Lenski experiment and falsifiability

(Daniel Fisher) #101

No, that has nothing to do with what I was saying. If there are, in fact, 50,000,000 variations between humans and chimps, then it is simply a fact that there are also 3^50,000,000 possible variations of how that evolutionary process could alternately have gone (and that is if we are limiting our consideration only to those particular 50,000,000 bases)

That is simply an inescapable observation, and in itself it cannot in any way be used to dispute common descent, nor was I or am I so doing. If there are, in fact, 50mil differences b/w chimps and humans, both descended from HCLCA, then those variations had to happen somewhere. The odds of a chimp having any one particular outcome of those possible 3^50mil variations has nothing to do with whether or not they descended from HCLCA.

(Matthew Pevarnik) #102

You’ve again reduced the problem despite everything everyone said to a very specific combination of mutations. In fact 50 million in particular mutations of which all need to be exactly specific to get from a common ancestor to the present difference between humans and chimps.

(Daniel Fisher) #103

It is simply easier to do a simple thought experiment dealing with a (relatively) smaller number. I thought that much at least was clear… That’s why I said I would “vastly, vastly oversimplify” in order to “limit my mental experiment.”

If we want to consider effective possible mutations of 50,000,000 mutations across the whole 3 billion base pair, my point is still identical. If you’d like me to spell it out…

  • If there are, say, 314,625,000,000,000,000,000,000,000 possible variations between HCLCA and humans given 50,000,000 variations across 3 billion base pair…

  • And of those, if, say, 314,625,999,999,999,999,999,999,900 possible variations that lead to intelligence… a vast, vast, vastly vast, very vast, supremely vast number…

  • This is still meaningless unless we know how many possible variations there are. If there were in fact 314,625,000,000,000,000,000,000,000 possible variations, it is still the case that for every one path that leads to intelligence, there are still 5.1 x 1047 paths that don’t, and I am still very dubious that the unguided process could accomplish that.

  • If, however, we could somehow determine that there are something like 314,625,999,999,999,999,999,999,990 paths that led to intelligence of the 314,625,000,000,000,000,000,000,000 options, then were are still talking about 1 path leading to intelligence for every merely 59,000 that don’t. And if that were the case, then I would not find the process particularly problematic. Those sorts of possibilities would not seem insurmountable for unguided natural,selection to find that path.

(Daniel Fisher) #104

Your math is backwards, I’m afraid. Ever base pair has three different (new) combinations. If there were only two bases, for instance, the possible combinations are 9 (3^2). If three, there are 27 possibilities (3^3).

So possible (new) combinations is not 60^3, but 3^60, or 4.2 x 10^28.

(Daniel Fisher) #105

I’m inviting you over to play poker, since you won’t notice anything odd if I deal myself 4 royal flushes in a row. After all, those hands are just as likely as any other combination of cards…!


(Steve Schaffner) #106

Quite right. Post in haste, repent at leisure.

(Steve Schaffner) #107

I don’t think I’d want to play poker by your rules, though – you’d deal yourself a 2 of clubs, 4 of hearts, 7 of spades, a jack and king of diamonds, and declare yourself the winner because the probability of your hand was so low.

(Steve Schaffner) #108

Yeah, if we could determine some things we could say something meaningful about the probabilities, but we can’t.

(Daniel Fisher) #109

All the more reason it seems better to withhold judgment one way or the other and keep an open mind both ways, no?

if it is true that we genuinely have no earthly idea how likely this process is in being able to achieve innovation, ought we not withhold judgment, no? - it could be 1/50,000; it could be 1 in a googol. The former is realistic given timeframes and numbers of generations and population size and as assisted by natural selection, the latter would be high fantasy… if we genuinely have no idea and ballpark numbers could be either, and no way of knowing at this point, wisdom and scientific method suggests to me we withhold judgment about the power of evolution and keep a genuinely open mind on the subject. If these kinds of numbers are genuinely a complete unknown, then I ought to be as dubious about the confident claims of ID folks that this could not happen as I ought to be about the claims of EC folks who claim it could happen. There just isn’t enough data yet to make such a judgment.

But at the moment, I feel I’m being asked to trust this evolutionary process (by faith?), and essentially told, “pay no attention to that math behind the curtain!”


(Daniel Fisher) #110

Certainly every outcome is equally improbable. But for our purposes, Once you apply some kind of criteria, whatever it is, and you see the same thing happening defying odds, you start getting suspicious.

(If for instance you noticed that i repeatedly dealt myself a 2, 4, 7, J, K every single time I dealt, you ought to be equally suspicious that something funny was going on.)

So yes, if I simply examine whatever “pops out” of the evolutionary mechanism, with no further criteria, and go, “wow, how improbable!” Then yes, absolutely, that is completely fallacious. Every outcome is theoretically as probable (or improbable) as any other.

But when you start to apply some kind of criteria to see if, in fact, someone is stacking the deck, and the same outcome keeps repeating…

If, for instance, we could hypothetically determine that for every one evolutionary path that led to genuine innovation there were a googol paths by which an organism could evolve and adapt while only maintaining or deteriorating its current level of complexity… and we saw this keep happening over and over and over across the history of life… if those were the hypothetical numbers, at that point, it would be equally fallacious to say, “well, the animals had to evolve in some direction, this is simply the direction they went,” without recognizing that when a googol to one circumstance keeps repeating, then someone is stacking the deck.

(Tim) #111

Do we have to just view evolution as a means to get from point A to point B?

