… with Mung. In a recent comment Mung asserted that
If Darwinists had to put up their hard earned money they would soon go broke and Darwinism would be long dead. I have a standing $10,000 challenge here at TSZ that no one has ever taken me up on.
Now, I don’t have $10,000 to bet on anything, but it is worth exploring what bet Mung was making. Perhaps a bet of a lower amount could be negotiated, so it is worth trying to figure out what the issue was.
Mung’s original challenge will be found here. It was in a thread in which I had proposed a bet of $100 that a Weasel program would do much better than random sampling. When people there started talking about whether enough money could be found to take Mung up on the bet, they assumed that it was a simple raising of the stake for my bet. But Mung said here:
You want to wager over something that was never in dispute?
Why not offer a meaningful wager?
So apparently Mung was offering a bet on something else.
I think I have a little insight on what was the “meaningful wager”, or at least on what issue. It would lead us to a rather extraordinary bet. Let me explain below the fold …
Mung accepted that Weasel programs reach their goal far faster than random sampling. However Mung also said (here) that
Weasel programs perform better than blind search because they are guided. I didn’t think the performance was in dispute, nor why the performance was better.
and elsewhere characterized Weasels as succeeding because of “intelligence” as opposed to ignorance.
So let’s imagine what might happen if we took Mung up on the $10,000 bet. We would bet that the Weasel would succeed, because of cumulative selection. Mung would bet that (because of “intelligence” or being “guided”) the Weasel would succeed. The stake would be held by a house of some sort, which would not take a commission.
The Weasel would be run. It would succeed. So the house would declare that we had all won. The stake would be given to the bettors, in proportion to their bets. But alas, no one actually bet against the Weasel. So the winnings would be zero. Everyone, Mung and the rest of us, would get their stake back, and that’s all.
To bet against Mung, we have to come up with some event that distinguishes cumulative selection from “intelligence” (or being “guided”). That seems to be the issue on which Mung was offering a $10,000 bet, and declaring (here) us all to be “pretender[s]” because we would not put up or shut up.
So there it is. We’re all betting on the same side, and no one will win or lose a penny. Unless Mung can come up with some test that distinguished “intelligence” or being “guided” from cumulative selection.
Now I am possibly misunderstanding what the bet actually would be. I hope that Mung will straighten us out on that, so that we can understand what test is proposed, and place our bets.
Which is more adapted to getting eaten by a bear? Which is more adapted to getting thrown into a volcano? Which is more adapted to my living room?
You see this is where you and Allan want to change your definition of fitness ONCE AGAIN (fuck off Allan). Does fitness mean which one can reproduce best if I decide the conditions which I will allow it to reproduce? If I put the fish and the dog inside a volcano, then neither can reproduce, so how do you know which is more fit?
Furthermore, if the dog and the fish both don’t reproduce, how do you know the fish is more adapted to life in water, and the dog is more adapted to life on land, maybe the dog didn’t find a mate and neither did the fish, so neither is adapted.
Yeah, but you have a tendency to not see very well, so the reason people parody him is likely lost on you.
Like when he claims that biomorphs have any relation whatsoever to evolution. Or when he says Nilsson-Pelger have modeled eye evolution. Or when he says, the best evidence we have shows there is no God.
I can come up with 150 other ridiculous things he said, if you give me about 2 minutes to read him.
And how is that, can you re-state it in your own words please?
What do “people” think about it?
Evolution as an explantion for life’s diversity does not depend on us having explained the origin of life.
But it is true that evolution as a real-world phenomenon depends on life existing, which in turn means at some point it had to originate.
I would agree in the sense that you can’t really say you have a full account of the phenomenon of life, if you don’t have an account for it’s origin. But you can still explain it’s diversification and many of the properties of living organisms that owe to the evolutionary process.
I don’t think anything is really adapted to any of those things as there has not been a process of differential reproductive success whereby some population of organisms increased their reproductive success by getting eaten by bears or thrown into volcanoes.
No it isn’t.
You can have your own personal definition of fitness if you like, it just won’t correspond the process by which populations of organisms adapt to their environments.
It is entirely possible that there is some environment (like inside a volcano) where their relative fitness is exactly equal. This isn’t an argument against, or any sort of problem with the concept of fitness in biology.
I am in awe of your continued inability to get this.
Well if I’m ignorant about their relative reproductive success in some given environment, then that is so, and then I don’t know which is more fit.
When you define the situation such that I don’t know, then by definition I don’t know. Obviously.
Holy hell. Once again, I am so glad you are here to inform me of these penetrating insights.
