Total Posts:64|Showing Posts:1-30|Last Page
Jump to topic:

No evidence that mutations are random

janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
3/31/2015 3:08:54 PM
Posted: 1 year ago
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)
tkubok
Posts: 5,044
Add as Friend
Challenge to a Debate
Send a Message
3/31/2015 11:01:02 PM
Posted: 1 year ago
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What do you mean by "predictable".

So you are able to predict which bacteria will produce offspring that is immune to antibiotics? You are able to predict which generation of bacteria will be able to digest Lactose?
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 12:00:01 AM
Posted: 1 year ago
At 3/31/2015 11:01:02 PM, tkubok wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What do you mean by "predictable".

So you are able to predict which bacteria will produce offspring that is immune to antibiotics? You are able to predict which generation of bacteria will be able to digest Lactose?

http://www.nytimes.com...
RuvDraba
Posts: 6,033
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 4:44:28 AM
Posted: 1 year ago
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.
If you have proof, then show it.
Mutations are NOT random, and they are predictable.

Jane, I believe you may be confusing changes in genotype (genetic code), and hereditary changes in phenotype (observable form and function, some of which can be inherited.)

Over time, benefits to phenotype can be inherited genetically -- that's evolution. So in a sense, you're right: if you know the species' current genotype and niche, you can guess certain ways it might evolve over time. An example of this is our confirmed expectation that bacteria will grow resistant to common antibiotics over time -- since common antibiotics change the niche for those bacteria.

But which specific individuals might receive these benefits, in what order and when, are not so easy to predict. And there may be other changes we don't predict, including changes that are neither harmful nor beneficial but just happen to carry down through heredity.

As the University of Utah explains:
Most DNA changes fall in the large areas of the genome that sit between genes, and usually they have no effect. When variations occur within genes, there is more often a consequence, but even then mutation only rarely causes death or disease. Mutation also generates new variations that can give an individual a survival advantage. And most often, mutation gives rise to variations that are neither good nor bad, just different.
-- University of Utah, What is mutation? http://learn.genetics.utah.edu...

I realise that's not 'proof' -- simply explanation. The proof lies in the way mutations actually occur, but I think the explanation above is the key missing insight. Happy to dig deeper if that would help.
tkubok
Posts: 5,044
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 7:44:11 AM
Posted: 1 year ago
At 4/1/2015 12:00:01 AM, janesix wrote:
At 3/31/2015 11:01:02 PM, tkubok wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What do you mean by "predictable".

So you are able to predict which bacteria will produce offspring that is immune to antibiotics? You are able to predict which generation of bacteria will be able to digest Lactose?

http://www.nytimes.com...

So youre asserting that you were able to predict which generation these bacteria evolved extra tails?
Otokage
Posts: 2,347
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 8:47:41 AM
Posted: 1 year ago
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

Mutations on a particular genome can not be predicted, so yes they are random by definition. Although mutation rates due to errors in the replication process can be calculated very vaguely, and it is often said that E.coli's DNA polymerase III has a chance of 10^(-7) of commiting a substitution type mutation every time it adds a nucleotide in the replication process of an haploid cell. Although I don't think this statistic takes into account the cell mechanisms to repair mutations, or even the own correction mechanism of the polymerase...
tkubok
Posts: 5,044
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 9:34:40 AM
Posted: 1 year ago
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What you seem to be talking about, is that mutations are predictable in the same way that we can predict that there will be a lottery winner, eventually. But the actual lottery, any lottery, is random by nature. And to claim that lotteries are rigged because we can predict that there will eventually be a winner, is absurd, to say the least.
Otokage
Posts: 2,347
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 12:43:03 PM
Posted: 1 year ago
At 4/1/2015 9:34:40 AM, tkubok wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What you seem to be talking about, is that mutations are predictable in the same way that we can predict that there will be a lottery winner, eventually. But the actual lottery, any lottery, is random by nature. And to claim that lotteries are rigged because we can predict that there will eventually be a winner, is absurd, to say the least.

Best analogy so far.
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 3:37:00 PM
Posted: 1 year ago
At 4/1/2015 4:44:28 AM, RuvDraba wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.
If you have proof, then show it.
Mutations are NOT random, and they are predictable.

Jane, I believe you may be confusing changes in genotype (genetic code), and hereditary changes in phenotype (observable form and function, some of which can be inherited.)

Over time, benefits to phenotype can be inherited genetically -- that's evolution. So in a sense, you're right: if you know the species' current genotype and niche, you can guess certain ways it might evolve over time. An example of this is our confirmed expectation that bacteria will grow resistant to common antibiotics over time -- since common antibiotics change the niche for those bacteria.

But which specific individuals might receive these benefits, in what order and when, are not so easy to predict. And there may be other changes we don't predict, including changes that are neither harmful nor beneficial but just happen to carry down through heredity.

As the University of Utah explains:
Most DNA changes fall in the large areas of the genome that sit between genes, and usually they have no effect. When variations occur within genes, there is more often a consequence, but even then mutation only rarely causes death or disease. Mutation also generates new variations that can give an individual a survival advantage. And most often, mutation gives rise to variations that are neither good nor bad, just different.
-- University of Utah, What is mutation? http://learn.genetics.utah.edu...

