Evolutionists also refer to comparative anatomy. But again, these do not tell us how the species arose. Similarity between species does not imply an evolutionary relationship. In fact, comparative anatomy commonly reveals contradictory patterns. Distant species share the same designs, and sister species show very different designs.
Evolutionists also refer to small-scale adaptation we can observe. But we don’t know that these adaptations generally accumulate to create large scale change evolution requires. In fact even evolutionists have agreed this is doubtful, and that some other, unknown, mechanism is required. In fact, these adaptations are produced as a consequence of profoundly complicated molecular structures and mechanisms, whose origin evolution does not explain.
Is there evidence for evolution? Sure, there is plenty of evidence for evolution. But there are significant problems with evolution. There is plenty of evidence for evolution just as there is plenty of evidence for geocentrism. But the science does not bode well for either theory.
So the evidence for evolution follows this general pattern: Even at its best, it does not prove evolution to be a fact. And furthermore, the evidence reveals substantial problems with evolution.
So how can evolutionists proclaim evolution to be a fact with such fervor? There seems to be a glaring mismatch between the evidence and the truth claims of evolutionists. The answer is that evolutionists use contrastive reasoning. Evolution is not claimed to be a fact based on how well it fits the evidence, but rather on how poorly the alternative fits the evidence. Evolution is proved by the process of elimination.
For example, evolutionists explain that nature’s apparently useless or harmful designs make no sense except on evolution. Such harmful designs are actually not predicted by evolution. They are low probability on evolution, but such harmful designs are at least understandable given evolution’s lack of planning. The designs may be low probability, but not altogether impossible.
But if the species were intelligently designed, then these useless or harmful designs make no sense. So we might say that evolution is proved not by positive evidences, but by negative evidences. And in fact the worse the evidence, the better for evolution, because such negative evidences are even worse for the alternative.
Indeed, there are no demonstrations of the fact of evolution that do not appeal to such contrastive reasoning. Evolutionists have a great many proofs for the fact of evolution, but they always entail some form of this contrastive reasoning. Here is how philosopher Eliott Sober explains contrastive reasoning:
This last result provides a reminder of how important the contrastive framework is for evaluating evidence. It seems to offend against common sense to say that E is stronger evidence for the common-ancestry hypothesis the lower the value is of [the probability of E given the common-ancestry hypothesis]. This seems tantamount to saying that the evidence better supports a hypothesis the more miraculous the evidence would be if the hypothesis were true. Have we entered a Lewis Carroll world in which down is up? No, the point is that, in the models we have examined, the ratio [the probability of E given the common-ancestry hypothesis divided by the probability of E given the separate-ancestry hypothesis] goes up as [the probability of E given the common-ancestry hypothesis] goes down. … When the likelihoods of the two hypotheses are linked in this way, it is a point in favor of the common-ancestry hypothesis that it says that the evidence is very improbable. [Evidence and Evolution, p. 314]
These evolutionary arguments and conclusions are very powerful. It seems that the evolutionist’s argument is compelling. The species must have arisen spontaneously via evolutionary mechanisms. But in all of this there is a catch.
Science cannot know all the alternative explanations for the origin of the species. When evolutionists conclude evolution is a fact via the process of elimination, they are making a subtle but crucial non scientific assumption—that they know all the alternative explanations.
So all of these powerful evolutionary arguments for the fact of evolution are non scientific. In other words, evolution has extremely powerful and compelling arguments, but the cost of building such a powerful case is that the idea is not scientific.
Without these powerful proofs, evolution would lie exposed to the many scientific problems and contradictions. The idea that the world, and all of biology, spontaneously arose is, from a strictly scientific perspective, extremely unlikely. But evolution is shielded from such problems by its powerful non scientific proofs.
This non scientific aspect of evolution is immense and would be difficult to underestimate. It has dramatically altered the very perception of science and its evidence. For given the fact of evolution, all of biology is interpreted according to the idea. The many scientific problems with evolution become more friendly “research problems.” And the theory becomes immune to scientific skepticism.
if all that wasn't bad enough, 'inconsistent identity' of cause (i.e. random cause) leads to failure of absolute truth claims for materialists;
ReplyDeleteWhat is the Evolutionary Argument Against Naturalism? (Alvin Plantinga) - video
http://www.youtube.com/watch?v=5yNg4MJgTFw
The following interview is sadly comical as a evolutionary psychologist realizes that neo-Darwinism can offer no guarantee that our faculties of reasoning will correspond to the truth, not even for the truth that he is purporting to give in the interview, (which begs the question of how was he able to come to that particular truthful realization, in the first place, if neo-Darwinian evolution were actually true?);
Evolutionary guru: Don't believe everything you think - October 2011
Interviewer: You could be deceiving yourself about that.(?)
Evolutionary Psychologist: Absolutely.
http://www.newscientist.com/article/mg21128335.300-evolutionary-guru-dont-believe-everything-you-think.html
What of the rival interpretation that it's meaningless to ask of gravity is actually a uniform natural force, in reality, or merely just a useful fiction to predict the motion of objects? After all, there are rival interpretations that accept the same evidence regarding the motions of objects, while indicating that gravity isn't actually a uniform law of nature in reality.
ReplyDeleteFor example, there is the interpretation that gravity is constantly being stabilized by supernatural beings for some reason we cannot understand, which matches all of the evidence we have to date. However, their labor contract goes up for renewal every 13.7 billion years or so. Therefore, gravity might cease to be uniform any day now should they go on strike
And there is the rival interpretation that, despite having a perfectly good law of nature to begin with, a designer decided to tinker with it, making it non-uniform and only recently required intervention by supernatural beings?
In other words, even in the case of gravity, there are an infinite number of rival interpretations that accept the same evidence, but suggest something completely different is going on in reality.
So, it would seem that, in regards to scientific theories, you're making specific assumptions about the relationship between formalisms, predictions and interpretations which doesn't appear constant.
Perhaps you could elaborate on what you think the relationship is, in detail, to help clear this apparent contradiction up?
This comment has been removed by the author.
ReplyDeleteUnknown
ReplyDeleteWhen evolutionists conclude evolution is a fact via the process of elimination, they are making a subtle but crucial non scientific assumption
"Evolutionists" don't conclude evolution is a fact via the process of elimination, That is Cornelius' rather ridiculous strawman.
Science concludes the historical reality of evolution is a fact due to the huge amount of consilient positive evidence.
People in the real scientific community read Creationist silly claims like CH's here and just laugh.
Where's the scientific methods that are reproducable in showing evidence that one fossil evolved into another in the fossil record???? Assumption is your evidence... Laugh all you want but many see evolution as the laughing matter especially since synthetic biology shows a certain amount of "Design by intelligent, living beings" to be a far greater 'possibility' than evolution that is ignored yet can be reproduced in the lab by peers without the same type of "assumption" that Evolution Theory rests on....
Delete@Thorton "People in the real scientific community read Creationist silly claims like CH's here and just laugh."
DeleteIn the scientific community they either "conform" or actually look at data and realize that evolution IS wrong and keep quiet for fear of loosing their jobs. I am no longer an academic in the field so I have NO fear and can relate to the truths as are being seen in the form of evidence in the lab without having to "make the data fit the evolutionary mold" when it doesn't fit!
HAha ha... I can laugh at the assumptive nature of evolution and I can see "where it goes wrong" and I don't care if you laugh at me, in fact it makes my point even stronger when you do! So go for it, you evolutionary hide-from-the-truth blogger (Grin) Note: Being called an "evolutionary" IS becomming quite a laughing matter these days :) "...And it was simple humiliation that showed the wrongs, thought to be truths, studied in science of our past"
Thorton:
DeletePeople in the real scientific community read Creationist silly claims like CH's here and just laugh.
I read your silly comments and worthless opinions on Hunter's blog and I ask myself, "Is this moron for real? How much does he get paid for this or does he do it for free?"
Sometimes I think Thorton is really just a character made up by CH.
DeleteSynthetic biology... Synthetic biology "shows" that a living, intelligence can be behind design in nature! When stated, "Science cannot know all the alternative explanations for the origin of the species. When evolutionists conclude evolution is a fact via the process of elimination, they are making a subtle but crucial non scientific assumption that - they know all the alternative explanations."
ReplyDeleteThe fossil record actually supports the idea that other advanced, intelligent beings designed life on Earth in our distant past many times, leaving a trail of fossils that "fit" this stance more completely than does the evolution theory.
Humanity is just in its infancy of genetic engineering and synthetic biology and space flight... What makes more sense, that evolution took millions of years to evolve (in which there is no scientific evidence backing this position) or that it only takes several thousand years for life to 'advance' to the level of space flight and synthetic biology to be able to venture out into the cosmos to synthetically produce life on another planet? Just something the evolution theory conveniently sweeps under the rug...
