Methinks it is sort-of like two weasels

[Rumraket (Mikkel R.)]

@lee_merrill

But no ID person thinks evolution probably cannot climb a hill where every step is selectable. A problem arises when several non-selectable steps present themselves, before fitness improves, then evolution bogs down, then the path becomes exponentially less probable, with the number of non-selectable steps.

If only there was some way of taking larger steps in sequence space than just substitutions, such as recombination, insertions and deletions, transposition, and internal duplications.

[colewd (Bill Cole)]

@Dan_Eastwood

The difficulty with “non-selectable steps” is that, as a hypothesis supporting design, it can never be falsified. The usual claim is that non-selectable steps can be falsified by demonstrating a selectable pathway. The difficulty is that a designer could also select this pathway, and so design is not falsified.

Hi Dan
If you could show enough selectable steps for the emergence of a gene/protein then design would be falsified as a direct explanation for that hypothesis.

[lee_merrill]

@John_Harshman

Usually, the people who talk about this problem fall victim to the Texas Sharpshooter fallacy.

Well, how so? I think Behe’s discussion of chloroquine resistance is sound, for instance.

[lee_merrill]

@Mercer

This is one of the reasons why ID people (a very strange label) avoid acknowledging both existing variation (the only thing Darwin knew about) and neutral evolution. They pretend that populations are genetically homogeneous and are “waiting” for new mutations to occur.

But the problem remains the same, if unselected steps happen in the past or in the future, the probability of getting through these steps gets exponentially smaller based on the number of steps.

[lee_merrill]

@Dan_Eastwood

The usual claim is that non-selectable steps can be falsified by demonstrating a selectable pathway. The difficulty is that a designer could also select this pathway, and so design is not falsified.

Thanks for the welcome! Now ID concedes that evolution can take a selectable pathway. Evolution can take a non-selectable pathway too, so evolution is never falsified, either. The question is then one of probability of natural processes generating a result, and the more unselectable steps there are, the lower the probability.

[Rumraket (Mikkel R.)]

@colewd

If you could show enough selectable steps for the emergence of a gene/protein then design would be falsified as a direct explanation for that hypothesis.

No it wouldn’t. “This doesn’t have a direct pathway through which it could evolve” is not a prediction of the hypothesis that “This was designed”.

ffs

[lee_merrill]

@Rumraket

If only there was some way of taking larger steps in sequence space than just substitutions, such as recombination, insertions and deletions, transposition, and internal duplications.

Yes, steps may include all these, and the question then is whether the steps are selectable.

[Dan_Eastwood]

@colewd

If you could show enough selectable steps for the emergence of a gene/protein then design would be falsified as a direct explanation for that hypothesis.

Not falsified. The designer could have selected those same steps.

@lee_merrill

Evolution can take a non-selectable pathway too, so evolution is never falsified, either.

Selectible Steps is not a hypothesis of evolution; it’s a hypothesis of design. Common Descent is a testable/falsifiable prediction of evolution. The discovery of Tiktaalik is a good example of an experiment in paleontology based on a prediction of evolution (that fossil with transitional feature should exist).

[Dan_Eastwood]

@lee_merrill

The question is then one of probability of natural processes generating a result, and the more unselectable steps there are, the lower the probability.

As a statistician, I find the probability arguments for design entirely flawed; there is no hypothesis to test, and the proposed calculations usually imply a hidden assumption that design is the only possible conclusion. That’s not good statistical inference.

[Rumraket (Mikkel R.)]

@lee_merrill

Yes, steps may include all these, and the question then is whether the steps are selectable.

I’m not aware of any real-world adaptation where there is any indication it couldn’t evolve. Are you?

[Mercer (John Mercer)]

@lee_merrill

Yes, steps may include all these, and the question then is whether the steps are selectable.

Again, you’re using a straw man version of evolution that ignores existing variation (your scenario of sequential steps) and neutral evolution.

[Mercer (John Mercer)]

@lee_merrill

…the probability of getting through these steps gets exponentially smaller based on the number of steps.

Please show your calculations. The big tell is your use of “exponentially” here; it doesn’t mean what you are using it to mean.

ID people tend to use mathy-sounding words without any math as they repeatedly invoke the Texas Sharpshooter fallacy.

So why is it that ID people routinely ignore existing variation and neutral evolution?