I realise that's not 'proof' -- simply explanation. The proof lies in the way mutations actually occur, but I think the explanation above is the key missing insight. Happy to dig deeper if that would help.

Thank you for your polite reply.

The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 3:40:04 PM
Posted: 1 year ago
At 4/1/2015 7:44:11 AM, tkubok wrote:
At 4/1/2015 12:00:01 AM, janesix wrote:
At 3/31/2015 11:01:02 PM, tkubok wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What do you mean by "predictable".

So you are able to predict which bacteria will produce offspring that is immune to antibiotics? You are able to predict which generation of bacteria will be able to digest Lactose?

http://www.nytimes.com...

So youre asserting that you were able to predict which generation these bacteria evolved extra tails?

No. I assume that the scientist would could accurately predict that if they did another 24 trials, 24 would produce the extra tails, given the same exact conditions.
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 3:47:20 PM
Posted: 1 year ago
At 4/1/2015 8:47:41 AM, Otokage wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

Mutations on a particular genome can not be predicted, so yes they are random by definition. Although mutation rates due to errors in the replication process can be calculated very vaguely, and it is often said that E.coli's DNA polymerase III has a chance of 10^(-7) of commiting a substitution type mutation every time it adds a nucleotide in the replication process of an haploid cell. Although I don't think this statistic takes into account the cell mechanisms to repair mutations, or even the own correction mechanism of the polymerase...

Random means the cause wasn't predetermined in any way.

But some predictions can be made, and the same mutations will occur under specific conditions. If all mutations were random, that wouldn't be the case.
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 3:52:34 PM
Posted: 1 year ago
At 4/1/2015 9:34:40 AM, tkubok wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What you seem to be talking about, is that mutations are predictable in the same way that we can predict that there will be a lottery winner, eventually. But the actual lottery, any lottery, is random by nature. And to claim that lotteries are rigged because we can predict that there will eventually be a winner, is absurd, to say the least.

I am claiming that mutations are predictable in that they will be the same under specific conditions.
Burzmali
Posts: 1,310
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 4:11:32 PM
Posted: 1 year ago
At 4/1/2015 3:52:34 PM, janesix wrote:
At 4/1/2015 9:34:40 AM, tkubok wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What you seem to be talking about, is that mutations are predictable in the same way that we can predict that there will be a lottery winner, eventually. But the actual lottery, any lottery, is random by nature. And to claim that lotteries are rigged because we can predict that there will eventually be a winner, is absurd, to say the least.

I am claiming that mutations are predictable in that they will be the same under specific conditions.

Depends on how specific you're getting with those conditions. Given identical conditions, lottery numbers will be the same each time they're drawn. But those conditions are so nebulous and uncontrollable as to effectively be random. The same goes for actual mutations, meaning mistakes in the copying of DNA.
RuvDraba
Posts: 6,033
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 4:23:31 PM
Posted: 1 year ago
At 4/1/2015 3:37:00 PM, janesix wrote:
The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.

Jane, mutation has been observed. It occurs through ordinary biochemistry, radiation, hybridisation, and a newly-explored area called epigenetics -- the impact of environment on genetic replication.

These changes themselves are unpredictable -- nobody knows where in the genome of which individuals the changes may occur, nor what the changes will produce.

However, evolution works across populations, so statistics apply. For example, in another thread we were talking about the parallel evolution of placental and marsupial dogs. If there were a gene for snout-length, then in a large population of dogs we could expect some dogs to have longer snouts and some shorter. If the dogs with longer snouts gain reproductive benefit (for example, they're better at tracking food), then we could expect the next generation of dogs to have longer snouts. So if we know the genome, the population size and the environmental pressures then we can predict certain mutations statistically.

But here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

This illusion of purpose in evolution can be compelling, even to someone professionally science-trained like me. I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Programs are first seeded 'randomly' using simple, blind functions like move, turn, shoot or kick. Scores are assessed based on the constraints of the game. A 'generation' is a population of random programs, and each new generation of programs is created by combining the most successful programs from previous generations. There is no human intervention in these tests, nor any guidance in how to fight or play soccer, but it really looks as though they're trying to learn.

That's just the weirdness that is evolution.
tkubok
Posts: 5,044
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 5:09:33 PM
Posted: 1 year ago
At 4/1/2015 3:52:34 PM, janesix wrote:
At 4/1/2015 9:34:40 AM, tkubok wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

What you seem to be talking about, is that mutations are predictable in the same way that we can predict that there will be a lottery winner, eventually. But the actual lottery, any lottery, is random by nature. And to claim that lotteries are rigged because we can predict that there will eventually be a winner, is absurd, to say the least.

I am claiming that mutations are predictable in that they will be the same under specific conditions.

Im pretty sure id always win the lottery under specific circumstances.

However, this is clearly demonstrated as being wrong, as demonstrated by Lenskis E. coli experiment, where only 1 of the 12 strains of E. coli managed to evolve to utilize citrate, while the other 11 did not, despite being placed in similar circumstances.
Mhykiel
Posts: 5,987
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 5:29:50 PM
Posted: 1 year ago
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
At 4/1/2015 3:37:00 PM, janesix wrote:
The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.

Jane, mutation has been observed. It occurs through ordinary biochemistry, radiation, hybridisation, and a newly-explored area called epigenetics -- the impact of environment on genetic replication.