Our past is riddled with evidence and rich in religion suggesting such 'people' from the cosmos visiting Earth. Even in the bible, these visitors could heal the sick and raise the dead, we are doing these very things right now yet we are not considered the Gods of the bible, but if you lived back in the times of the bible, you too may 'think' that any visitor that came from the heavens (outer space) and could do things that have never been seen before, would have to be Gods (Plural, as Monotheism is a fairly recent evolution of religion. Before monotheism even in the Hebrew texts, Elohim is plural) By being human, we are not gods and when we design things, they are NOT perfect so when the evolutionary claims that no God would design such things with such flaws, only backs the stance of intelligent design in that the designers were NOT Gods but most likely advance people, just as we are now becoming in our advancements in the sciences.
You sir, are proof positive of why all rational people should fear the bills that allow teachers to present all "theories" that contrast evolution.
DeleteID? sure
Aliens? Why not.
Popol Vuh?
The semen of Atum?
The slaying of Ymir?
Falsify any of those kindly.
Michael Behe on Falsifying Intelligent Design - video
Deletehttp://www.youtube.com/watch?v=N8jXXJN4o_A
Orr maintains that the theory of intelligent design is not falsifiable. He’s wrong. To falsify design theory a scientist need only experimentally demonstrate that a bacterial flagellum, or any other comparably complex system, could arise by natural selection. If that happened I would conclude that neither flagella nor any system of similar or lesser complexity had to have been designed. In short, biochemical design would be neatly disproved.- Dr Behe in 1997
Cornelius,
ReplyDeleteWhen evolutionists conclude evolution is a fact via the process of elimination, they are making a subtle but crucial non scientific assumption—that they know all the alternative explanations.
It's a fact that the milk in my fridge is there because I bought it this morning. Sure, it is possible that (1) a pan-dimensional being put it there, (2) someone else in my family bought it, (3) there is no milk in my fridge etc. There are all these alternative hypotheses I haven't considered and yet I have declared something a fact.
It's almost like Cornelius is speaking rubbish.
Yet, the milk bottle and the fridge have actually been manufactured. It is also a fact.
DeleteThis is getting beyond ridiculous.
ReplyDeleteHere is one evolutionist who does NOT state that "evolution is a fact", who considers it misleading, but who is perfectly happy to accept as a fact that Darwinian mechanisms have been observed to result in adaptation in both lab and field, and who considers the evidence for common descent overwhelming, and sees no reason to think that the major driver of adaptation is not Darwinian natural selection plus drift.
In my experience, most people would concur with the above view, and might even abbreviate as "evolution is now an established a fact". However, if push came to shove, I bet every one of them would say - but all scientific conclusions are provisional, all theories are subject to constant refinement and we are almost certainly wrong about many of the details.
So count me out of your "evolutionists" above, please.
Also, who was it who talked about "the evolving synthesis?" Doesn't he get a mention?
Oh, yes, Allan McNeill.
So that's two. I shall keep a tally.
EL said:
Delete"So count me out of your "evolutionists" above"
Well seems to me that you are out or in according how the argument goes.
"evolution is now an established a fact" is not equal to "Darwinian mechanisms have been observed to result in adaptation in both lab and field, and who considers the evidence for common descent overwhelming, and sees no reason to think that the major driver of adaptation is not Darwinian natural selection plus drift."
EL:
DeleteIn my experience, most people would concur with the above view, and might even abbreviate as "evolution is now an established a fact". However, if push came to shove, I bet every one of them would say - but all scientific conclusions are provisional, all theories are subject to constant refinement and we are almost certainly wrong about many of the details.
Sure, of course. Not only would they say that, they do say it. But saying that the details are likely to change and that the theory is subject to refinement is inconsequential to the claim, the idea is still a fact on par with heliocentrism or gravity. As Gould put it, I suppose apples might suspend in midair tomorrow, but there’s no sense worrying about today (or something like that). All of this doesn’t obviate or lessen the claim that evolution is a fact.
"But saying that the details are likely to change and that the theory is subject to refinement is inconsequential to the claim, the idea is still a fact on par with heliocentrism or gravity."
DeleteYes, pretty well.
That's why you have a complete red herring here.
On the one hand you claim that "evolutionists" are being metaphysical by claiming that their "idea" is an absolute fact (which they don't) an don the other hand you accept that by saying, rather loosely, that "evolution is a fact", they are saying that the broad theory is so well established we can assume it to be true.
Which is not a metaphysical claim at all.
It's just extreme confidence that we have the framework broadly right.
As for your crusade against "contrastive reasoning" - is this new?
And what kind of model-evalution do you suggest if not "contrastive"?
(I asked you this in another thread - I'll try to remember where, and check if you responded).
Elizabeth,
Delete"natural selection plus drift"
As I said in another thread, in this case evolution is just a mode of explanation because there is nothing really that can't be explained by means of evolution.
No one disputes the definition of evolution as per change of frequencies over generations. But how far can one go with this?
In Darwin's days they thought that only natural selection existed. Now that the evidence can't be squashed into it any more, they are adding genetic drift.
Well, maybe it not that hopeless after all and they will one day recognise evidence that only agency credibly produces genuine novelty.
So evolution as a paradigm is just patchwork. It is time the paradigm shifted. The sooner, the better for science.
Cornelius: I have responded to your response to mine about Sober and "contrastive reasoning" on your Gnosis thread. Apologies for the delay.
DeleteLizzie, "evidence for common descent overwhelming:"
Delete--
Lizzie, what you have is observations of bounded variation, and an ungrounded assumption of the historical fossil record. What's overwhelming about that?
Neal -
DeleteLizzie, what you have is observations of bounded variation...
Tell us about these 'bounds'.
What are they, specifically? Where do they lie? What evidence do you have that they exist at all?
Ritchie, just plot the observations. Future speculation is not observation nor is it evidence.
DeleteEL:
DeleteCH: As we saw in the previous post, the consensus position among evolutionists is that evolution is a fact, every bit as much as gravity, the round Earth and heliocentrism are facts.
EL: you claim that "evolutionists" are being metaphysical by claiming that their "idea" is an absolute fact (which they don't) ...
I thought you agreed evolutionists claim their idea is a fact. Are you now denying this?
Neal: Lizzie, what you have is observations of bounded variation, and an ungrounded assumption of the historical fossil record. What's overwhelming about that?
DeleteWell, you have yet to demonstrate that the variation is bounded in any way that would halt incremental adapation.
As for the overwhelming part - phylogenies, in a word.
EL:
DeleteWhy do phylogenies provide overwhelming evidence for evolution?
Aftermath: In Darwin's days they thought that only natural selection existed. Now that the evidence can't be squashed into it any more, they are adding genetic drift.
DeleteThis is a complete misrepresentation.
In Darwin's day they did not known about genetics at all. Moreover Darwin knew that the mechanism of inheritance was unknown, and also the mechanism of variation. He did not think that "natural selection was all that existed".
Now we know what carries inheritance (the genome) and how spontaneous genetic variations (mutations) have phenotypic effects. That means we can model heritability mathematically (population genetics) and see that even neutral mutations result in population change - that populations do not "regress to the mean" but drift about, and that "natural selection" can be modelled as a bias on the drift.
Drift is not "added" to natural selection - in fact, in some senses it's better to think of natural selection as being "added" to drift, or, if you like, as a "special case" of drift.
And these models turn out to be very good.
CH: I thought you agreed evolutionists claim their idea is a fact. Are you now denying this?
DeleteYou missed out my word "absolute".
If they were to declare evolution to be an absolute fact, that would be a metaphysical claim.
They don't. They use it, as you suggest, to mean "we have huge confident in this model, to the extent that we can regard it as a fact".
That's not a metaphysical claim. It's essentially a practical one.
"Why do phylogenies provide overwhelming evidence for evolution?"
DeleteThey provide overwhelming support for a common descent model and paint a consilient picture of the putative tree of life.
The question then becomes: can our postulated evolutionary processes account for the longitudinal change down these putative lineages?
And a great deal, again, of consilient evidence, much of it independent, strongly suggests "yes".
EL:
DeleteBut why to phylogenies provide overwhelming support for CD? There are all kinds of examples of strikingly similar designs in distant species, which cannot be chalked up merely to gene transfer. Likewise, examples of substantial differences in similar species. Those don't fit the CD model.
Neal -
DeleteRitchie, just plot the observations.
Which observations?
Be specific, please.
The Sober quote is just a popular exposition of a statistical technique known as likelihood, which is foundational to all of statistics and used in hundreds of fields in addition to evolutionary biology. If Cornelius's arguments go through, you have to throw out likelihood-based estimation and model-testing in medicine, physics, linear regression, etc. After all, you can never absolutely rule out all alternative explanations, especially when miracles are one of the alternatives, etc., etc., (insert more of Cornelius's relativist, postmodern silliness).
ReplyDeleteFunny NickM that you defend a mere 'likelihood' quote from Sober, of all things, but when shown evidence that renders neo-Darwinian evolution severely implausible, and even impossible, all of the sudden the 'likelihood' does not matter anymore.
