Sunday, January 24, 2010

Has a supporter of evolution decreed that it is false?

Pvblivs,

I forgot this bit:

(In what follows, "it" refers to "universal common descent")

"Lateral tranfers of information (which occur in bacteria) are consistent with it."

Nope, Lateral gene transfer is NOT consistent with it.

[Name withheld]


     Well, Wikipedia has an article on lateral gene transfer. Lateral gene transfers are a known reality. This poster says they are inconsistent with universal common descent. Logically, that would mean that universal common descent is false.
     In all likelihood, he probably meant to claim that lateral gene transfer among macroscopic life forms is inconsistent with universal common descent. Of course, that's not true either. That is just taking something we believe not to occur and declaring it a "potential falsifier" for that which is intrinsicly untestable.

16 comments:

Anonymous said...

Ups, should have answered here.

Nope, Evolution is not false. Universal Common Descent might be.

G.E.

Anonymous said...

Hello Pvblivs,

Answering here because of the quote-mining that Ray might do (which might happen anyway):

Yes, I am inclined to the side that there is no such thing as Universal Common Descent (UCD).

Let me explain a bit:

1. Lateral Gene Transfer (LTG) happens in Bacteria and Archaea (similar to Bacteria, but some important differences in their central genetic system).

2. UCD would mean we all, living forms, descend from a single common ancestor.

3. The problem of LGT as a falsifier, is determining how much of it happens, and whether it makes sense to think of a single cellular line at the bottom, despite, at least for Bacteria and Archaea, there is a probability that not a single gene in the present can be traced back to such a single cell (I do not think that if no gene goes back to a single cell, UCD would make sense).

4. Some math shows that even a modest amount of LGT might break the whole thing of UCD (genetically speaking), but the models assume any gene can be transferred (successfully). I think any gene can be transferred, but not necessarily with the same ease. If some genes will just not be displaced. Then there would be a case for something like UCD, but only regarding this set of basic genes. Which would mean that, as a concept for life itself, it would be somewhat meaningless. This is where I stand.

5. Yeah, I know. Some parts sound more like philosophical problem, rather than a scientific one. The thing is to now come with ways of putting these questions into testable hypotheses, then have the community agree on the philosophical side.

6. We would still need to explain why we all use the same genetic code, and other commonalities before abandoning completely the possibility of UCD.

Best,

G.E.

(Now this comment is where it should be)

Anonymous said...

Pvb,

It is funny that you still tried to accommodate my meaning.

Since UCD might be false for Bacteria and ARchaea, it is clear that it would be false for us, because if we take Bacteria and Archaea out, well, it is no longer universal, is it?

I am inclined to think that Eukaryotic Common Descent might not properly be true for us either. Common descent, as in humans and chimps share common ancestors, sure. With gorillas? Sure. With Monkeys? Sure. The problem is where to put the barrier. But ECD ... hum tricky. Have you read about coalescence?

We do have a bit of demonstrable LGT by the way (have you heard of ERVs? Have you heard of ERVs becoming promoter sequences?)

I have not thought a lot about this one (Eukaryotic Common Descent). So, do not take my word for it. I am just thinking out loud.

G.E.

Anonymous said...

So Pvb,

Are you still convinced that UCD is intrinsically untestable? (In all likelihood you are! ... well, you tried to guess me, I have the right to guess you!

I really needed to start re-thinking about this whole mess. Thanks Pvb.

G.E.

P.S. By the way. LGT, I was awfully critical and skeptical when it was heralded as the most important thing in Bacterial evolution. After criticizing yet looking carefully at the data (to find the faults), well, I changed my mind. Yet, I cannot say that it is definitive. Can I?

Anonymous said...

Did you see the base of that tree in the wikipedia article?

Does it look like a Universal Common Ancestor?

I think that article already "decrees" UCD as false.

G.E.

Pvblivs said...

     "It is funny that you still tried to accommodate my meaning."
     How so? I am not in the habit of misrepresenting people. (I am also not in the habit of appealing to ridicule, which is why I withheld the name.)
     "Are you still convinced that UCD is intrinsically untestable?"
     I am. But it appears that we may be in disagreement over what would qualify as an instance of UCD. I will then have to leave the matter here.

Anonymous said...

How so? I am not in the habit of misrepresenting people. (I am also not in the habit of appealing to ridicule, which is why I withheld the name.)

I gues I misworded. I meant to say you still tried to accommodate my meaning to what you supposed would be my position (that LGT can falsify UCD). You thought I would say something about it being a falsifier, just because I do not think that happens in us. Thus the potential falsifier is no such thing.

But now I puzzled. So, to you, a potential falsifier should be something we do expect to happen?

Am I clear on that? I mean. After years of studying evolution, of course you come not to expect somethings to happen, because evolution is, by all evidence around, well, true. Yet, that a devastating falsifier is something we do not expect, what else could it be?