These changes themselves are unpredictable -- nobody knows where in the genome of which individuals the changes may occur, nor what the changes will produce.

However, evolution works across populations, so statistics apply. For example, in another thread we were talking about the parallel evolution of placental and marsupial dogs. If there were a gene for snout-length, then in a large population of dogs we could expect some dogs to have longer snouts and some shorter. If the dogs with longer snouts gain reproductive benefit (for example, they're better at tracking food), then we could expect the next generation of dogs to have longer snouts. So if we know the genome, the population size and the environmental pressures then we can predict certain mutations statistically.

But here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

This illusion of purpose in evolution can be compelling, even to someone professionally science-trained like me. I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Programs are first seeded 'randomly' using simple, blind functions like move, turn, shoot or kick. Scores are assessed based on the constraints of the game. A 'generation' is a population of random programs, and each new generation of programs is created by combining the most successful programs from previous generations. There is no human intervention in these tests, nor any guidance in how to fight or play soccer, but it really looks as though they're trying to learn.




That's just the weirdness that is evolution.

Wow Selection criteria actually effect the outcome who would have thought?

guess what those programs will only become better and better at fighting and playing football, They won't ever evolve into anything else.
RuvDraba
Posts: 6,033
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 5:41:09 PM
Posted: 1 year ago
At 4/1/2015 5:29:50 PM, Mhykiel wrote:
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Wow Selection criteria actually effect the outcome who would have thought?

Yes. They affect the outcome, but don't determine the strategy by which the outcome is achieved. Evolution is blind, but very creative, offering multiple strategies to satisfy selection criteria, combining them to produce even more complex strategies, and adapting to competing strategies to produce yet more sophistication. In the videos we saw many examples of these.

guess what those programs will only become better and better at fighting and playing football,

What's interesting is that:
1) this is achieved over small populations with very limited functions in a very small number of generations. Yet the results are remarkable. Clearly, evolution can be a very powerful adaptive mechanism;
2) the selection criteria for these games are simple, yet complex 'intelligent' strategies and sophisticated functions emerge. More complex criteria are unlikely to reduce the sophistication of the strategies.

They won't ever evolve into anything else.

Regarding them not evolving into anything else, Mhykiel, arguably they already are. In the fighting example, fighters seem to be speciating toward snipers and meleers, while In the soccer example, individuals seem to be speciating toward specialist attackers and defenders.

You have an epistomological question to reflect on here: what does it mean to be something else? If we gave them obstacles to hide behind and they began to change size, would that be enough?
joepalcsak
Posts: 409
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 6:04:34 PM
Posted: 1 year ago
At 4/1/2015 4:44:28 AM, RuvDraba wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.
If you have proof, then show it.
Mutations are NOT random, and they are predictable.

Jane, I believe you may be confusing changes in genotype (genetic code), and hereditary changes in phenotype (observable form and function, some of which can be inherited.)

Over time, benefits to phenotype can be inherited genetically -- that's evolution. So in a sense, you're right: if you know the species' current genotype and niche, you can guess certain ways it might evolve over time. An example of this is our confirmed expectation that bacteria will grow resistant to common antibiotics over time -- since common antibiotics change the niche for those bacteria.

But which specific individuals might receive these benefits, in what order and when, are not so easy to predict. And there may be other changes we don't predict, including changes that are neither harmful nor beneficial but just happen to carry down through heredity.

As the University of Utah explains:
Most DNA changes fall in the large areas of the genome that sit between genes, and usually they have no effect. When variations occur within genes, there is more often a consequence, but even then mutation only rarely causes death or disease. Mutation also generates new variations that can give an individual a survival advantage. And most often, mutation gives rise to variations that are neither good nor bad, just different.
-- University of Utah, What is mutation? http://learn.genetics.utah.edu...

I realise that's not 'proof' -- simply explanation. The proof lies in the way mutations actually occur, but I think the explanation above is the key missing insight. Happy to dig deeper if that would help.

Of course this begs the question: how do mutations actually occur? clearly, random mutations exist, but random mutations have never done the kind of work that Darwinian evolution needs them to do; indeed, random mutations, when completely understood, inevitably diminish function and do not add information to the genome. If this strain of bacteria is rapidly gaining additional tails 100% of the time, and these tails are due to a specific mutation 100% of the time, then we have strong reason to believe that the word "random" is not appropriate in this case.

Indeed, the sort of (truly) beneficial evolution that actually does occur; that actually can be observed; that changes entire populations rapidly; happens too quickly and too regularly to be ascribed to random mutations. Barbara McClintock was awarded a reluctant nobel prize for the work she did in this area. I say reluctant because if the important mutations turn out to be non-random, then the current dogma is dead in the water, which is repugnant to the high priests of the current dogma.
RuvDraba
Posts: 6,033
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 6:32:33 PM
Posted: 1 year ago
At 4/1/2015 6:04:34 PM, joepalcsak wrote:
random mutations have never done the kind of work that Darwinian evolution needs them to do;

Random mutations, by definition, do everything possible eventually -- some things beneficial, some detrimental, and many neutral or unobservable. But they are a very slow and error-prone way to innovate, and this is well-understood by biologists.

if the important mutations turn out to be non-random, then the current dogma is dead in the water, which is repugnant to the high priests of the current dogma.