DeleteDarwin and the Mathematicians - David Berlinski
“The formation within geological time of a human body by the laws of physics (or any other laws of similar nature), starting from a random distribution of elementary particles and the field, is as unlikely as the separation by chance of the atmosphere into its components.”
Kurt Gödel, was a preeminent mathematician who is considered one of the greatest to have ever lived. Of Note: Godel was a Christian Theist!
http://www.evolutionnews.org/2009/11/darwin_and_the_mathematicians.html
“Darwin’s theory is easily the dumbest idea ever taken seriously by science."
Granville Sewell - Professor Of Mathematics - University Of Texas - El Paso
Here’s That Monumental Evolution Blunder About Probability Again - March 2012
Excerpt: Laplace didn’t rebuke this argument two centuries ago for no good reason—the fallacy has been around forever and evolutionists continue to employ it.,,, It is truly incredible to see evolutionists work their chicanery so they can uphold complete nonsense as the truth. So the evolutionists would credulously accept all manner of bizarre events. If all their roulette wheel bets turned out winners, if their poker hands always gave a royal flush, if random Scrabble letters spelled out CONSTANTINOPLE, it all would be just another small probability event from which nothing can be concluded. This monumental blunder leads them into all kinds of ridiculous conclusions:
http://darwins-god.blogspot.com/2012/03/heres-that-monumental-evolution-blunder.html
The Real Barrier to Unguided Human Evolution - Ann Gauger - April 25, 2012
Excerpt: Their results? They calculated it would take six million years for a single base change to match the target and spread throughout the population, and 216 million years to get both base changes necessary to complete the eight base binding site. Note that the entire time span for our evolution from the last common ancestor with chimps is estimated to be about six million years. Time enough for one mutation to occur and be fixed, by their account.
To be sure, they did say that since there are some 20,000 genes that could be evolving simultaneously, the problem is not impossible. But they overlooked this point. Mutations occur at random and most of the time independently, but their effects are not independent. (Random) Mutations that benefit one trait (are shown to) inhibit another (Negative Epistasis; Lenski e-coli after 50,000 generations).
http://www.evolutionnews.org/2012/04/the_real_barrie058951.html
"He who would deny this must reject all the truths, which we know by induction." -- Daniel Bernoulli, 1735, on contrastive thinking.
Delete"you have to throw out likelihood-based estimation and model-testing in medicine, physics, linear regression, etc." -- Nick Matzke, 2012 on contrastive thinking.
At least you're in good company!
And lest anyone be misled by Nick on this one:
DeleteThe Sober quote is just a popular exposition of a statistical technique known as likelihood, which is foundational to all of statistics and used in hundreds of fields in addition to evolutionary biology.
This is, of course, completely false. Imagine the physicist saying "Gee, the lower the probability of this experimental result is on my theory, the better it actually is for my theory. After all, it just means the alternative would be even worse off!!"
It's just another example of evolutionists saying whatever to avoid the facts.
Oh, and then they blame me for the "relativist, postmodern silliness." You can't make this stuff up.
Cornelius, I have to insist, you are still completely missing Sober's point. And I think even Nick has missed your missing, so fundamental is it.
DeleteThe issue isn't whether Sober is using a likelihood estimate or not (he is, of course, we all agree), it's what his p represents. You are forgetting that p is the probability of an event prior to the information gained from direct observation, that has occurred at least once.
Let p be the probability of the spontaneous appearance of a specific novel sequence.
Sober then assumes (for the same of simple math) that if the novel sequence occurs, the probability that all the descendents of the individual in whom it occurs will carry it is 1.
So the probability of the sequence occurring spontaneously twice is p^2.
However, given the occurrence of the sequence once, the probability of observing the sequence in any descendent, is 1.
If therefore we observe the sequence in two separate populations there are two possiblities:
One is that we are seeing a second spontaneous occurrence of the sequence.
The other is that we are seeing two related populations, both of which are descended from a common ancestor in which the sequence spontaneously occurred.
Now the reasoning is: if p is large, both the explanations shared ancestry and independent occurrences of the sequence are both quite plausible. We can't really plump for one or the other. p and p^2 will be quite similar.
But if the spontaneous occurrence of the sequence is a low frequency event (if p is small) then we can pretty confidently rule out independent occurrences and plump for common ancestry, because p and p^2 will be very different.
So your putative physicist is not saying "Gee, the lower the probability of this experimental result is on my theory, the better it actually is for my theory. After all, it just means the alternative would be even worse off!!"
She is just saying "Gee, the lower the probability of this experimental result, in the event my theory is false, the better this result actually is for my theory!!! "
Which is absolutely fine. It's precisely the reasoning (but on flawed premises) Dembski uses to infer ID.
Oh, and then they blame me for the "relativist, postmodern silliness." You can't make this stuff up.
Maybe you should let Dembski know?
Again, we never speak of the existence of dinosaurs, millions of years ago, as an interoperation of our best theories of fossils. Rather, we say that dinosaurs are the explanation for fossils. Nor is the theory primarily about fossils, but about dinosaurs, in that they are assumed to actually exist as part of the explanation.
DeleteAnd we do so despite the fact that there are an infinite number of rival interpretations of the same data that make all the same predictions, yet say the dinosaurs were not there, millions of years ago, in reality.
We also do so despite the fact that some designer could have intervened in regards to fossils using some unknowable means in some yet to be conceived way. We do this because we do not think any theory is exhaustively true. Science makes no assumption that our knowledge is be exhaustively true. Even the idea that the earth was flat has been falsified, it still contains the truth that the earth appears flat when observe from it's surface.
Perhaps you'd like to explain in detail, how we could have this sort of exhaustive knowledge, and provide quotes of those claiming to hold it in the sense you're implying?
In the same way, evolutionary process are the explanation of biological adaptations. They have an overwhelming impact on the results For there to be additional explanations, you'd need to provide one. However, what we get are easily varied, shallow explanations like "That's just what the designer must have wanted"
So, as usual, your argument is parochial in that it makes several undisclosed assumptions about science, intent, etc., which are not universally held.
Lizzie:
DeleteCornelius, I have to insist, you are still completely missing Sober's point. And I think even Nick has missed your missing, so fundamental is it.
Actually I'm not. This can all be easily cleared up by simply going through the math, which is pretty straightforward. I did so in my response to you in that other thread:
http://darwins-god.blogspot.com/2012/04/here-is-that-secret-gnosis.html?showComment=1336251503193#c7172612187080351263
Lizzie:
DeleteSober then assumes (for the same of simple math) that if the novel sequence occurs, the probability that all the descendents of the individual in whom it occurs will carry it is 1.
Actually no, that assumption is not needed to simplify the math. What does help to simplify the math is to assume that if the trait is not in an ancestral population, then it will not arise in the extant populations. So:
Prob(0-->1) = 0
So the probability of the sequence occurring spontaneously twice is p^2. However, given the occurrence of the sequence once, the probability of observing the sequence in any descendent, is 1.
No, again, you are not following the math.
According to all the evolutionary biology textbooks, evolution is DEFINED as " a change in the frequency of alleles in a population over many generations." There are several variants of this definition but they all mean pretty much the same thing.
ReplyDeleteThe question before us is whether such change has been observed and documented. I maintain that it has: therefore evolution is a fact.
Would it really hurt you to admit that evolution is a known fact that's just as well established as gravity? Why do you insist on using your own (unstated) definition of evolution instead of the one used by scientists?
There are some scientists who are unfamiliar with the proper definition of evolution but the ones who state unequivocally that evolution is a fact know what they're talking about.
Though neo-Darwinists are infamous for claiming that Darwinian evolution is as well established as gravity. This claim is false! For one thing Gravity, as formulated within General Relativity, can be falsified:
DeleteThen the Principle of Equivalence states that
'the inertial and gravitational masses are identical.'