[colewd (Bill Cole)]

@Dan_Eastwood

Not falsified. The designer could have selected those same steps.

It is falsified as the direct cause. If for instance the origin of the ribosome is the result of design its action can be observed in terms of chemistry and physics. If design is true it is the ultimate cause of everything as you argue but it is often not the direct cause as we can explain certain phenomena with natural laws.

[lee_merrill]

@Rumraket

I’m not aware of any real-world adaptation where there is any indication it couldn’t evolve. Are you?

Behe points out the bacterial cilium and flagellum (which is not apparently descended from the T3SS), and the human blood-clotting cascade.

[lee_merrill]

@Dan_Eastwood

As a statistician, I find the probability arguments for design entirely flawed; there is no hypothesis to test, and the proposed calculations usually imply a hidden assumption that design is the only possible conclusion. That’s not good statistical inference.

No, the null hypothesis is that evolution did it, and then if the probability of this is low, then we reject the hypothesis.

[John_Harshman]

@lee_merrill

I think Behe’s discussion of chloroquine resistance is sound, for instance.

I would start correcting that by reading Larry Moran’s blog Sandwalk on the subject, beginning here:

The Edge of Evolution

In more general terms, the assumption is often that what we happen to see now is the only sequence that would work, and that only one path of mutations would lead from the prior condition to the current one, often assumed to happen only in a particular order. Behe further seems to assume that each mutation except the last, the one that suddenly confers function, is deleterious.

[lee_merrill]

@Dan_Eastwood

Selectible Steps is not a hypothesis of evolution; it’s a hypothesis of design.

Both evolution and design claim to be able to make it past non-selectable steps, though.

@Dan_Eastwood

Common Descent is a testable/falsifiable prediction of evolution. The discovery of Tiktaalik is a good example of an experiment in paleontology based on a prediction of evolution (that fossil with transitional feature should exist).

Actually, Tiktaalik is not a good example of descent, “but fortunately there are still the Zachelmie trackways discovered a few years ago (Niedźwiedzki et al. 2010) in Poland and dated as not only older than Tiktaalik , but actually older than all other lobe-finned ancestors of tetrapods (the so-called elpistostegids)” (Evolution News).

[RonSewell (RonS)]

@lee_merrill

Yes, steps may include all these, and the question then is whether the steps are selectable.

Could not the duplication just be drift, and later, a substitution be selectable?

[Rumraket (Mikkel R.)]

@lee_merrill

Behe points out the bacterial cilium and flagellum (which is not apparently descended from the T3SS), and the human blood-clotting cascade.

I’m sorry but I missed where you pointed out what it is about them that indicates they couldn’t evolve.

[lee_merrill]

@John_Harshman

I would start correcting that by reading Larry Moran’s blog Sandwalk on the subject…

Well, to quote Larry Moran: “Behe has correctly indentified [sic] an extremely improbably evolution event; namely, the development of chloroquine resistance in the malaria parasite. This is an event that is close to the edge of evolution, meaning that more complex events of this type are beyond the edge of evolution and cannot occur naturally.” (Sandwalk here)

@John_Harshman

In more general terms, the assumption is often that what we happen to see now is the only sequence that would work, and that only one path of mutations would lead from the prior condition to the current one, often assumed to happen only in a particular order. Behe further seems to assume that each mutation except the last, the one that suddenly confers function, is deleterious.

Multiple paths could work, though, Behe acknowledges this. "… an important paper by Summers et al. shed additional light on the development of chloroquine resistance. Although limited to experiments involving frog oocytes in the lab, this research provided solid experimental evidence detailing the specific mutations involved. The researchers identified two initial routes to chloroquine resistance, with additional mutations leading to ‘the attainment of full transport activity.’ " (Evolution news here)

And Behe uses measurements of the development of chloroquine resistance, he does not just assume here, the rate of about 1 in 10^20 indicates that two non-selectable mutations are involved.

“Behe for his part correctly noted that, if anything, the new research supported his primary argument. Indeed, one of the key takeaways of Summers et al. is that chloroquine resistance is a multi-mutational event, with both of the identified routes to resistance requiring ‘a minimum of two mutations’ to get started. Behe’s 2007 prediction that chloroquine resistance did not result from a series of individually beneficial mutations, but required a multi-mutational event, turned out to be correct.” (Evolution News here)