Your rhetoric aside, Joe, it's possible that there are faster adaptive mechanisms than random mutation. For example, epigenetics and interspecies hybridisation are being explored productively.

However, random mutation has been shown to deliver benefit in experiment [http://www.newscientist.com...] so additional mechanisms do not invalidate it as a key mechanism -- especially for simpler species -- nor do additional mechanisms imply destiny or any other kind of supernatural intervention.

And evolution isn't dogma. Like all science, it can be amended or replaced by good evidence. And unlike politicians, theologians and philosophers, scientists routinely admit error, change their minds, and do not consider apologia more important than evidence.

Moreover, many theologically-motivated objections are arguments of ignorance and bigotry rather than independent scientific critique. Errors of ignorance can be corrected with information, but errors of bigotry disqualify the speaker from intelligent discourse, since it's pointless when someone is convinced they have all the answers, and won't distinguish evidence from rhetoric.

With that in mind, I'd invite you to reconsider your language, Joe. The idea of insulting scientists by calling them priests is both ignorant (because it's untrue), and ironic, given your own theology and its history.
Otokage
Posts: 2,347
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 7:33:58 PM
Posted: 1 year ago
At 4/1/2015 3:47:20 PM, janesix wrote:
At 4/1/2015 8:47:41 AM, Otokage wrote:
At 3/31/2015 3:08:54 PM, janesix wrote:
There is no evidence that mutations that cause new species, or adaptations to the environment are random.

If you have proof, then show it.

Mutations are NOT random, and they are predictable.

(there can, of course, be accidents in nature, but these are usually detrimental, and not the cause of adaptation or new species)

Mutations on a particular genome can not be predicted, so yes they are random by definition. Although mutation rates due to errors in the replication process can be calculated very vaguely, and it is often said that E.coli's DNA polymerase III has a chance of 10^(-7) of commiting a substitution type mutation every time it adds a nucleotide in the replication process of an haploid cell. Although I don't think this statistic takes into account the cell mechanisms to repair mutations, or even the own correction mechanism of the polymerase...

Random means the cause wasn't predetermined in any way.

Random: Statistics. of or characterizing a process of selection in which each item of a set has an equal probability of being chosen. (http://dictionary.reference.com...)

Nucleotides of a genome have an equal probability of being mutated by a replication error. So mutations are random.

But some predictions can be made, and the same mutations will occur under specific conditions.

That's not a predictable event, because you can not predict anything about it, just that it will happen. Someone explained it to you with a good analogy. Is the lotery winner random? Yes. The only thing you know about lotery is that someone will eventualy win it.

If all mutations were random, that wouldn't be the case.
Mhykiel
Posts: 5,987
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 9:28:41 PM
Posted: 1 year ago
At 4/1/2015 5:41:09 PM, RuvDraba wrote:
At 4/1/2015 5:29:50 PM, Mhykiel wrote:
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Wow Selection criteria actually effect the outcome who would have thought?

Yes. They affect the outcome, but don't determine the strategy by which the outcome is achieved. Evolution is blind, but very creative, offering multiple strategies to satisfy selection criteria, combining them to produce even more complex strategies, and adapting to competing strategies to produce yet more sophistication. In the videos we saw many examples of these.

guess what those programs will only become better and better at fighting and playing football,

What's interesting is that:
1) this is achieved over small populations with very limited functions in a very small number of generations. Yet the results are remarkable. Clearly, evolution can be a very powerful adaptive mechanism;
2) the selection criteria for these games are simple, yet complex 'intelligent' strategies and sophisticated functions emerge. More complex criteria are unlikely to reduce the sophistication of the strategies.

They won't ever evolve into anything else.

Regarding them not evolving into anything else, Mhykiel, arguably they already are. In the fighting example, fighters seem to be speciating toward snipers and meleers, while In the soccer example, individuals seem to be speciating toward specialist attackers and defenders.

You have an epistomological question to reflect on here: what does it mean to be something else? If we gave them obstacles to hide behind and they began to change size, would that be enough?

Where's the fighter that sprouts wings and flies?

Again a great example that instructions can only arise in a possible set of changes. That there is a limit to the domain of these possibilities.

Still waiting for that dog breeder to make a dog the size of a horse or with gills from a ancient ancestor.
RuvDraba
Posts: 6,033
Add as Friend
Challenge to a Debate
Send a Message
4/1/2015 11:46:10 PM
Posted: 1 year ago
At 4/1/2015 9:28:41 PM, Mhykiel wrote:
At 4/1/2015 5:41:09 PM, RuvDraba wrote:
At 4/1/2015 5:29:50 PM, Mhykiel wrote:
those programs will only become better and better at fighting and playing football,
They won't ever evolve into anything else.
Regarding them not evolving into anything else, Mhykiel, arguably they already are. In the fighting example, fighters seem to be speciating toward snipers and meleers, while In the soccer example, individuals seem to be speciating toward specialist attackers and defenders.

Where's the fighter that sprouts wings and flies?

That's a different experiment.

For a video showing how evolution of simple, composable functional elements can assemble genetically to produce different kinds of movement, please see the supplied links.

Of course, complex organ changes are rare and difficult to observe in nature, so you might conclude that we haven't observed any yet.