The whole of the General Theory of Relativity rests on this postulate, and will fail if one can find a material for which the inertial and gravitational masses have different values.
http://physics.ucr.edu/~wudka/Physics7/Notes_www/node85.html
Whereas, neo-Darwinism has no identifiable falsification criteria:
Science and Pseudoscience – Imre Lakatos
“nobody to date has yet found a demarcation criterion according to which Darwin can be described as scientific” – Imre Lakatos (November 9, 1922 – February 2, 1974) a philosopher of mathematics and science, , quote as stated in 1973 LSE Scientific Method Lecture
Science and Pseudoscience - Imre Lakatos - exposing Darwinism as a ‘degenerate science program’, as a pseudoscience, using Lakatos's rigid criteria for falsification
https://docs.google.com/document/d/1LpGd3smTV1RwmEXC25IAEKMjiypBl5VJq9ssfv4JgeM/edit
Moreover, General Relativity has been confirmed to stunning degree of accuracy:
Einstein’s General Relativity Tested Again, Much More Stringently - 2010
Excerpt: As Müller puts it, “If the time of freefall was extended to the age of the universe – 14 billion years – the time difference between the upper and lower routes would be a mere one thousandth of a second, and the accuracy of the measurement would be 60 ps, the time it takes for light to travel about a centimetre.”
http://www.universetoday.com/56612/einsteins-general-relativity-tested-again-much-more-stringently/
Whereas neo-Darwinists have yet to demonstrate that even a single protein can arise by purely material processes:
Evolution vs. Functional Proteins - Doug Axe - Video
http://www.metacafe.com/watch/4018222
,,, Moreover, though neo-Darwinists refuse to accept any reasonable falsification criteria for Darwinism, the fact is that science itself could care less and falsification has been provided against neo-Darwinism by advances in quantum mechanics:
Falsification Of Neo-Darwinism by Quantum Entanglement/Information
https://docs.google.com/document/d/1p8AQgqFqiRQwyaF8t1_CKTPQ9duN8FHU9-pV4oBDOVs/edit?hl=en_US
Does Quantum Biology Support A Quantum Soul? – Stuart Hameroff - video (notes in description)
http://vimeo.com/29895068
Professor Moran:
DeleteAccording to all the evolutionary biology textbooks, evolution is DEFINED as " a change in the frequency of alleles in a population over many generations." There are several variants of this definition but they all mean pretty much the same thing. The question before us is whether such change has been observed and documented. I maintain that it has: therefore evolution is a fact.
When evolutionists claim their idea is a fact, they are referring to much more than merely changes in allele frequencies. You can see this in the example quotes given in the previous post:
http://darwins-god.blogspot.com/2012/05/you-wont-believe-what-evolutionists.html
For example, Lewontin states that “It is a fact that all living forms come from previous living forms. Therefore, all present forms of life arose from ancestral forms that were different. Birds arose from nonbirds and humans from nonhumans. No person who pretends to any understanding of the natural world can deny these facts any more than she or he can deny that the earth is round, rotates on its axis, and revolves around the sun.”
Likewise Futuyma states that “the statement that organisms have descended with modifications from common ancestors--the historical reality of evolution--is not a theory. It is a fact, as fully as the fact of the earth's revolution about the sun.”
Similarly Jerry Coyne states that “Now, when we say that “evolution is true,” what we mean is that the major tenets of Darwinism have been verified. Organisms evolved, they did so gradually, lineages split into different species from common ancestors, and natural selection is the major engine of adaptation. No serious biologist doubts these propositions.”
So evolutionists are quite clear about their claims. Evolutionists use the word “evolution” in the same way that people generally understand it, that is to represent the origin of species, not merely changes in allele frequencies. I’m not interested in playing semantic games. If you want to say that the origin of species is not “evolution,” then fine. Use whatever word you like, but that is what evolutionists claim to be a fact.
Would it really hurt you to admit that evolution is a known fact that's just as well established as gravity? Why do you insist on using your own (unstated) definition of evolution instead of the one used by scientists?
I’m afraid I’m not the one using a peculiar definition. Nor is it unstated. I’m addressing precisely what evolutionists claim to be a fact, nothing more, nothing less.
There are some scientists who are unfamiliar with the proper definition of evolution but the ones who state unequivocally that evolution is a fact know what they're talking about.
Well now you are contradicting yourself. Lewontin, Futuyma, Mayr, Coyne and the others are most definitely stating that evolution is a fact, and they are not referring to allele frequency changes.
The question before us is whether such change has been observed and documented. I maintain that it has: therefore evolution is a fact.
DeleteChanges in the frequency of alleles has been observed, therefore the idea that all changes, every single species and everything that can be observed in biology arose spontaneously is also a fact.
I wonder why people still debate such things then. After all, you could just point to your nose and say something like: "Look here at this evidence of a change, therefore I'm on my way to a knowledge of the origins of everything that can be observed in biology!"
Change happens, it's like excrement in that respect. And so, here is your knowledge...
Well now you are contradicting yourself. Lewontin, Futuyma, Mayr, Coyne and the others are most definitely stating that evolution is a fact, and they are not referring to allele frequency changes.
DeleteThe truth of what people have said and how they have actually represented evolution/change to the public doesn't matter. Those closer to the top of the cartel of knowledge are just charlatans like the occult magicians who originally gave them their "out of chaos, order" creation myths. So when illusions of knowledge fail they will shift toward manipulation and coercion through the use of professional identity and associations.
I like the rules of the game given by those closer to the top of the pyramid scheme of gnostic knowledge though. For instance, always allowing imagining what God wouldn't do but never allowing people to imagine what God would do. Always allowing for invocations of "chance"/ignorance within your stories but never allowing the same for people who would imagine things in any other way.
Given enough "chance" of that sort how could you ever lose? No wonder chance/ignorance is being treated as the equivalent of a fact/knowledge these days...
Laurence, welcome to the discussion.
DeleteThe majority of us here understand the definitions of evolution that are used and Cornelius' reply nails the equivocation that evolutionists make.
Having studied the definitions of evolution for years and how it is used in both scientific and popular literature, I see the equivocation all the time... we observe variation in bird beak sizes and moth coloration, therefore all of life is descended from a single cell organism. It's like saying we observe flat fields, therefore the earth is flat.
This comment has been removed by a blog administrator.
DeleteCornelius Hunter
DeleteI’m not interested in playing semantic games.
LOL! Really CH, we're all pals here. Who do you think you're kidding? Silly semantic games are ALL you're interested in, as you bounce back and forth and equivocate over the many different meanings and usages of the word "evolution".
You sure aren't interested in providing your alternate explanations for any of the scientific findings you present. But we understand. You regurgitate the same tired old propaganda, your handful of Creationist/IDiot groupies get to vent their frustration over being so scientifically impotent, you collect your paycheck from the DI. Life goes on.
Neal: "we observe variation in bird beak sizes and moth coloration, therefore all of life is descended from a single cell organism. It's like saying we observe flat fields, therefore the earth is flat."
DeleteCan you give me a citation for somebody claiming making this inference?
Because as far as I can tell, your hyperbole materially affects the truth of your claim.
Lizzie, that was a simplification of evolutionary thought, but if one peels away the rhetoric of evolutionary thinking like an onion, this is quite representative of what's left. Bird beaks are representative of the majority of the mountain of "evolutionary evidence" which only shows oscillating patterns of variation and nothing more.
DeleteEvolutionary interpretation of the fossil record is a joke, if not a fraud. The fossil record does not show anything near a consistent record of bottom up gradualism, but a top down pattern with major life forms showing up without ancestry followed by stasis with bounded variation.
The mixing and matching of genes showing up in supposedly distantly related species sinks Darwinism. Have you any reply regarding this critical data yet?
mymyn: For instance, always allowing imagining what God wouldn't do but never allowing people to imagine what God would do
DeleteYou mean, like imaging a designer that decided to create the world we observe last Thursday?
But if this was the case, dinosaurs wouldn't be the explanation for fossils. Rater than designer would have created them when it created the world we observe last Thursday.
Yet, we never speak of the existence of dinosaurs as an interoperation of our best theories of fossils. Rather, we say that dinosaurs *are* the explanation for fossils. And we do this despite the fact that there are an infinite number of rival interpretation that accept the same evidence, yet indicate that dinosaurs never existed millions of years ago.
Does this mean that the existence of dinosaurs is somehow held hostage by imagining what God would do?
If I understand these comments correctly, it seems that IDiots and scientists all agree that evolution, as defined, is a fact.
DeleteNow we're just quibbling over how far that facthood can be extended to the history of life on this planet. Is it a fact that the human populations of North America and Asia have a common ancestor even though they differ in the frequency of blood type alleles?
Is it a fact that all the Galapogos finches share a common ancestor?
Is is a fact that humans and chimps share a common ancestor? I say "yes" to all three as do Michael Behe and Michael Denton, among others.
But even if you deny these obvious facts, you can't deny that evolution is a fact at some level ...
Oops, I take that back, some people can deny anything even the most blatantly obvious facts.
If evolution is a known fact then how do you explain the Cambrian explosion?
DeleteProfessor Moran:
DeleteIf I understand these comments correctly, it seems that IDiots and scientists all agree ...
Please keep the language respectful.
Laurence,
DeleteEquivocation:
1: to use equivocal language especially with intent to deceive
2: to avoid committing oneself in what one says
Synonym: Evolution
It is a fact that the field in my back yard is flat... it is a fact that the earth is flat
Flat earth theory circa 5000BC- One must differentiate between the fact of flat earth and the theories of flat earth. While theories of flat earth change, the fact of flat earth has been confirmed repeatedly by farmers observing flat earth all over our provinces.
DeleteBut even if you deny these obvious facts, you can't deny that flat earth is a fact at some level ...
Oops, I take that back, some people can deny anything even the most blatantly obvious facts.
This comment has been removed by a blog administrator.
DeleteCornelius Hunter
DeleteProfessor Moran:
If I understand these comments correctly, it seems that IDiots and scientists all agree ...
Please keep the language respectful.