Why not just view it as a means to thrive and exist in the here and now? Adaptation is not a sin, to be shunned. We look to ID, common decent, but what about, having simularities just because we all exist in the same environment? Some animals even share environments. We have the basic three environments: air, land, and water. Then the three, have a variety of temperature differentials. There is plenty of diversity between these variables. The fine tuning in physics should be an indication of fine tuning in genetics, whereby biological entities can all exist in a somewhat narrow ecosystem without drastic changes being constantly experienced. I realize this is an ongoing debate. My point is not a why/because statement. This is not about moving from point A to B.

Since we can observe there seems to be a narrow environment whether on purpose or not, compared to the whole universe, we are stuck with the observation. Otherwise our view of even evolution itself would be totally different. There would be no simularities at all. Every single biological entity would have it’s own separate DNA, not from a single source, but so different, we would have to figure out how they all exist in the narrow window called life. Fine-tuning does not disprove evolution. Nor do we really have any observable way to see life any other way.

(Stephen Matheson) #112

And then you might think “I wonder what explains this,” and then you might read Darwin 1859 or perhaps Dawkins 1986, and think, “ah, yes, that makes sense.”

(Matthew Pevarnik) opened #114

(Steve Schaffner) #115

No. We cannot calculate from first principles the probability of, say, a novel functional protein arising. There are a great many outcomes in lots of areas that we can’t calculate from first principles. We can’t calculate the probability that a protein will fold into a particular three-dimensional structure (although we’re slowly getting better at that), or when a major earthquake will occur. Ought we therefore to withhold judgment about whether moving tectonic plates cause earthquakes and whether known mechanical and chemical processes cause proteins to fold? Of course not.

When we can’t calculate probabilities from first principles, we look for explanations that are consistent with the data. What biological data tell us is that new species descend from existing species, that the genetic differences between species look like the result of mutational processes we understand, that new genes look exactly like they have arisen from existing DNA (DNA in transposons, or in existing genes, or in noncoding sequence), and that the changes all look like ones that could be explained by known mutational processes. These observations are entirely consistent with what we would expect to see if biological diversity results from evolution, and can often be predicted in advance.

As an alternative to this, we have intelligent design. ID provides no mechanism for changes to species, no evidence for the presence or activity of the intelligent designer, no description of what exactly happened or when or where or how in the formation of species, and no predictions about what we should see in new data.

Given these choices, no, we should not withhold judgment.

But there are no odds to defy – those are precisely the odds we’ve just been discussing.

(Roger A. Sawtelle) #116

There are several problems with the way you pose the question. First is factual. You assume that the development of a new species is the only test of evolution. The Lenski experiment has shown some interesting changes, so it shows real evidence of evolution.

It has not evolved into a new species because 1) E. coli is a simple life form which cannot change that much, and 2) to create a new species requires a separate novel niche which is not a part of the experiment. There are a number of strains of e. coli which are difference but do not qualify as different species.

(Daniel Fisher) #117

It has shown many interesting micro evolutionary changes, the kind which no one (not even Ken Ham) disputes.

But also, to my understanding, we have not seen E. coli evolve into a new species across the globe outside the experiment over the last 50+(?) years since we’ve been able to observe them?

So, it essentially keeps coming back to that superhero in Mystery Men… who can turn invisible, so long as no one is watching. Macroevolution only happens in those cases where it would be impossible for us to observe it.

Higher species take too long to generate for us to observe such evolution. Species that have extraordinarily short generation times? no, we can’t see evolution there either, because those organisms take far longer to evolve. And even if we could observe them over long enough periods, there is already an excuse ready (they haven’t found the proper niche) to explain why we don’t see evolution happening at the pace it is understood to have happened in the past.

It is all a bit too convenient for my skeptical mind, I’m afraid.

So it all essentially confirms the question I asked at the outset… the macroevolutionary mechanism is believed, embraced, and assumed, but neither witnessed, observable, repeatable, testable, nor falsifiable. It only happens when we’re not watching.

(Marvin Adams) #118

I find it amusing that the same setup is used on the other side of the pond to demonstrate evolution at an industrial scale, Evotec being one of the spinoffs of the process
Looking in disbelief at your question for not knowing about Manfred Eigen I was very sad to find that he died in February, what a loss to science.
Die Evolutionsmaschine was debated in the eighties and the 1992 working model is in the German Museum so if anything, it proves the opposite.

(Mitchell W McKain) #119

So what about evolutionary experiments with eukaryotic organisms? This has been done with great success using aphids and drosophilia, with adaptations to new food sources or to low oxygen environments.

Testability let alone falsifiability is hardly limited to reproducibility. Any predictions that come from a theory provides a means of testing and falsifying the theory. This includes things we would expect to find in the fossil record or in the genetics of both living organisms and in the remains of those long dead. This has produced and endless number of tests of evolutionary theory as well as opportunities of falsifying the theory. Instead all tests have confirmed the predictions of evolution 100%.

(Daniel Fisher) #120

Fossils and such genetics could at best only confirm the theory of common descent. They cannot confirm the mechanism of unguided macroevolution.

(Stephen Matheson) #121

Comments like this make you look dishonest. The more likely explanation is that you don’t understand the topic you claim to be skeptical about. But if you actually think that you have to be able to observe something to believe it as an explanation, then you just aren’t credible as a skeptic.

In related news, I’ve never seen a mountain erode away, and neither has anyone else.