Except, we have. Here are two f'rinstances:

1. Rapid large-scale evolutionary divergence in morphology and performance associated with exploitation of a different dietary resource
(Herrel et al, PNAS: http://www.pnas.org...)
Here we show how lizards have rapidly evolved differences in head morphology, bite strength, and digestive tract structure after experimental introduction into a novel environment. Despite the short time scale (~36 years) since this introduction, these changes in morphology and performance parallel those typically documented among species and even families of lizards in both the type and extent of their specialization. Moreover, these changes have occurred side-by-side with dramatic changes in population density and social structure, providing a compelling example of how the invasion of a novel habitat can evolutionarily drive multiple aspects of the phenotype.

This mutation included the formation of completely new organs in the lizards' gut known as cecal valves, which helped the lizards switch to a new, strictly plant diet. So, these lizards have evolved new stomachs to become vegetarian.

2. FGF signaling delineates the cardiac progenitor field in the simple chordate, Ciona intestinalis.
(Davidson et al, NCBI: http://www.ncbi.nlm.nih.gov...)

The abstract is full of jargon but in plain English, the sea squirt Ciona Intestinalis is a well-studied organism with its entire genome sequenced. Scientist discovered that a simple mutation changed its simple, single-chambered heart into a more complex multichambered one. So the sea-squirt now has a new kind of heart -- in fact, one much more like a mammal's.

We also have examples of new speciation, new genetic material, beneficial mutations, big morphological changes, and even the evolution of a multicellular organism from smallar single-cells: http://phylointelligence.com...

So will you lift the bar indefinitely, Mhykiel? Demand that science now produce a reconstituted dinosaur and show it mutating feathered wings in time-lapse photography, only to insist that the supernatural snuck between the photo-frames?

At some point one has to define a line between skepticism and bigotry -- and your own line keeps moving.
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/2/2015 12:35:54 PM
Posted: 1 year ago
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
At 4/1/2015 3:37:00 PM, janesix wrote:
The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.

Jane, mutation has been observed. It occurs through ordinary biochemistry, radiation, hybridisation, and a newly-explored area called epigenetics -- the impact of environment on genetic replication.

These changes themselves are unpredictable -- nobody knows where in the genome of which individuals the changes may occur, nor what the changes will produce.

However, evolution works across populations, so statistics apply. For example, in another thread we were talking about the parallel evolution of placental and marsupial dogs. If there were a gene for snout-length, then in a large population of dogs we could expect some dogs to have longer snouts and some shorter. If the dogs with longer snouts gain reproductive benefit (for example, they're better at tracking food), then we could expect the next generation of dogs to have longer snouts. So if we know the genome, the population size and the environmental pressures then we can predict certain mutations statistically.

But here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

This illusion of purpose in evolution can be compelling, even to someone professionally science-trained like me. I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Programs are first seeded 'randomly' using simple, blind functions like move, turn, shoot or kick. Scores are assessed based on the constraints of the game. A 'generation' is a population of random programs, and each new generation of programs is created by combining the most successful programs from previous generations. There is no human intervention in these tests, nor any guidance in how to fight or play soccer, but it really looks as though they're trying to learn.




That's just the weirdness that is evolution.

I agree that some mutations are accidents(and thus random). What I'm saying is that the mutations that drive evolution aren't random. How else to explain the studies that show predictable results in mutations with repeatable, specific conditions? How could the same mutation "accidentally" occur time and time again?
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/2/2015 12:42:32 PM
Posted: 1 year ago
At 4/1/2015 6:32:33 PM, RuvDraba wrote:
At 4/1/2015 6:04:34 PM, joepalcsak wrote:
random mutations have never done the kind of work that Darwinian evolution needs them to do;

Random mutations, by definition, do everything possible eventually -- some things beneficial, some detrimental, and many neutral or unobservable. But they are a very slow and error-prone way to innovate, and this is well-understood by biologists.

if the important mutations turn out to be non-random, then the current dogma is dead in the water, which is repugnant to the high priests of the current dogma.

Your rhetoric aside, Joe, it's possible that there are faster adaptive mechanisms than random mutation. For example, epigenetics and interspecies hybridisation are being explored productively.

However, random mutation has been shown to deliver benefit in experiment [http://www.newscientist.com...] so additional mechanisms do not invalidate it as a key mechanism -- especially for simpler species -- nor do additional mechanisms imply destiny or any other kind of supernatural intervention.

And evolution isn't dogma. Like all science, it can be amended or replaced by good evidence. And unlike politicians, theologians and philosophers, scientists routinely admit error, change their minds, and do not consider apologia more important than evidence.

Moreover, many theologically-motivated objections are arguments of ignorance and bigotry rather than independent scientific critique. Errors of ignorance can be corrected with information, but errors of bigotry disqualify the speaker from intelligent discourse, since it's pointless when someone is convinced they have all the answers, and won't distinguish evidence from rhetoric.

With that in mind, I'd invite you to reconsider your language, Joe. The idea of insulting scientists by calling them priests is both ignorant (because it's untrue), and ironic, given your own theology and its history.

"nor do additional mechanisms imply destiny or any other kind of supernatural intervention."

It could imply that there are mechanisms that are as yet undetected. No supernatural or "destiny" needed. Just natural laws.
tkubok
Posts: 5,044
Add as Friend
Challenge to a Debate
Send a Message
4/2/2015 2:03:29 PM
Posted: 1 year ago
At 4/2/2015 12:35:54 PM, janesix wrote:
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
At 4/1/2015 3:37:00 PM, janesix wrote:
The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.