Yes, please remember the Creationist guidelines:
It's OK to tell the most blatant lies about scientific research, and to accuse the researchers of being either hopelessly incompetent and/or deliberate frauds, as long as you use respectful language while doing it.
Laurence A. Moran said:
Delete"Is it a fact that the human populations of North America and Asia have a common ancestor even though they differ in the frequency of blood type alleles?
Is it a fact that all the Galapogos finches share a common ancestor?
Is is a fact that humans and chimps share a common ancestor? I say "yes" to all three as do Michael Behe and Michael Denton, among others."
Then as this:
"http://www.touropia.com/natural-rock-formations/"
is the product of natural process, giving enough time natural process will produce this:
"http://en.wikipedia.org/wiki/Stonehenge"
PS: I hope my hyperbole do not materially affect the truth of my claim in the eyes of Elizabeth.
Thorton:
Delete"Yes, please remember the Creationist guidelines: It's OK to tell the most blatant lies about scientific research, and to accuse the researchers of being either hopelessly incompetent and/or deliberate frauds, as long as you use respectful language while doing it."
Was that another random mutation at work?
Cornelius Hunter
DeleteThorton: "Yes, please remember the Creationist guidelines: It's OK to tell the most blatant lies about scientific research, and to accuse the researchers of being either hopelessly incompetent and/or deliberate frauds, as long as you use respectful language while doing it."
Was that another random mutation at work?
Sadly, it's an empirically observed fact.
"Was that another random mutation at work?
DeleteSadly, it's an empirically observed fact."
There you go again, another "fact" claim ...
Cornelius Hunter
Delete"Was that another random mutation at work?
Sadly, it's an empirically observed fact."
There you go again, another "fact" claim ...
Yep, just like it's a fact you've never once offered your "correct" alternate explanations for any of the data you claim evolutionary scientists get horribly wrong.
What was that definition of "equivocation" again?
Equivocation:
1: to use equivocal language especially with intent to deceive
2: to avoid committing oneself in what one says
This comment has been removed by a blog administrator.
DeleteLouis Savain says,
Delete"Stop it with the blatant lies already. You may fool the choir but not everybody sings in your stupid choir."
It will be interesting to see how long it takes for Cornelius Hunter to caution you for using disrespectful language.
Waiting ....
Louis gets smacked down fairly regularly. Usually there is no warning.
DeleteProfessor Moran:
Delete“It will be interesting to see how long it takes for Cornelius Hunter to caution you for using disrespectful language.”
The guidelines are no foul language, personal attacks or derision. So for example, it is OK to call someone's idea [insert pejorative], but not not OK to apply that to the person. Admittedly these guidelines can be subjective, but they were necessary. So in this case, "your stupid choir" is OK but "cowardice" is not.
Nor is IDiot or LOL. You received a warning out of respect for you and because you may have been unaware of the guidelines.
Professor Moran:
DeleteNow we're just quibbling over how far that facthood can be extended to the history of life on this planet. Is it a fact that the human populations of North America and Asia have a common ancestor even though they differ in the frequency of blood type alleles?
Is it a fact that all the Galapogos finches share a common ancestor?
Is is a fact that humans and chimps share a common ancestor? I say "yes" to all three as do Michael Behe and Michael Denton, among others.
But this is the very point in contention, so it seems to me that you are begging the question. In this case, you seem to be saying that humans and chimps sharing a common ancestor is merely a minor step beyond what would be required for the Galapagos finches to share a common ancestor. But that would not be true.
The good news is we all can agree on the evidence at hand, what the known biological processes are (such as adaptation), and what the level of inference is for different conclusions. It is not controversial that there are no mechanisms that are known to be capable of changing a small primate into a human. Indeed, there are all manner of traits and characteristics of the millions of species on Earth, including humans, the origin of which evolution is not able to explain beyond broad, cartoon-level, speculation. I don’t mean this as an attack, this is simply the state of the science.
So give the state of the science, how can evolution, or common descent, be a fact?
Hunter:
DeleteThe guidelines are no foul language, personal attacks or derision. So for example, it is OK to call someone's idea [insert pejorative], but not not OK to apply that to the person. Admittedly these guidelines can be subjective, but they were necessary. So in this case, "your stupid choir" is OK but "cowardice" is not.
OK, but I'm sure Moran got my message loud and clear.
Cornelius Hunter
DeleteSo give the state of the science, how can evolution, or common descent, be a fact?
Because the evidence that an event occurred, and the understanding of the specific mechanisms of how the event occurred are two different things.
It's a fact that TWA flight 800 crashed off of Long Island in 1996 due to an on-board explosion. There is more than enough physical evidence to support facthood.
It's a well supported theory that the crash was caused by a build up of fuel vapor and subsequent explosion due to a spark in a near empty fuel tank.
It's a fact that evolution by common descent over deep time occurred. There is more than enough physical evidence to support facthood.
It's a well supported theory that such evolution occurred by mechanisms such as genetic variation (due to mutations, drift, HGT) and natural selection leading to an accumulation of beneficial traits.
You of course know the difference between the fact of evolution and the theory of evolution, but will continue to make the false equivocation as long as you still get propaganda mileage with your Creationist sycophants from the canard.
And considered the source.
DeleteCornelius Hunter
DeleteIt is not controversial that there are no mechanisms that are known to be capable of changing a small primate into a human.
You may want to consider taking some remedial science courses. Science has known the mechanisms to a fair amount of detail for over 50 years.
Indeed, there are all manner of traits and characteristics of the millions of species on Earth, including humans, the origin of which evolution is not able to explain beyond broad, cartoon-level, speculation.
Not knowing every detail doesn't mean not knowing any detail. Amazing that you still get that wrong even after all the times you've been corrected.
I don’t mean this as an attack, this is simply the state of the science.
Of course you meant it as an attack. Attacking the sound sciences that support evolutionary theory is what you do. You don't present the state of science; you present the twisted dishonest mangling of the state of science the Discovery Institute pushes.
Amen, Amen, Amen.
ReplyDeleteSmall scale adaption is just a line of reasoning when invoked to explain macro evolution claims.
It could only be that small adaptive abilities exist.
People are case in point of potential if we are from the same parents.
Fossilism is just pictures of former life in a moment and is not saying anything about biological relationship.
Thats only deduced once geological relationships are accepted.
yet srtill conclusions for evolution are not from biological investigation.
Its a special case of comparing things said to be unrelated in time existence.
comparative anatomy presumes like equals same.
A common blueprint would account for this and so the case is not made because things look alike.
In fact they deny this with convergent evolution ideas they invoke when desperately needed.
Just watch the marsupial wolf on youtube in still or moving pictures.
Either thats just another wolf with a pouch or indeed shows a important biological reality about creatures.
Fossilism is just pictures of former life in a moment and is not saying anything about biological relationship.
DeleteNot to mention that the picture is generally based on people imagining things about the past based on the skeletal remains of organisms. Not to mention the fact that the way that they've been conditioned to imagine things about biology of living things based on dead bones has been linked to their professional identity and the way they make a living. Due to the actions of the cabal that established the ancient "out of chaos, order" memes in modern times there is generally fame to be found in "missing links" and little to be found in imagining things about the past in any other way. Not that imagining things about dead bones is the equivalent of actual biological knowledge anyway. And it's telling that even when it is granted as almost some sort of equivalent for actual biological knowledge there are still deep problems with even imagining things about the past. We're a long way from biologists having a "theory of evolution" by which they can predict a trajectory of adaptation within a living population here.
Even within the field of "knowledge" of imagining things about the past based on the skeletal remains of organisms, the type of illusions of knowledge typical to the cabal of the "illuminated" that established modern creation myths in the minds of imbeciles these days is evident:if stasis could not be explained away as missing information, how could gradualism face this most prominent signal from the fossil record? The most negative of all strategies-a quite unconscious conspiracy of silence-dictated the canonical response of paleontologists to their observations of stasis. Again, a “culprit” may be identified in the ineluctable embedding of observation within theory. Facts have no independent existence in science, or in any human endeavor; theories grant differing weights, values, and descriptions, even to the most empirical and undeniable of observations. Darwin’s expectations defined evolution as gradual change. Generations of paleontologists learned to equate the potential documentation of evolution with the discovery of insensible intermediacy in a sequence of fossils. In this context, stasis can only record sorrow and disappointment.
DeletePaleontologists therefore came to view stasis as just another failure to document evolution. Stasis existed in overwhelming abundance, as every paleontologist knew. But this primary signal of the fossil record, defined as an absence of data for evolution, only highlighted our frustration-and certainly did not represent anything worth publishing. Paleontology therefore fell into a literally absurd vicious circle. No one ventured to document or quantify-indeed, hardly anyone even bothered to mention or publish at all-the most common pattern in the fossil record: the stasis of most morpho-species throughout all their geological duration.
(The Structure of Evolutionary Theory (Harvard College) by Stephen Jay Gould :759-760)
Gould wrote: Darwin’s expectations defined evolution as gradual change.