Just clarifying. At least I should not misrepresent you. At Ray's I left thinking you think a falsifier is not such if it is shown after the fact. As if, when a falsifier succeeds, but was not tested as such, would invalidate the falsifier. Am I clear somehow?

G.E.

Anonymous said...

Oh, here is where you "accommodated my meaning":

That is just taking something we believe not to occur and declaring it a "potential falsifier" for that which is intrinsicly untestable.

G.E.

Pvblivs said...

     "But now I puzzled. So, to you, a potential falsifier should be something we do expect to happen?"
     It should be something that would reasonably likely to happen on the assumption that the idea being falsified is, in fact, false. The example I like to use is that the sun rising in the west is not a meaningful potential falsifier for anything, because we can safely predict that that will not happen independent of the idea being "falsified." If you have good reason to believe that the "falsifier" will not happen that does not depend on its inconsistency with the hypothesis under test, it is not a useful falsifier.
     I remember one time someone defended the pre-cambrian rabbit as a potential falsifier for evolution on the grounds that there would be no food supply for the rabbit in precambrian times. But the fact that rabbits need a food supply to exist is not dependent on evolution. So, he was actually showing why the "falsifier" should not be found even if evolution is false.
     I hope I have made my position clear. A useful potential falsifier is something that someone who believes the hypothesis false would expect to find, or, at least, wouldn't be ruling out in advance.

Anonymous said...

Hey Pvb,

But the fact that rabbits need a food supply to exist is not dependent on evolution.

Of course it is. Evolution is variation and selection. Without a food supply the rabbit should not be there.

I am still not clear about what you meant. Let me see:

It should be something that would reasonably likely to happen on the assumption that the idea being falsified is, in fact, false.

But how would it not be reasonable for something to happen if the idea is false to prove that the predictions of the idea (those you deem as not true falsifiers) fail? Perhaps if we continue ...

The example I like to use is that the sun rising in the west is not a meaningful potential falsifier for anything, because we can safely predict that that will not happen independent of the idea being "falsified."

I still do not get it. The sun not coming for the west will not happen. This we know from experience (is this where you are going?), and it is consistent with gravity and such model of the Universe. So, would you say that the theory of gravity is not falsifiable?

If you have good reason to believe that the "falsifier" will not happen that does not depend on its inconsistency with the hypothesis under test, it is not a useful falsifier.

But the infamous rabbits should not happen. This for reasons uniquely related to evolution. If evolution is true, rabbits cannot appear before its predecessors. Rabbits cannot appear without an adequate environment either ... I do not see the problem.

But please clarify: the long gone potential falsifiers (like the molecular thingies which might erode UCD from Bacteria and Archaea), those you do not like because they "were identified after evolution was safe" (whatever that last part means). Why are they inadequate? Because we mention them after we knew they confirm evolution? Is that truly logical to you, or am I missing something?

G.E.

Pvblivs said...

     It was known that food is required for rabbits to exist long before evolution was conceived. Furthermore, even if evoltution is assumed false, rabbits still need a food supply. It is not dependent on evolution. (Evolution may require that life forms need to eat. But life forms needing to eat does not necessarily imply evolution.
     "But how would it not be reasonable for something to happen if the idea is false to prove that the predictions of the idea (those you deem as not true falsifiers) fail?"
     Here we come to the rub. It seems that evolution only predicts things that we already know. More to the point it seems that evolution only predicts things after we already know them to be true.
     "The sun not coming for the west will not happen. This we know from experience (is this where you are going?)"
     Actually, this is exactly where I am going. We already know that it won't happen. Incidentally, it's not predicted by any theory of gravity. The apparent motion of the sun is attributed to the rotation of the earth. If I recall correctly, gravity only accounts for a 6 minute adjustment in the time. But even at that, the mere fact that objects fall down could not be used to test any theory of gravity, because it falls under prior knowledge.
     "Why are they inadequate? Because we mention them after we knew they confirm evolution?"
     Because I find no indication that the "falsifiers" were considered such when it was unknown what the outcome of the relevant experiment would be.

Anonymous said...

Hey Pvb,

Thanks for the patience.

Yes, my bad about the sun gravity thing. The important point is that we know from experience that the sun will not come from the west. Which cannot be used as falsifier for Earth's rotation. Earth's rotation explain the data. But something else might explain it just as well if we were able to falsify Earth's rotation. Right?

Now, we cannot disprove gravity either. Can we? Gravitational theory explains what we see. So, how do we falsify it? (I think we might be getting somewhere here.)

Because I find no indication that the "falsifiers" were considered such when it was unknown what the outcome of the relevant experiment would be.

But how does that matter? If the outcome had been the falsity of evolution it would have been a valid falsifier, right? (To be consistent you would have to say no.) So, why since it "failed" it was not a falsifier at its time? I still cannot see the logic. At its time, whether they were trying to falsify evolution or not, it was a potential falsifier. I truly do not see your logic.

G.E.

Pvblivs said...