Jane, mutation has been observed. It occurs through ordinary biochemistry, radiation, hybridisation, and a newly-explored area called epigenetics -- the impact of environment on genetic replication.

These changes themselves are unpredictable -- nobody knows where in the genome of which individuals the changes may occur, nor what the changes will produce.

However, evolution works across populations, so statistics apply. For example, in another thread we were talking about the parallel evolution of placental and marsupial dogs. If there were a gene for snout-length, then in a large population of dogs we could expect some dogs to have longer snouts and some shorter. If the dogs with longer snouts gain reproductive benefit (for example, they're better at tracking food), then we could expect the next generation of dogs to have longer snouts. So if we know the genome, the population size and the environmental pressures then we can predict certain mutations statistically.

But here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

This illusion of purpose in evolution can be compelling, even to someone professionally science-trained like me. I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Programs are first seeded 'randomly' using simple, blind functions like move, turn, shoot or kick. Scores are assessed based on the constraints of the game. A 'generation' is a population of random programs, and each new generation of programs is created by combining the most successful programs from previous generations. There is no human intervention in these tests, nor any guidance in how to fight or play soccer, but it really looks as though they're trying to learn.




That's just the weirdness that is evolution.

I agree that some mutations are accidents(and thus random). What I'm saying is that the mutations that drive evolution aren't random. How else to explain the studies that show predictable results in mutations with repeatable, specific conditions? How could the same mutation "accidentally" occur time and time again?

The same reason why you can have 10 children with downs syndrome, with the exact same mutation, despite being genetically unrelated.

Some mutations occur more easily due to the nature of the environment or the organism in question. We know of mechanisms that speed up mutation rates in bacteria, especially when placed under stress.
RuvDraba
Posts: 6,033
Add as Friend
Challenge to a Debate
Send a Message
4/2/2015 2:33:09 PM
Posted: 1 year ago
At 4/2/2015 12:35:54 PM, janesix wrote:
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

I agree that some mutations are accidents(and thus random). What I'm saying is that the mutations that drive evolution aren't random. How else to explain the studies that show predictable results in mutations with repeatable, specific conditions? How could the same mutation "accidentally" occur time and time again?

The benefits aren't random -- they're quite predictable. That's because only certain changes will be beneficial.

The kinds of mutations that can occur -- beneficial or not -- can also be modeled statistically. That's because replication works as a series of steps, and we can study each step and work out how any step can go wrong, and the chance of doing so.

On the other hand, if you run an actual experiment -- putting bacteria into a test-tube and then introducing selection or competition elements, say, or introducing lizards into a remote environment and watching how they evolve over decades (scientists actually do this), the results can surprise. [http://www.newscientist.com...]

Among the big surprises are that:
* populations can evolve much faster than expected; and
* how you set up the experiment can make a big difference to the outcome.

At 4/2/2015 12:42:32 PM, janesix wrote:
At 4/1/2015 6:32:33 PM, RuvDraba wrote:
"nor do additional mechanisms imply destiny or any other kind of supernatural intervention."
It could imply that there are mechanisms that are as yet undetected. No supernatural or "destiny" needed. Just natural laws.

That's exactly what's emerging, Jane. To be clear, reproductive mutation does occur, and accounts for organic changes, including new organs. However the changes sometimes go faster than expected, and in unexpected ways. I'm not a biologist, but here are some I know of:

* The Founder Effect: when you introduce a small population to a new environment, that population won't exactly resemble the whole species. So when it starts breeding with only itself, it can begin to produce a new species. That species may not have all the benefits of the parent species, but may also develop faster along new lines; [http://en.wikipedia.org...]

* Founder Mutation: In particular, any mutations already in the founders may become dominant in the population very quickly, and some may be beneficial. There are examples of lizards introduced onto a hostile island, growing new stomachs to become vegetarian in less than four decades. They also got bigger heads, a stronger bite and a bunch of other changes that otherwise might take millions of years to produce; [http://www.pnas.org...]

* Hybridisation: inter-species breeding occurs more than first realised, both in animals and especially in plants. It can represent another jump in adaptation, since it can blend the adaptations of two different gene-pools. Examples include cross-breeds between subspecies, like Bengal and Siberian tigers, and cross-breeds between species, like sheep and goats. Hybridisation can quickly sow the seeds of new species and create major adaptive changes, so sexual selection as well as ecological selection can play a big part in evolution. [http://www.newscientist.com...]

* Epigenetics: Only discovered in the 20th century, a vast range of environmental factors determine how much genes get switched 'on' in individuals. These switches can persist through cell division, which means they can influence the form, function and behaviour of the organism, and some effects can persist through reproduction too -- so they can be inheritable. So, suppose you could clone your own body and a surrogate mother agreed to carry that child. Who she was, where she lived and the foods she ate could greatly affect how that child looked, grew up and acted, so that the child may not look or act exactly like your identical twin. [http://www.newscientist.com...]