ReplyDeleteAnd how did Darwin's memes emerged and what were they defined by? Look to the work of his father and his involvement in the same type of occult societies which established the Royal Society to spread Darwinian memes. Note that the "cabal" of the so-called illuminated was so successful that it established new creation myths in the West. Its power is also evident in the fact that those who imagine themselves as closer to the top of the pyramid $cheme have gained control over the economic language of the world. Control over creation stories and the creation of money, what more could be asked for?
Still, the gnostic establishment of a new order for the ages is a long time coming... it's almost as if the Brave New World envisioned by the illuminated elite of the occult/hidden is being held back by something or other.
A Huxley reference?
ReplyDeleteThat aside, are the new post guidelines now defunct? A pity if so.
EL:
Delete" your hyperbole materially affects the truth of your claim"
Classic
Yes, it's a reference to the occult/hidden nature of the invisible college and the gnosis/knowledge which is symbolized in the secret societies that Huxley and the Darwin's were involved in.
ReplyDeleteThey successfully shaped scientia/knowledge in the image of their old memes through the Royal Society and so on.
It's too bad about the eugenics movement toward world war though. Pay no mind to the mound of skulls at the bottom of pyramid $chemes of the Darwinian sort with their ivory towers of nothing but babble which trace back to the monuments to ignorance that liter the ancient world. If the illuminated have been successful in explaining your mind/knowledge in terms of ignorance then is it fair to say that they've made you lose your mind? Lol...
Mynyn,
DeleteYour hyperbole materially affects the truth of your claim
You regurgitate the same tired old propaganda, your handful of Creationist/IDiot groupies get to vent their frustration over being so scientifically impotent, you collect your paycheck from the DI.
ReplyDeleteIf only you could eliminate his funding or end his career then you could manufacture a broader form of consensus to cite as the equivalent of specified and verifiable scientific theory, right? I suppose it worked for charlatans seeking to built pyramid $chemes based on ignorance in the past.
The thing about it these days is that technology (which comes about based on intelligent design) may be decentralizing knowledge and information more quickly than you can fund the manufacture of a consensus.
Your hyperbole materially affects the truth of your claim.
ReplyDeleteShrug... I doubt that you even recognize the truth of my claims, symbolically speaking.
Actually, I'm probably underestimating the way that the herd can be guided by intelligent people. Even with all the new forms of technology which decentralize knowledge, it seems that those at the base of the American pyramid $cheme are too busy trampling people to death at Walmart lift their eyes up and focus on the truth of things.
ReplyDeleteMeanwhile, scientists who consider themselves to be more knowledgeable and intelligent than others are still dumb enough to allow themselves to be manipulated through their professional identity and funding. If or when American civilization declines further perhaps it can be said that Darwinism was a self-fulfilling form of prophecy in which people are totally governed by the carnal or animal side of things.
Perhaps stimulus funding could be directed to stimulate the scientific herd to form a firmer consensus on modern creation myths and more profitable ideas like global warming?
ReplyDeleteThat way Americans could at least entertain themselves with the idea that they are preventing imaginary catastrophes in the future by driving a different car as their civilization declines.
Here we are now, entertain us.
I didn't realize that those were claims, sorry. I have a general rule,whenever I read" mound of skulls at the bottom of pyramid $chemes of Darwinian sort with their ivory tower of babble" ,I assume no factual claims are present.
ReplyDeleteLet's see,Walmart ,global warming,stimulus funding,American civilization declines,imaginary catastrophes,pyramid $cheme,eugenics,the illuminated,Royal Society,cabal.
Cheap, d'uh I live in Texas, good idea,how can you tell,better than real ones,$10,000 pyramid?, against, good on a dark night, still around? , against generally
If I drive a black 2002 Tacoma pickup,Trd, two wheel drive, 4 cylinder, am I preventing an imaginary catastrophe? If so,you all are welcome.
You miss the point Mynym,you are the entertainment.
This comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteCornelius, I have responded to your comment re Sober on the Gnosis thread here.
ReplyDeleteLizzie, you have a couple of minor errors in your typing (I think you meant P*N, not P^N, and you have some unmatched parentheses). But that's minor. The main problem is that the example you provide doesn't change my point. That is to say, in your example as well as in the likelihood argument Sober elucidates, the argument for common ancestry (CA) becomes powerful when the probability of the evidence, given CA, worsens. CA is made to look good by virtue of contrastive reasoning, in spite of the the evidence. And in fact, the worse the evidence, the better for CA! Your posterior probability is not aloof of the contrastive reasoning. I replied here:
Deletehttp://darwins-god.blogspot.com/2012/04/here-is-that-secret-gnosis.html?showComment=1336347901458#c5527093997842785181
Sober elucidates why the argument for CA is so powerful. Sure there is all kinds of evidence, but he shows that the really powerful arguments are those where the probability of the evidence is really bad. In those cases, CA looks really good, because of the contrastive reasoning.
To use your example, in your posterior probability calculation, CA looks really good as the numerator *decreases*. The trick is that the denominator decreases even faster, courtesy of contrastive reasoning.
CH: Lizzie, you have a couple of minor errors in your typing (I think you meant P*N, not P^N, and you have some unmatched parentheses)
DeleteWell, spotted, thanks!
CH: That is to say, in your example as well as in the likelihood argument Sober elucidates, the argument for common ancestry (CA) becomes powerful when the probability of the evidence, given CA, worsens.
You keep talking about "the probability of the evidence". That may be misleading you. It's not "the probability of the evidence", it's simply the probability of the data. But as the data are given (which is what the word means) their posterior probability is 1!
CH:CA is made to look good by virtue of contrastive reasoning, in spite of the the evidence. And in fact, the worse the evidence, the better for CA! .
No. This is the mistake you keep making. The evidence for CA is good, by virtue of the low prior probability of the data. It's not "the worse the evidence, the better for CA" - it's "the lower the prior probability of the data, the better the evidence for CA".
You do in fact understand this, as it applies directly to your plagiarism example. The lower the prior probability of the suspect words, the better the evidence for plagiarism, given those words. The higher the prior probability of the suspect words, the worse the evidence for plagiarism.
It's just that you are mapping your argument on to the wrong bits of math. What makes the case potentially unreasonably strong for CA is not that the fact that p is low makes the "evidence" for CA worse, but that p(CA) is assumed to equal 1-p(SA).
As you agree, I think. You've just got the Bayesian bit of the argument wrong, as is nicely illustrated here:
CH:Sober elucidates why the argument for CA is so powerful. Sure there is all kinds of evidence, but he shows that the really powerful arguments are those where the probability of the evidence is really bad. In those cases, CA looks really good, because of the contrastive reasoning.
No, the probability of the evidence is not "really bad". The prior probability of the data is really low, making the evidence really good. Yes, this is to do with "contrastive reasoning" but has nothing to do with the prior probability of the "evidence [being] really bad".
CH:To use your example, in your posterior probability calculation, CA looks really good as the numerator *decreases*. The trick is that the denominator decreases even faster, courtesy of contrastive reasoning.
No, that is not why the numerator decreases, and, I repeat yet again, the numerator is not the "evidence" for CA. It is simply the probability of the data, given CA. This is simply not the same thing.
The odds ratio between p(data|CA) (numerator) and p(data|SA) is not the ratio between the evidence for each. If we set our priors for CA and SA as equal (as I did, and Sober does) then the odds ratio is a way of expressing the posterior probability for CA but it is not the posterior probability. To compute the posterior probability we simply add the numerator to the denominator, so that we get a value between 0 and 1. That will tell us the strength of the evidence for CA. And that gets stronger as p gets weaker. This is not a "trick" of "contrastive reasoning" as I will demonstrate below:
Let's suppose that instead of two hypotheses to account for the appearance of M in two species, we have three.
DeleteThe first is IM for "Independent Mutations" (as on the other thread).
The second is CA for "Common Ancestry" - the hypothesis that the two populations have a common ancestor.
The third is SC for "Special Creation" - the idea that although the species do not share a common ancestor, they do share a common creator who recycled DNA sequences.
I will also add NU for "null" or "none of the above" (could include HGT, for instance, but also "contaminated sample").
Because I have an explicit null, I can say that p(CA)+p(SC)+p(IM)=1-p(NU).
I shall give them each a prior of .25, so that the priors will cancel out and make the math easier.
Now, what I want to know is the p(CA|data) i.e. the evidence for common ancestry aka the posterior probability of CA.
So, on the top line goes:
p(data|CA)*p(CA)
which we compute as before.
On the bottom line goes:
p(data|CA)*p(CA) +p(data|SA)*p(SA)+p(data|ID)*p(ID)+p(data|NU)*p(CA)
That is more tricky, because we need a specific hypothesis for p(data|ID), and we also need some kind of value for p(data|NU).
However, that is OK, because, regardless of those values, as long as they are non-zero, you can immediately see that by adding two new hypotheses we have increased the value of the denominator. After factoring out the priors and cancelling, our old denominator was:
p(data|CA)+p(data|SA).