     "If the outcome had been the falsity of evolution it would have been a valid falsifier, right?"
     Well, that's just it. If I am correct, had the experiments gone differently, the "potential falsifiers" would be considered nothing of the sort. I remember hearing about the test for chromosome fusion. Leading up to the experiment, they said that, if evolution is true, they might find evidence of chromosome fusion in the 23 human chromosomes (as opposed to the 24 in other primates.) They were prepared to do an exhaustive search. The only conclusion is that they considered that they might do an exhaustive search and find no evidence of fusion and that that would not falsify evolution. Now, people tell me it's a potential falsifier. The only thing that has changed is that what is now considered to be a potential falsifier has been ruled out by experiment.
     Plausible arguments for and against several things being potential falsifiers can be made. And I admit I cannot judge them directly. But it seems to fit the idea that the people involved believe evolution to be true and are deciding whether something is a potential falsifier for evolution based on which decision supports evolution.
     It is a known fact that people tend to interpret data to fit their most cherished beliefs. (Yes, that applies to me, too. Yes, it's possible I'm doing it. These things tend to be unconscious.) It looks like whether something is a falsifier for evolution is subject to that interpretation. If the idea of a falsifier for evolution is, indeed, "that which we know can't happen" then it is unfalsifiable, even in principle.
     That's why "before it was safe" is such a big deal to me. If something had been identified as a potential falsifier for evolution before having been ruled out by observation, it could not have been reinterpreted as a potential falsifier after learning that it would not be found.

Anonymous said...

Pvb,

But you are mixing two things:

1. A potential falsifier being authentic. In other words, it would have falsified evolution had it gone the opposite way.

2. Something that could just be interpreted still as compatible with evolution, yet it is presented as if the opposite result would be a real falsifier.

My complain about "1" is that you think it should have been perceived as a falsifier before testing. My answer, no such thing, because ad it gone the other way around, the scientists might have even published it without noticing that they have a result that falsifies evolution. In other words, intent does not invalidate a falsifier whichever way the results pointed to in the end.

Few noticed that LGT could be a falsifier for UCD. Lots of articles appeared before anybody talked about this possibility ... does that take away LGT from being a potential falsifier for UCD? I doubt it.

I completely disagree that, had the molecules told a different story, evolution would be sustainable as the explanation for life's diversity. If you do not believe it, well, I can do nothing about it.

G.E.

Pvblivs said...

Photosynthesis:

     "But you are mixing two things:"
     It gets worse. I have admitted that I cannot reliably distinguish between #1 and #2. (Is there really anyone both sufficiently informed and sufficiently impartial to do so?) And confirmation bias casts a shadow on any determination made after the fact. It is simply my impression that #1 doesn't exist; and we have #2 passed off as #1. And I came to that conclusion because it just looked like an unreasonable coincidence that the determination of falsifier was always made after the fact. And don't get me wrong. I'm sure it's all in good faith. If I am right and people are reinterpreting things as falsifiers, I'm sure that it is one of those unconscious reinterpretations.
     "Few noticed that LGT could be a falsifier for UCD. Lots of articles appeared before anybody talked about this possibility ... does that take away LGT from being a potential falsifier for UCD? I doubt it."
     That, actually, gets very interesting. I still do not see it as a falsifier for (my interpretation of) UCD. But let's assume for the sake of argument that it does falsify what the scientists mean by it. LGT has been known for a while. UCD has not yet been tossed. That would imply that it has (for the time being) been interpreted as "not a falsifier." That would defeat my belief that UCD cannot be falsified in principle, with the consolation that it would establish my belief in a bias in interpreting data. It will be interesting to see how that turns out. (If I am wrong and UCD can, and does, fall, I know of some people that will have a field day with it.)

Anonymous said...

Hey Pvb,

Well, not been able to distinguish the two does not help. It would help if you just said such, and that, for now, you think most potential falsifiers presneted to you are things that could have been interpreted in favor of evolution had they go "wrong." I would think that, even if something were tested to falsify evolution today, and failed to falsify anything tomorrow, you would still think that a creative mind could have interpreted it the other way around.

But then the problem is not whether UCD is falsifiable, but that creative minds make falsifiability a nebulous thing. Right? (Or both, of course.)

That, actually, gets very interesting. I still do not see it as a falsifier for (my interpretation of) UCD. But let's assume for the sake of argument that it does falsify what the scientists mean by it.

Sure, lets.

LGT has been known for a while. UCD has not yet been tossed. That would imply that it has (for the time being) been interpreted as "not a falsifier."

Well, no, it has been interpreted as a potential falsifier, needing more data before we jump all around. But many microbiologists are already convinced.

That would defeat my belief that UCD cannot be falsified in principle, with the consolation that it would establish my belief in a bias in interpreting data.

A bias that needs no conformation. This is true. But there are quite the good number of very self-critical scientists.

It will be interesting to see how that turns out. (If I am wrong and UCD can, and does, fall, I know of some people that will have a field day with it.)

Tell me about it.

G.E.