So in summary, Jane, some adaptations can be modeled effectively, but even inside a test-tube, how many, how fast, and how arrived at can surprise. Outside a test-tube in the natural world, it's very wild, and it may be that the adaptations of new species, and species introduced to new environments, are much faster than previously realised -- so much so that some biologists are now talking about the possibility of explosions in biodiversity from human activity -- though this still is not assured, and may not be all beneficial to humans if it does occur.

I hope that helps.
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/2/2015 2:35:42 PM
Posted: 1 year ago
At 4/2/2015 2:03:29 PM, tkubok wrote:
At 4/2/2015 12:35:54 PM, janesix wrote:
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
At 4/1/2015 3:37:00 PM, janesix wrote:
The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.

Jane, mutation has been observed. It occurs through ordinary biochemistry, radiation, hybridisation, and a newly-explored area called epigenetics -- the impact of environment on genetic replication.

These changes themselves are unpredictable -- nobody knows where in the genome of which individuals the changes may occur, nor what the changes will produce.

However, evolution works across populations, so statistics apply. For example, in another thread we were talking about the parallel evolution of placental and marsupial dogs. If there were a gene for snout-length, then in a large population of dogs we could expect some dogs to have longer snouts and some shorter. If the dogs with longer snouts gain reproductive benefit (for example, they're better at tracking food), then we could expect the next generation of dogs to have longer snouts. So if we know the genome, the population size and the environmental pressures then we can predict certain mutations statistically.

But here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

This illusion of purpose in evolution can be compelling, even to someone professionally science-trained like me. I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Programs are first seeded 'randomly' using simple, blind functions like move, turn, shoot or kick. Scores are assessed based on the constraints of the game. A 'generation' is a population of random programs, and each new generation of programs is created by combining the most successful programs from previous generations. There is no human intervention in these tests, nor any guidance in how to fight or play soccer, but it really looks as though they're trying to learn.




That's just the weirdness that is evolution.

I agree that some mutations are accidents(and thus random). What I'm saying is that the mutations that drive evolution aren't random. How else to explain the studies that show predictable results in mutations with repeatable, specific conditions? How could the same mutation "accidentally" occur time and time again?

The same reason why you can have 10 children with downs syndrome, with the exact same mutation, despite being genetically unrelated.

Some mutations occur more easily due to the nature of the environment or the organism in question. We know of mechanisms that speed up mutation rates in bacteria, especially when placed under stress.

"Some mutations occur more easily due to the nature of the environment or the organism in question"

How?
tkubok
Posts: 5,044
Add as Friend
Challenge to a Debate
Send a Message
4/2/2015 2:43:41 PM
Posted: 1 year ago
At 4/2/2015 2:35:42 PM, janesix wrote:
At 4/2/2015 2:03:29 PM, tkubok wrote:
At 4/2/2015 12:35:54 PM, janesix wrote:
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
At 4/1/2015 3:37:00 PM, janesix wrote:
The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.

Jane, mutation has been observed. It occurs through ordinary biochemistry, radiation, hybridisation, and a newly-explored area called epigenetics -- the impact of environment on genetic replication.

These changes themselves are unpredictable -- nobody knows where in the genome of which individuals the changes may occur, nor what the changes will produce.

However, evolution works across populations, so statistics apply. For example, in another thread we were talking about the parallel evolution of placental and marsupial dogs. If there were a gene for snout-length, then in a large population of dogs we could expect some dogs to have longer snouts and some shorter. If the dogs with longer snouts gain reproductive benefit (for example, they're better at tracking food), then we could expect the next generation of dogs to have longer snouts. So if we know the genome, the population size and the environmental pressures then we can predict certain mutations statistically.

But here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

This illusion of purpose in evolution can be compelling, even to someone professionally science-trained like me. I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Programs are first seeded 'randomly' using simple, blind functions like move, turn, shoot or kick. Scores are assessed based on the constraints of the game. A 'generation' is a population of random programs, and each new generation of programs is created by combining the most successful programs from previous generations. There is no human intervention in these tests, nor any guidance in how to fight or play soccer, but it really looks as though they're trying to learn.




That's just the weirdness that is evolution.

I agree that some mutations are accidents(and thus random). What I'm saying is that the mutations that drive evolution aren't random. How else to explain the studies that show predictable results in mutations with repeatable, specific conditions? How could the same mutation "accidentally" occur time and time again?

The same reason why you can have 10 children with downs syndrome, with the exact same mutation, despite being genetically unrelated.

Some mutations occur more easily due to the nature of the environment or the organism in question. We know of mechanisms that speed up mutation rates in bacteria, especially when placed under stress.




"Some mutations occur more easily due to the nature of the environment or the organism in question"

How?

For example, by being located in an area that is affected on by the increase in mutation rate.
janesix
Posts: 3,437
Add as Friend
Challenge to a Debate
Send a Message
4/2/2015 2:52:01 PM
Posted: 1 year ago
At 4/2/2015 2:43:41 PM, tkubok wrote:
At 4/2/2015 2:35:42 PM, janesix wrote:
At 4/2/2015 2:03:29 PM, tkubok wrote:
At 4/2/2015 12:35:54 PM, janesix wrote:
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
At 4/1/2015 3:37:00 PM, janesix wrote:
The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.

Jane, mutation has been observed. It occurs through ordinary biochemistry, radiation, hybridisation, and a newly-explored area called epigenetics -- the impact of environment on genetic replication.