And our new denominator is:
p(data|CA)+p(data|SA)+p(data|ID)+p(data|NU).
And so, simply by adding in more competing hypotheses we have increased the value of the denominator and thus reduced the posterior probability of CA (and we can do this in turn for each hypothesis).
But the value of p has not changed. In other words, the issue is not that a low p means small evidence for CA. It's that the evidence for CA depends not only on p but on how many hypotheses you are contrasting it with, what their priors are, and what the value is for the probability of the data under each of those hypotheses.
All hypothesis testing is contrastive. We don't have another method. Bayesian methods on the whole are, IMO, less misleading than Fisherian methods because you are actually computing the probability of your hypothesis, which is what you want, not the probability of the data given the null, as in Fisherian statistics, which isn't.
But either way, your confidence that your hypothesis is correct depends on what alternatives you have considered. The problem is not in "contrastive reasoning" per se but in figuring out what you are contrasting your hypothesis with.
I have responded to you here:
Deletehttp://darwins-god.blogspot.com/2012/04/here-is-that-secret-gnosis.html?showComment=1336414357451#c4833905046418511504
Lizzie:
DeleteNo. This is the mistake you keep making. The evidence for CA is good, by virtue of the low prior probability of the data. It's not "the worse the evidence, the better for CA" - it's "the lower the prior probability of the data, the better the evidence for CA".
We’re talking past each other because I’m looking at it from a scientific perspective and you are looking at it from a metaphysical perspective. It’s a very instructive example. From a scientific perspective, we have a hypotheses and we have data which have low probability given that hypothesis. It is only with evolutionary metaphysics that the evidence is better for CA. But from a metaphysical perspective, yes no question about it, the evidence is strong for CA.
No, the probability of the evidence is not "really bad". The prior probability of the data is really low, making the evidence really good. Yes, this is to do with "contrastive reasoning" but has nothing to do with the prior probability of the "evidence [being] really bad".
No, of course it has something to do with the prior probability of the evidence being really bad. That’s where the metaphysical contrastive reasoning pays off.
CH:To use your example, in your posterior probability calculation, CA looks really good as the numerator *decreases*. The trick is that the denominator decreases even faster, courtesy of contrastive reasoning.
EL: No, that is not why the numerator decreases, and, I repeat yet again, the numerator is not the "evidence" for CA. It is simply the probability of the data, given CA. This is simply not the same thing.
Fascinating discussion. You seem to be taking every avenue possible to avoid the obvious. “No, that is not why the numerator decreases”. What is “not why the numerator decreases”? “and, I repeat yet again, the numerator is not the "evidence" for CA.” But of course I never said that.
the issue is not that a low p means small evidence for CA
And I never said it was. This is becoming comical.
All hypothesis testing is contrastive. We don't have another method.
Regardless, in the hands of evolutionists it is metaphysical. So when you say the evidence is overwhelming, that is a non scientific claim.
Oh, for goodness' sake, Cornelius! I have no idea what a "metaphysical perspective" even is (at least I thought I did, but I have no idea what you mean by it), I'm just a scientist who tests hypotheses for a living. That is my perspective.
DeleteCH:From a scientific perspective, we have a hypotheses and we have data which have low probability given that hypothesis.
Yes. But we also have that data. The posterior probability of those data is 1. The uncertainty is gone. It's like saying that the probability that the coin is heads, once it has fallen heads, is still .5! That's the part you keep ignoring! There's nothing "metaphysical" about those data being actually observed! In fact, if anything is "metaphysical" it's the concept of probability itself. What we are really talking about here is normalised frequency, and once an even infrequent event has occurred, then it isn't improbable any more - it has occurred.
CH:It is only with evolutionary metaphysics that the evidence is better for CA. But from a metaphysical perspective, yes no question about it, the evidence is strong for CA.
The evidence is strong for CA if CA is the only hypothesis that can account for the data. If it isn't, then it isn't strong. But as Sober lays it out, CA is the only hypothesis on the table under which the occurrence of same mutation in two different species is non-independent. Nothing metaphysical about that - just a challenge to provide an alternative non-independent account of those data.
CH:No, of course it has something to do with the prior probability of the evidence being really bad. That’s where the metaphysical contrastive reasoning pays off.
Please explain where this "Metaphysical" stuff kicks in. This is making no sense at all. A low probability of a spontaneous mutation is not "metaphysical", and the fact that where this probability ("p") is low, the posterior probability of CA is raised is not "metaphysical" either. Nor, indeed, is the fact that the likelihood ratio is higher when p is low "metaphysical".
The problem is not "metaphysical" but logical. If a middle has been excluded, your argument is not valid. That's why Dembski's argument is invalid, and why Sober's is.
Nothing to do with "contrastive" reasoning, nothing to do with a low value of p.
continued...
DeleteCH:Fascinating discussion. You seem to be taking every avenue possible to avoid the obvious. “No, that is not why the numerator decreases”. What is “not why the numerator decreases”? “and, I repeat yet again, the numerator is not the "evidence" for CA.” But of course I never said that.
You have variously called it:
the conditional probability for CA, which you "clarified" as: the probability for the CA case
You also called it: "the probability of the evidence given CA" and "the conditional probability of the evidence given CA"
Those last two are closest, though I'd rather call it the data, than the evidence, although I'll buy that last thing. But what it is not is "the probability for the CA case".
CH:And I never said it was. This is becoming comical.
Well that is how I read your description of the numerator as "the probability for the CA case". But it isn't what you mean, fine.
However, given that it isn't what you meant, then why are you concerned with a low p(data|CA) when we actually have that data? Why worry on Tuesday that the forecast said it might rain on Monday, when in fact, it did not?
CH:Regardless, in the hands of evolutionists it is metaphysical. So when you say the evidence is overwhelming, that is a non scientific claim.
Yes, it is, rather. To be a scientific claim I'd need to do a proper literature review and preferably a meta-analysis. Which in this case would be wildly inappropriate because "evolution" is a vast body of interlinked theory, some of which is better supported than others.
But do I take it that you now agree that all hypothesis testing is contrastive? And do I take it that you also agree that to make a valid contrastive inference, we need to ensure we have not excluded a middle? (As in the unwarranted assumption implicit in Sober's example that p(CA)=1-p(SA)?
And, if so, would you agree that Dembski has made a major error?
EL:
DeleteOh, for goodness' sake, Cornelius! I have no idea what a "metaphysical perspective" even is (at least I thought I did, but I have no idea what you mean by it), I'm just a scientist who tests hypotheses for a living. That is my perspective.
Well it’s all well and good to say “I’m just a scientist” but you’re the one who claimed the evidence for CA is overwhelming and who provided posterior probabilities for CA based on a non scientific premise about CA and SA being the alternatives. Both are claims are outside of science.
There's nothing "metaphysical" about those data being actually observed!
Agreed.
But as Sober lays it out, CA is the only hypothesis on the table under which the occurrence of same mutation in two different species is non-independent. Nothing metaphysical about that
No, that is metaphysical. The evolutionary premise is that it knows all the alternatives, in this case labeled CA and SA. You can’t get that from science.
Please explain where this "Metaphysical" stuff kicks in. This is making no sense at all.
Whether in the likelihood ratio, or in the form of Bayes’ Rule that you used, your premise is that you have knowledge of all the alternatives. In this case, CA and SA. That doesn’t come from science. That premise occurs before the science begins. It is above the science, or meta-scientific. And this runs all through the literature. In one form or another, the high claims for evolution always come from some version of these metaphysics.
But in fact, the results tell you merely how your CA model compares with your SA model. It tells you nothing about the probability of either model being true, or approximating reality. It does not warrant the claim that evolution or CA is a fact, or is astronomically likely to be true, or is overwhelmingly supported by the scientific evidence, or any other way you want to parse the language to make it palatable.
Continued …
Continued …
DeleteYou have variously called it:
the conditional probability for CA, which you "clarified" as: the probability for the CA case
You also called it: "the probability of the evidence given CA" and "the conditional probability of the evidence given CA"
Those last two are closest, though I'd rather call it the data, than the evidence, although I'll buy that last thing. But what it is not is "the probability for the CA case".
Well in the context of discussing likelihood ratios I abbreviated the conditional probability. That was hasty, clumsy and understandably confusing. You asked about it I clarified it unambiguously. But then for comment after comment after comment after comment you continued to refer to that initial phrase, as though that pretty much summed it up. You continued to point out how it was a false statement. And each time I continued to remind you of my clarification. But that didn’t seem to matter. And now, even when you finally acknowledge this, you still include that initial comment. You can’t say I didn’t try.
which you "clarified" as: the probability for the CA case
No, there you go again. I did not clarify it that way, but apparently you didn’t read beyond the first sentence. Here was my clarification, yet once again:
I thought that was clear, but my summary was not. To be clear, “In other words, the worse the conditional probability of the evidence given CA, the better the case for CA, because the conditional probability of the evidence given SA got even worse yet.”
