my strength, and my redeemer. [Psalm 19:14]
Posts: 7772 | From: Canada; Washington DC; Phoenix; it's complicated | Registered: Oct 2001
| IP: Logged
|
|
ken
Ship's Roundhead
# 2460
|
Posted
quote:
Originally posted by sharkshooter:
I have been trying to understand the main argument here. I think it can be summarized:
"Similarity in design means evolution occurred."
Is that correct?
No it is completely wrong.
--------------------
Ken
L’amor che move il sole e l’altre stelle.
Posts: 39579 | From: London | Registered: Mar 2002
| IP: Logged
|
|
sharkshooter
Not your average shark
# 1589
|
Posted
quote:
Originally posted by ken:
quote:
Originally posted by sharkshooter:
I have been trying to understand the main argument here. I think it can be summarized:
"Similarity in design means evolution occurred."
Is that correct?
No it is completely wrong.
That certainly seemed to be Glenn Oldham's point.
--------------------
Let the words of my mouth, and the meditation of my heart, be acceptable in thy sight, O LORD, my strength, and my redeemer. [Psalm 19:14]
Posts: 7772 | From: Canada; Washington DC; Phoenix; it's complicated | Registered: Oct 2001
| IP: Logged
|
|
Alaric the Goth
Shipmate
# 511
|
Posted
Sharkshooter, there is something called 'convergent evolution'. According to this idea, you will end up with similar features evolving in unrelated organisms who are adaprting to similar environmental conditions. Thus you have a dolphin-like shape evolving in the icthyosaurs (marine reptiles contemporaneous with the dinosaurs) millions of years before the actual cetaceans (whales and dolphins) evolve. The marine mammals evolve to fill the ecological niche(s) left 'vacant' by the mass extinction of sea reptiles etc. at the end of the Cretaceous period.
It also explains why the marsupial Thylacine evolved a wolf-like shape as it was (until its 20th century extinction) a predator in the same 'niche' as the unrelated placental mammals of Eurasia and North America. (Australasia having no placentals till humans introduced the dog (dingo) a few thousand years ago.)
Posts: 3322 | From: West Thriding | Registered: Jun 2001
| IP: Logged
|
|
sharkshooter
Not your average shark
# 1589
|
Posted
quote:
Originally posted by Alaric the Goth:
...'convergent evolution'. According to this idea, you will end up with similar features evolving in unrelated organisms who are adaprting to similar environmental conditions. ...
So, "similar features for similar environments"? (I put it as a question, in case I did not understand your point. My intention is not to twist your words, but to understand your meaning.)
This supports your "convergent evolution" (I'll take that as a given) as well as a Creator who understood what different species of His creation needed for these environments.
Would it have been necessary for God to have created each species completely different for people to believe that He created them? So, for example, no two species with the same number of vertebrae, no two species with the same number or design of eyes, no two species with the same basic limb design, etc.
--------------------
Let the words of my mouth, and the meditation of my heart, be acceptable in thy sight, O LORD, my strength, and my redeemer. [Psalm 19:14]
Posts: 7772 | From: Canada; Washington DC; Phoenix; it's complicated | Registered: Oct 2001
| IP: Logged
|
|
Alan Cresswell
Mad Scientist 先生
# 31
|
Posted
quote:
Originally posted by sharkshooter:
quote:
Originally posted by ken:
quote:
Originally posted by sharkshooter:
I have been trying to understand the main argument here. I think it can be summarized:
"Similarity in design means evolution occurred."
Is that correct?
No it is completely wrong.
That certainly seemed to be Glenn Oldham's point.
No, the point was more that the same basic feature occurs even though it has no real reason to be the same.
So, in your house analogy, it could be something like an "ancestral house" having a fireplace then subsequent houses maintain that feature even though in many of them the fire becomes a couple of electric bars and there is no chimney, and that the introduction of central heating has made it superfluous anyway.
--------------------
Don't cling to a mistake just because you spent a lot of time making it.
Posts: 32413 | From: East Kilbride (Scotland) or 福島 | Registered: May 2001
| IP: Logged
|
|
Faithful Sheepdog
Shipmate
# 2305
|
Posted
quote:
Originally posted by Glenn Oldham:
Neil,
You say that:
quote:
Originally posted by Faithful Sheepdog:
To my mind Darwinism is more akin to the constrained evolutionary algorithm. An evolving life-form must acquire beneficial mutations in the population and remain functional at all times. This suggests that a more accurate model of the biological process is a highly constrained numerical algorithm, nearly optimised and functional at all stages of the iterative process.
I suspect that a more constrained algorithm is always going to be much less inventive than a less constrained one – i.e. a constrained algorithm may generate improvements on existing designs
which, (apart from the odd quibble I have with it) that I have is correct. But then you say that:
quote:
... but it won’t generate wholly new designs that haven’t already been thought of. In biological terms we may generate new varieties and even the occasional sub-species, but no more than that.