That would be the case inside a volcano, yes. I don’t think life as we know it even can adapt to life inside a volcano, as all the relevant chemistry doesn’t work inside volcanoes, and the very molecules of which living organism are constituted, break down in microseconds at those pressures and temperatures.
What a truly fantastic and paradigm shifting revelation.
I’m just back to the thread after an interlude of real life. Although much has been said since, it may be worth going back to the issue of “cumulative” genetic drift and cumulative selection:
It is cumulative, but we can imagine genetic drift that isn’t.
In fact, in teaching my courses, I used to raise just such a distinction when I tried to clarify the concept of genetic drift. I analogized one generation of genetic drift to tossing a coin
times (for a population of 
diploid individuals). The probability of Heads would be the gene frequency 
in the parents. Then I would ask why, if we tossed 
times each generation, the probability of Heads did not converge in the long run to 
, as it would with ordinary coin tossing.
If we continued to toss with Heads probability
in each generation, it would converge to that. But we don’t, because each generation starts from the outcome of the previous generation. The gene frequency keeps changing, until it ends up at 0 or at 1.
So the convergence to
is for non-cumulative genetic drift, and the ultimate convergence to 0 or to 1 when we just have reproduction with genetic drift, is “cumulative” genetic drift, which is the ordinary kind.
There has been much sturm und drang here about how to have a Weasel in which selection is not cumulative. Most of the suggestions have been to have the target string change every so often. In between those changes, selection would be cumulative and there would be some progress toward that most recent target string.
Actually, far upthread, I suggested a simpler way to have selection not be cumulative:
This in fact would be what Dawkins called “single step selection”. We start each generation afresh from the same genotype, or we start each from a random genotype that has nothing to do with the outcome of selection in the previous generation.
We could also change the target every generation to a randomly drawn string. That way any change in the previous generation will have carried us no closer to the new target.
I suspect that Mung will not be willing to bet that these cases will do as well as cumulative selection, which is where the target does not change and each generation is the outcome of the same selection in the previous generation.
Rumraket,
If the definition of fitness involves defining the exact environment in which something exists, then all one has to do is define the environment differently and all of the determinations of fitness completely change. Allan can claim the new bacteria in lenskis are more fit, only because he gets to decide what they are more fit for? What if we decide to blast them with a blowtorch, which are most fit? What if we bath them in alcohol?
If I have a dog on dry land and a fish in water, but I decide I am going to shot the dog with a rifle, then the fish is more fit. The fish must have lucky genes.
If I decide to raid a village and kill all the male children, then being female is more fit.
So this idea that you can measure fitness is nonsense, because why do you get to decide what the environment you are going to measure is going to be like? The bacteria in lenskis flasks are equally fit, then someone decides they are going to poison some with their choice of poison, or starve them, like by giving them nylonese for food.
So you can’t say which gazelle is most fit, until I get to say, well, I am going to put this one in 3 foot of water next to a crocodile, and this one I am going to feed it rat poison, and this third one is going to be put in a zoo with other gazelles and be given grains to eat. Ah ha, the third gazelle is most fit!
There is no static environment, there are never equal conditions for all living things, so any definition you give for fitness is just made up. If you happen to go to college with a girl you like and you marry her, but she does not want to have kids, but some hick in West Virginia marries a girl with one leg who has 17 babies, then you are forced to say that the couple in West Virginia is more fit than you and your college sweetheart.
But there is no science to that, it is just you saying, I chose this as fit.
So no, you can’t measure fitness, not in a lab, and not in the wild. Its completely subjective depending on who is making the conditions.
In the context of fitness as differential survival of alleles, yes. If you die without producing offspring, your particular allele set, with any unique advantageous new variation that you might possess, disappears with you.
Alan Fox,
So then we can’t measure fitness. Something either exists or it doesn’t, because there is no way to know beforehand what the environment is. Maybe after you attempt to measure it, someone shots and kills what you just measured, now it can’t reproduce.
I think you can. You can observe and record numbers of offspring in populations. It’s certainly easier to observe in smaller organisms with short generation times, bacteria, for example.
So far as I understand you, I don’t agree. You can set up a controlled experiment. Lenski’s classical and ongoing experiment with E. coli for instance.
It’s statistically biased sampling. Many individual chancy events stack up to leave the next generation richer in beneficial alleles.
phoodoo,
Joe G used to say something similar, before he was banned.
Sure, it’s possible a cosmic ray could strike the single bacteria with a mutation that makes it outcompete all its relatives, and yet despite that this still happens: https://www.youtube.com/watch?v=plVk4NVIUh8
So it seems you simply are thinking at the wrong level. Sure, the fastest gazelle might trip and fall and break it’s neck. So what? The next fastest one is now the fittest, if that’s how you are measuring fitness at that moment. You have to think about populations rather then individuals.