These changes themselves are unpredictable -- nobody knows where in the genome of which individuals the changes may occur, nor what the changes will produce.

However, evolution works across populations, so statistics apply. For example, in another thread we were talking about the parallel evolution of placental and marsupial dogs. If there were a gene for snout-length, then in a large population of dogs we could expect some dogs to have longer snouts and some shorter. If the dogs with longer snouts gain reproductive benefit (for example, they're better at tracking food), then we could expect the next generation of dogs to have longer snouts. So if we know the genome, the population size and the environmental pressures then we can predict certain mutations statistically.

But here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

This illusion of purpose in evolution can be compelling, even to someone professionally science-trained like me. I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Programs are first seeded 'randomly' using simple, blind functions like move, turn, shoot or kick. Scores are assessed based on the constraints of the game. A 'generation' is a population of random programs, and each new generation of programs is created by combining the most successful programs from previous generations. There is no human intervention in these tests, nor any guidance in how to fight or play soccer, but it really looks as though they're trying to learn.




That's just the weirdness that is evolution.

I agree that some mutations are accidents(and thus random). What I'm saying is that the mutations that drive evolution aren't random. How else to explain the studies that show predictable results in mutations with repeatable, specific conditions? How could the same mutation "accidentally" occur time and time again?

The same reason why you can have 10 children with downs syndrome, with the exact same mutation, despite being genetically unrelated.

Some mutations occur more easily due to the nature of the environment or the organism in question. We know of mechanisms that speed up mutation rates in bacteria, especially when placed under stress.




"Some mutations occur more easily due to the nature of the environment or the organism in question"

How?

For example, by being located in an area that is affected on by the increase in mutation rate.

I'm sorry, I'm not trying to be obtuse. I really don't understand what you mean.
tkubok
Posts: 5,044
Add as Friend
Challenge to a Debate
Send a Message
4/2/2015 2:56:45 PM
Posted: 1 year ago
At 4/2/2015 2:52:01 PM, janesix wrote:
At 4/2/2015 2:43:41 PM, tkubok wrote:
At 4/2/2015 2:35:42 PM, janesix wrote:
At 4/2/2015 2:03:29 PM, tkubok wrote:
At 4/2/2015 12:35:54 PM, janesix wrote:
At 4/1/2015 4:23:31 PM, RuvDraba wrote:
At 4/1/2015 3:37:00 PM, janesix wrote:
The problem I see is that the evidence doesn't support random mutation as a means to evolve a new species. As I see from the article I posted, there was the same mutation 24 times in a row during the experiment. That leads me to believe that mutations aren't as random as they are proclaimed to be.

Jane, mutation has been observed. It occurs through ordinary biochemistry, radiation, hybridisation, and a newly-explored area called epigenetics -- the impact of environment on genetic replication.

These changes themselves are unpredictable -- nobody knows where in the genome of which individuals the changes may occur, nor what the changes will produce.

However, evolution works across populations, so statistics apply. For example, in another thread we were talking about the parallel evolution of placental and marsupial dogs. If there were a gene for snout-length, then in a large population of dogs we could expect some dogs to have longer snouts and some shorter. If the dogs with longer snouts gain reproductive benefit (for example, they're better at tracking food), then we could expect the next generation of dogs to have longer snouts. So if we know the genome, the population size and the environmental pressures then we can predict certain mutations statistically.

But here, we also encounter a sort of confirmation bias. If the only changes we observe are successful ones we were already looking for, then we will ignore all the other changes that were either unsuccessful, didn't manifest in the phenotype, or were neutral. So it can appear that evolution is purposeful because we were looking too narrowly in the first place.

This illusion of purpose in evolution can be compelling, even to someone professionally science-trained like me. I've linked these videos before, but take a look. The first one shows two robots 'learning' to fight; the second shows some amoebas 'learning' to play soccer together. Each experiment uses genetic programming.

Programs are first seeded 'randomly' using simple, blind functions like move, turn, shoot or kick. Scores are assessed based on the constraints of the game. A 'generation' is a population of random programs, and each new generation of programs is created by combining the most successful programs from previous generations. There is no human intervention in these tests, nor any guidance in how to fight or play soccer, but it really looks as though they're trying to learn.




That's just the weirdness that is evolution.

I agree that some mutations are accidents(and thus random). What I'm saying is that the mutations that drive evolution aren't random. How else to explain the studies that show predictable results in mutations with repeatable, specific conditions? How could the same mutation "accidentally" occur time and time again?

The same reason why you can have 10 children with downs syndrome, with the exact same mutation, despite being genetically unrelated.

Some mutations occur more easily due to the nature of the environment or the organism in question. We know of mechanisms that speed up mutation rates in bacteria, especially when placed under stress.




"Some mutations occur more easily due to the nature of the environment or the organism in question"

How?

For example, by being located in an area that is affected on by the increase in mutation rate.

I'm sorry, I'm not trying to be obtuse. I really don't understand what you mean.

I wrote before:

"We know of mechanisms that speed up mutation rates in bacteria, especially when placed under stress."

This increase in mutation doesnt just occur everywhere in the genome. That would be problematic. It occurs at specific points that are capable of handling such increased mutation rates, i.e. parts of the genome that arent necessarily essential.

The mutation itself is random, but an increase in mutation increases the likelihood, a lot, that similar mutations will follow.