But on what grounds do you say that? That is very puzzling. The continual gradual modification of existing structures will produce new structures that may eventually be so altered that they are so different from their original structure that the original is uninferable from them.
<snip>
This is where I wish I could see the various equations lying behind the selection constraints in these evolutionary algorithms. Since neither antenna design nor electronic micro-circuits are my technical speciality, I can only talk in mathematical generalities at the moment.
I based my previous comments on the limitations of constrained evolution on deductions from the webpage comments on evolutionary algorithms in relation to electronic micro-circuits. It does seem intuitively obvious to me that, in a complex iterative process trying to optimise the parameters of a multi-dimensional problem, the mathematics of the selection constraints can be varied at the discretion of the programmer, and that this will have an enormous influence on the result.
A very loose set of selection constraints may mean that the software never converges to anything specific, or oscillates about a mean, or even becomes unstable. On the other hand, a very tight set of constraints may produce little or no improvement on existing (non-evolutionary) designs.
I have occasionally used some technical graphing software to produce contour graphs, and this gives me the choice of various algorithms for producing contours. There have been some cases where one algorithm gave me sensible results, and another gave me garbage. Plenty of things can go wrong in large scale numerical manipulation.
To sum up, I think the direct analogy between evolutionary algorithms and biological evolution breaks down at the point of the programmed selection constraints. Natural selection in biology is probably too complex a phenomenon to be modelled accurately at present.
I haven’t forgotten your homology post – I’m still thinking about it.
Neil
--------------------
"Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world." ~ Michael J. Behe
Posts: 1097 | From: Scotland | Registered: Feb 2002
| IP: Logged
|
|
The sceptical Atheist
Shipmate
# 379
|
Posted
So, "similar features for similar environments"? (I put it as a question, in case I did not understand your point. My intention is not to twist your words, but to understand your meaning.)
This supports your "convergent evolution" (I'll take that as a given) as well as a Creator who understood what different species of His creation needed for these environments.
Would it have been necessary for God to have created each species completely different for people to believe that He created them? So, for example, no two species with the same number of vertebrae, no two species with the same number or design of eyes, no two species with the same basic limb design, etc.
No, it is not necessary, God could create how he wanted, but evolution can't. Evolution is severely constrained.
It is in the detail that the differences become apparent. Let us take for an example birds and bats. Both fly and so both have wings.
The wings are different and yet look superficially the same because of convergent evolution. Bats cannot have feathers because they are not birds, and birds cannot feed their chicks with milk because they are not mammals.
And yet, both of these have backbones.
There is a strict heirachy of living things and creatures cannot pick and choose what they have. If they were all designed seperately, then any design could work, and a nested hierachy is not necessary, you could take the best bits of each and put them together. Evolution demands a strict hierachy.
Of course, this in itself is not evidence one way or the other. To understand why evolution is the way life developed, all the other evidence needs to be considered, but just the fact that life does fall into a nested hierachy with no exceptions supports evolution in a very important way.
I haven't explained this very well at all, sorry.
--------------------
"Faith in God and seventy-five cents will get you a cup of coffee."
[Wayne Aiken]
Posts: 293 | From: Staffordshire | Registered: Jun 2001
| IP: Logged
|
|
sharkshooter
Not your average shark
# 1589
|
Posted
quote:
Originally posted by The sceptical Atheist:
...
The wings are different and yet look superficially the same because of convergent evolution. ...
This is the part with which I disagree. I think the following is as reasonable a statement:
quote:
The wings are different and yet look superficially the same because they were created for different species but the same purpose.
Either statement works - so it does not support evolution.
--------------------
Let the words of my mouth, and the meditation of my heart, be acceptable in thy sight, O LORD, my strength, and my redeemer. [Psalm 19:14]
Posts: 7772 | From: Canada; Washington DC; Phoenix; it's complicated | Registered: Oct 2001
| IP: Logged
|
|
The sceptical Atheist
Shipmate
# 379
|
Posted
Sharkshooter,
I understand what you mean. If that was the only consideration then you would have a point.
We can see that a bat is a mammal and that a bird is, well, a bird. This is from the physiology.
We can see from the fossil evidence that birds evolved from reptiles, probably dinosaurs.
We can see that mammals evolved from reptiles, too, but a lot earlier.
We can examine the embryology and see that bothe birds and mammals are related to reptiles, but that the embryos show fundamental differences that highlight the different evolutionary pathways (especially based on the circulation system).
We can examine the genetics and see the way that birds and mammals are related to reptiles is exactly what we would expect if evolution had evolved them in exactly the way that the fossil record / embryology and physysiology would expect.
Each of these is evidence for evolution, but none of them are enough on their own. Together, it becomes clear that birds and bats evolved, they were not seperately created in isolation from reptiles.
--------------------
"Faith in God and seventy-five cents will get you a cup of coffee."
[Wayne Aiken]
Posts: 293 | From: Staffordshire | Registered: Jun 2001
| IP: Logged
|
|
|
Printer-friendly view