And if a population gets “shot and killed” by, say, a meteor, then another population may be able to expand into the now free niche. Or not.
In measuring fitnesses, we use multiple individuals (large numbers if we can) and look at all individuals of a given genotype. So one individual accidentally dying does not make the fitness pf the genotype zero. For viability, the ones and zeros for individuals get averaged across all individuals of their genotypes, so yes, we can have viabilities of with values like 0.892.
And, as usual, phoodoo ignores fertility. If population members of a given genotype have probability 0.892 of surviving to adulthood, and each of the survivors then has on average 2.4 offspring, the fitness of that genotype is the product of these numbers.
This has been explained to phoodoo multiple times, but to no avail.
I didn’t know you were from West Virginia.
Joe Felsenstein,
No Joe, we are talking about individuals NOT genotypes. You have no justification for trying to change the playing field to genotypes. The Weasel program doesn’t measure genotypes, it measures strings, each string. If it is closer to the target than another string, then you call it fit.
This has been explained to you many times. Genotypes don’t get mutations, individuals do. Furthermore, there is no such thing as one genotype, because every individual has a slightly different one. So you can’t claim part of an individual is fit, and part isn’t.
Again, this has been explained to you many times.
No, the definition of fitness is always the same: relative reproductive success. The measure of fitness depends on the environment. Just like measuring the speed of a runner, depends on the conditions where they run.
The fact that their performance can vary on different surfaces or wind conditions, for example, does not mean there is no such thing as a running performance, or that we cannot compare the velocities different runners can obtain.
I suspect there is no set of conditions that exist where I would outperform Usain Bolt. And even if there were, then that would just mean for that obscure circumstance, I would be a better runner than Usain bolt.
In the same way, the concept of fitness isn’t rendered nonsensical just because the relative reproductive success of carrier so different alleles depends on the environment.
No, their performance might chance, but the measure is still relative reproductive success.
Just like velocity measures distance traveled in a set amount of time.
No, because in the environment where they are tested, they ARE more fit.
Why are polar bears white and brown and black bears brown and black? Could this have something to do with their ability to hunt? Is it a stretch to imagine that he polar bear fares better in the arctic, than brown or black bears would? Then the polar bear has higher fitness in the arctic. It is better adapted to life in arctic conditions.
Then they both die, and they are equally unfit.
If you blow the legs off the lineup of olympic runners, then they all get nowhere. That doesn’t mean that before you did this, one of them wasn’t faster than the others.
If you systematically decide to kill dogs yet let fish live, and this has a measurable impact on dog populations, then yes, whatever genes make dogs be dogs which will cause you to shoot them, are associated with lower reproductive success.
You have yet think of even a single thing that is an actual problem with the concept of fitness in biology.
No because females (at least human females) can’t reproduce naturally without males. So their relative reproductive success wouldn’t change.
But you CAN measure fitness, and you can do it in different environments.
Just like you can measure velocity of different runners under different weather and surface conditions, and get actual measures of their performance.
So no, it’s not at all nonsensical just because the value you get depends on the conditions you use to measure.
No, they simply aren’t. They show rather large differences in fitness performance under different testing conditions.
Stop saying stuff that is trivially demonstrably false.
And then their performance would change, just like if you poison or starve runners, their performance changes. That doesn’t mean they don’t actually have different performance levels. And we can measure them.
Yes I can, I just have to find an environment and test it. And then we could move to a different environment and test it there too, and their relative performance might change.
Yes, yes it would be. Thank you for demonstrating that we could still measure their relative reproductive success under those circumstances, and that you like me, find it obvious that third Gazelle in your exmaple would be more fit that the two others.
Environments aren’t static, but their rate of change is usually very slow. It takes a very long time for a river to carve a different path in the landscape. It takes a very long time for mountains to form and cliffs to erode. For forests to disappear and new ones to grow tall.
It’s true they’re not truly static, but they’re also not impossibly erratic. The arctic doesn’t become a dense jungle by summer time, there is usually ice there all year round.
So to practically measure fitness in some environment, it just means we have to measure the relative performance of different genotypes by counting lots and lots deaths and births and deriving an average.
They very well might be. That isn’t a problem with the concept of fitness.
But there is science in that, it is direct observation.
Of course I can, as I can just measure relative reproductive success.
No, all organisms will objectively have some particular performance in some particular environment, and we can measure it.
Wrong phoodoo, individuals have genotypes, and those genotypes might affect their reproductive success. It could be, for example, that some men like girls with brown eyes and black hair.
Or some mice with a dark color genotype might be harder to spot for cats, so avoid predators on average more often.
I was one of the people who suggested a Weasel with a randomly changing target string, but I don’t believe this would make the selection process non-cumulative. It was merely meant to reflect a dynamic fitness landscape as might exist in nature, with a changing environment.
phoodoo,
Yes, each individual has a slightly different genotype in most populations. But when we study them, we usually concentrate on one locus, or on a small set of loci.
It is also possible to measure phenotypes, such as body size, and see how fitness depends on that.
Lots of evolutionary biologists are doing these kinds of studies. There are two possibilities:
1. phoodoo has seen a logical problem that has been overlooked by tens of thousands of evolutionary biologists. Once this is realized, it is inevitable that phoodoo will be widely acclaimed and showered with honors. Or,
2. phoodoo doesn’t understand how evolutionary biologists measure fitness.
We will each have to make our choice among these two possibilities. I think I know which I choose.
Again Joe, I understand perfectly why you want to change the problem from identifying fit individuals to identifying traits which you think are fit. So then you can just say, well, more individuals are big, I guess big is fit. More individuals are small, I guess small is also fit. That’s a cope out Joe, as has been explained to you many times.
Individuals survive, individuals reproduce, not genotypes. Weasel also doesn’t deal with genotypes. Please pay attention.
Rumraket,
Wrong Rumraket.
You have an isolated population of 500 dogs, and its March 13th, 2018 at 5:00 pm. Of those 500 dogs, how do you determine which is most fit and which is least fit?
phoodoo,
Wait a generation?
OMagain,
We already did, that’s how we got to 500, now what?
Wait another one and count how many offspring individual dogs manage to produce.
The same way I would if it was 500 runners and the question was which one is the fastest over some distance: Let them run and measure it.
Let the dogs reproduce and count how many offspring the different dogs have.
Well, we got some problems Rumraket.
First, Joe claimed that we can check things like size and determine fitness with such variables. So we throw out that idea.
Next, does it matter if the dogs offspring can produce offspring? What if all of some of the dogs offspring are infertile, is the dog that has the most dogs still the most fit?
And when is the generation over, so we can decide which dog is fit and which isn’t? Do we wait 1 year, 5 years, 10 years? When do we do the count?
And then one dog might live only one year and then die as its giving birth to five dogs, but another dogs lives to be 38 years old, but only gives birth to four dogs, so it is less fit than the dog that died at 1?
And what if some of the dogs are kept in cages, and not allowed to mate, does this mean they are not as fit as the 1 year old that died?
And ten dogs get struck by lightning, so they were unfit?
And some dogs are runts with mange and tapeworms, and only have three legs, and a deformed cleft palate, but the manage to impregnate another dog that gives birth to a bunch of dogs with cleft palates and three legs, but those dogs are all more fit than the ones struck by lighting, because the dog with three legs couldn’t go outside to look for food, but the dogs that got struck by lightning could?
So when does the counting start? After they all die?
So the moral of the story is, if you have a population of animals and you want to know which ones are fit, wait until they are dead, then you can know if they were fit? Even if their bloodline completely dies off in the next generation from illness?
Dawkins isn’t trying to explain life’s diversity. Even creationists accept life’s diversity.
Focus Rumraket.
Yes he is. The blind watchmaker is evolution.
They accept it exits, they don’t accept the only scientific explanation for it there is.
So we take a string and compare it to a target string and then we assign it a fitness score based on how closely it resembles the target string. We do this for every member of the population. Then we take the one with the highest fitness score, kill off all the rest, and then make copies of the one with the highest fitness score.
So it’s relative reproductive success is higher and how could it not be? What is it’s fitness? 1.0? What is the fitness of the others? 0.0?
And this is what we mean when we say “cumulative selection”?
PRATT.
If you have killed off all the other dogs, what is the point of counting the number of offspring of the one remaining dog?
You kill off 499 of them and hope the remaining one will reproduce.
No, I think you do.
Yes, we can measure the contribution of individual phenotypes on fitness. That is correct.
No, we don’t.
That’s how fitness is practically measured by biologists, by counting offspring over multiple consecutive generations, so yes.
That depends on the fraction of the offspring that are infertile, obviously.
Is it really the case that you don’t know what a generation is?
That depends on the generation time: https://en.wikipedia.org/wiki/Generation_time
“Definitions and corresponding formulas
The existing definitions of the generation time fall into two categories: those that treat the generation time as a renewal time of the population, and those that focus on the distance between individuals of one generation and the next. “
Which is why in reality, you measure over multiple successive generations to average out the effects of individual life histories.
In the grand scheme of things yes.
If ten runners are struck by lightning, are they slow runners?
I have to say your questions are stadily becoming increasingly stupid. I’m tired of interacting with someone who becomes a child on a particular subject. Go out and play with the other kids and let the adults speak for a while.
In think we have interacted enough on this subject already as I have comprehensively shown the concept of fitness is both sensible and measurable in the real world, just like the performance of runners are. Any rational person who comes to this thread to read this exchange between us will see that. I’m done, thank you for playing, time for you to act like a child again.
No that’s just a fact.
Who the hell says we kill off 499 of the dogs? Why would we do that?
Mung,
Stephen Hawking -1
Novak Djokovic-0
Cristiano Ronaldo-0
LeBron James-0
Caroline Wozniaki-0
Peter Dinklage-1
Kobe Bryant-0
Ronda Rousey-0
Myrtle Corbin-1
I don’t think this fitness thing is very useful.
I think Allan also suggested it. But you’ve already seen the flaw in that analogy, as organisms don’t take on the characters of their environment. In Weasel, the candidate solutions become more and more like the target phrase.
Part of the issue here is the way Dawkins framed the problem. The word cumulative doesn’t modify the word selection, and the phrase single-step” doesn’t modify the word selection.
You have selection, and you have accumulation. Dawkins tries to make it appear as if you can have selection without having selection and Joe makes the same mistake.
When phoodoo wrote that there is no genotype in Weasel my initial reaction was that he is wrong. But I can actually see how he could be right about that.
What is it, exactly, that the target phrase represents? It is supposed to represent some complex feature, like perhaps an eye. It certainly doesn’t represent an organism.
That would be like taking all the runners, gluing them to the ground and only letting one guy run. Then you come and declare that the concept of running performance is meaningless because you have artificially constructed a situation where you can’t meaningfully compare the the individual runner’s performances to each other.
Well I guess nobody every runs faster than others and the olympics is all staged, or something. More and more people are boycotting the olympics and I have this growing list of professional athletes signing this statement that they doubt the veracity of running. Only through censorship at athletic institutions by the running-lobby are alternative types of locomotion-theories kept out.
Or.. something.
LoL. By analogy. Because that’s the way it is done in Weasel. In Weasel you pick one member of the population to be the sole survivor and kill off all the others.
If you are going to argue with yourself, what’s the fun for me?
Were you there? No? Then you can’t prove there weren’t invisible cyclists carrying runners in the past.
Even young earth creationists accept common descent and speciation. They do not believe that all extant species were on the ark. Your claim remains a PRATT.
Once again, for the fifth time at least, nobody has claimed the Weasel was an accurate reflection of all aspects of biological evolution as they take place in the wild.
Once again, it was merely meant to demonstrate the performance of cumulative selection as compared to random guessing without selection (what dawkins calls single-step selection).
We could also alter the weasel so instead of just making a sole survivor every generation, we had that 500 population and killed off 20 random individuals among the bottom (with respect to the degree of match to the target string) half of the population.
And it’d still massively outperform random guessing, demonstrating that with cumulative selection evolution isn’t at all like the random guessing that creationist debaters try to portray it as.
They accept limited descent within what they call “kinds”. They think there’s some mystical magical barrier that prevents diversification beyond certain limitations they are unable to actually define or identify.
So no, they don’t accept evolution as the explanation for life’s diversity. They think life’s diversity was primarily seeded by God with a large number of biblical “kinds”, and that these then (ironically) micro-evolved at hyper-speed into the full diversity of life we see today in about 4000 years (their timing of the purported great flood).
Yes, having a long range target which can be aimed for over many generations is essential to having cumulative selection. You win the prize Joe.
Please explain how you have not just admitted to this.
Gee, what a surprise. You tell a computer what to look for and when it does you keep it, and it does much better than if you don’t tell it what to look for.
The miracle of evolution: an analogy. (Ok, not a perfect analogy. Not an accurate analogy. But still.)
So? They accept the “scientific explanation” for life’s diversity. You claimed they did not. Your claim was false. So now you shift the goalposts to universal common ancestry. If that’s what you meant to say originally that’s what you should have said.
As it is, what you actually did say, is false. And a PRATT.
Weasel is actually an argument for ID, lol. That may deserve it’s own OP.
So we agree that Dawkins’ argument depends on having selection without selection.
Selection without selection would have to be one of them.
That should make for a great thread…