However, given that it isn't what you meant, then why are you concerned with a low p(data|CA) when we actually have that data? Why worry on Tuesday that the forecast said it might rain on Monday, when in fact, it did not?
The low p(data|CA) is a good indicator of the metaphysics at work. Again, the key is that the p(data|SA) is even lower. That’s how CA wins. It is not that the data are highly probable given CA. They’re not. But the data are ridiculously improbable on the alternative, SA. And since CA and SA are the only possibilities, CA wins. It all hinges on CA and SA being the only possibilities. It’s all metaphysical.
CH:Well it’s all well and good to say “I’m just a scientist” but you’re the one who claimed the evidence for CA is overwhelming and who provided posterior probabilities for CA based on a non scientific premise about CA and SA being the alternatives. Both are claims are outside of science.
DeleteCornelius you are again equivocating between the specific and the general.
Yes, I (this is a non-scientific statement) consider the argument for common ancestry of extant species overwhelming. However, that is a very loose statement,because I am aware of various horizontal mechanisms of heritability that also operate making "common ancestry" too simplistic a claim.
What I calculated was a hypothetical posterior probability for the common ancestry of two hypothetical species sharing a hypothetical rare sequence (i.e. Sober's example). I demonstrated that the posterior probability of CA increases as p reduces. I also agreed with you that this is fallacious, but argue that this is because it assumes a non-excluded middle, without justification (and we know for sure that there is a middle that was excluded).
My objection is not to your conclusion (that Sober's likelihood and my posterior is generous to CA; it is), but with your contention that this generosity is a) due to "contrastive reasoning" per se and b) anything to do with the the smallness of the numerator of the likelihood ratio.
The fallacy lies simply in the exclusion of possible but unknown middles.
I'll have to reply to the rest of your responses later.
CH:You continued to point out how it was a false statement. And each time I continued to remind you of my clarification. But that didn’t seem to matter. And now, even when you finally acknowledge this, you still include that initial comment. You can’t say I didn’t try.
DeleteNo, and thanks. I thought you were still holding to your second version ("the probability for the CA case"), and thought you thought it was somehow equivalent to "the conditional probability of the evidence given CA". I won't mention it again!
CH:The low p(data|CA) is a good indicator of the metaphysics at work.
This is where you are still wrong. This has nothing to do with "the metaphysics at work", or,at any rate, has nothing to do with the fallaciousness of the reasoning.
Again, the key is that the p(data|SA) is even lower. That’s how CA wins. It is not that the data are highly probable given CA. They’re not. But the data are ridiculously improbable on the alternative, SA. And since CA and SA are the only possibilities, CA wins.
Although you are now using a correct definition of the numerator, you are still laboring under the misapprehension that the fact that the data are "improbable" means that CA is also "improbable", just less "improbable" than SA. It doesn't matter what the prior probability of the data were: we actually have them in our hands. Their posterior probability is 1!
However, this statement is correct:
It all hinges on CA and SA being the only possibilities. .
Yes. But all you need to do to fix the problem is to redo the hypotheses so that they are legitimately binary.
I suggested that a while back, but let me suggest it again: Instead of SA, let us have "IM" for "independent mutations" and instead of "CA" let's have "NM" for "non-independent mutations".
Now, legitimately, p(NM)=1-p(IM).
Now let's run the likelihood and Bayes calcs again. We get exactly the same answer (except that we won't be as precise about the value of p(data|NM), but we still know it's greater than p(data|IM)) and, hey presto, we get a larger posterior probability and a larger likelihood ratio for NM when p is small than when p is large, just as we got it for CA before.
There is now no fallacy, no excluded middle. The numerator of the likelihood ratio is still tiny, but the posterior probability of NM is now approaching 1. We have still used "contrastive reasoning".
But we now have a legitimate conclusion for the very simple reason that we no longer have an excluded middle.
We can now say, with justified confidence, that some non-independent factor led to the appearance of M in both population. We just can't claim that that factor is Common Ancestry, at least not as it stands.
As I keep saying, excluding a middle creates a real problem. You are right that there is a problem. You are right that it arises from an excluded middle (the "metaphysical" assumption that there are only two alternatives, when in fact there are more). But you are wrong when you say it has anything to do with "contrastive reasoning" and wrong when you point to the low value of the numerator as "a good indicator of the metaphysics at work". As my example shows, a low value of that numerator is perfectly consistent with a perfectly legitimate conclusion.
All that matters is that we do not exclude a middle.
As Dembski does.
Are you going to confront the Dembski issue?
EL:
DeleteYes, I (this is a non-scientific statement) consider the argument for common ancestry of extant species overwhelming. However, that is a very loose statement, because I am aware of various horizontal mechanisms of heritability that also operate making "common ancestry" too simplistic a claim.
But the species do not fall into a pattern like this. With HGT you could account for some genes showing up in odd places, but that is not the observed pattern. It is not as though there is a CA pattern with some odd genes here and there.
What I calculated was a hypothetical posterior probability for the common ancestry of two hypothetical species sharing a hypothetical rare sequence (i.e. Sober's example). I demonstrated that the posterior probability of CA increases as p reduces. I also agreed with you that this is fallacious, but argue that this is because it assumes a non-excluded middle, without justification (and we know for sure that there is a middle that was excluded).
What you are calling fallacious is fundamental to evolutionary reasoning. It runs all through the literature, from Darwin to Coyne and everywhere in between. All the proofs for evolution spring from it, in one way or another. That’s why Sober gives it so much attention.
My objection is not to your conclusion (that Sober's likelihood and my posterior is generous to CA; it is), but with your contention that this generosity is a) due to "contrastive reasoning" per se and …
Well my point is that contrastive reasoning as used in evolutionary thought is metaphysical. If there is a way to use contrastive reasoning without the metaphysics, then great. I’m not making a universal claim that all contrastive reasoning necessarily is metaphysical. That’s not my point. I’m focusing on evolutionary thought.
b) anything to do with the smallness of the numerator of the likelihood ratio.
Yes, you have objected to this several times, which seems strange since Sober and the math make this quite clear. With the CA and SA models, the smaller the p(data|CA), the better the case for CA. I still don’t understand why you think this is false.
Continued …
Continued …
DeleteCH: The low p(data|CA) is a good indicator of the metaphysics at work.
EL: This is where you are still wrong. This has nothing to do with "the metaphysics at work", or, at any rate, has nothing to do with the fallaciousness of the reasoning.
No, this is where you are still wrong, and I think I’m seeing why, so I think I can explain. Imagine a case where this likelihood ratio approach (or your Bayesian calculation) was scientific. That is, the two theories were known from science to be the only two possibilities. For instance, let’s say that the CA and SA models were known to be the only alternatives. Then there would be no metaphysical assumptions at play, and the low p(data|CA) would be no problem. I suspect this is the way you are thinking of it.
But even in this case, CA wins not with data which has high p(data|CA), but rather low p(data|CA). And the lower the better. The reason for this is that as p(data|CA) reduces, p(data|SA) reduces at an even faster rate. So by finding those unlikely data, you drive up the case for CA dramatically. You can see this to be true if you just look at the models, and the math.
You need to understand the point that the reason why the low p(data|CA) is used is due to how the problem is set up. It is due to the structure of the CA and SA models, and the fact that they are the only alternatives.
But now switch back to the actual case. The assumption that these particular two models, CA and SA, are the only alternatives, is a metaphysical assumption. So reason why the low p(data|CA) is used, and drives up the case for CA so well, is metaphysical. You should be able to see this easily in the math.
I won't mention it again! …
Although you are now using a correct definition of the numerator
There you go again.
you are still laboring under the misapprehension that the fact that the data are "improbable" means that CA is also "improbable",
No, I didn’t say that either.
Yes. But all you need to do to fix the problem is to redo the hypotheses so that they are legitimately binary.
Well don’t tell me, tell the evolutionists.
I suggested that a while back, but let me suggest it again: Instead of SA, let us have "IM" for "independent mutations" and instead of "CA" let's have "NM" for "non-independent mutations".
Now, legitimately, p(NM)=1-p(IM).
No, you are still making metaphysical assumptions. Science does not know that bad design X arose either from NM or from IM.
We can now say, with justified confidence, that some non-independent factor led to the appearance of M in both population.
This is just more evolutionary metaphysics at work.
Are you going to confront the Dembski issue?
I didn’t want to move to another topic without resolving these basic evolutionary metaphysics.
EL:
DeleteCH: But now switch back to the actual case. The assumption that these particular two models, CA and SA, are the only alternatives, is a metaphysical assumption. So reason why the low p(data|CA) is used, and drives up the case for CA so well, is metaphysical. You should be able to see this easily in the math.
BTW, another way to see this is to think of the plagiarism example. Same thing. You see a share typo and you suspect cheating. Then you see an entire erroneous (completely random) sentence that is shared and your suspicions go through the roof. It is not because the data is more probable on cheating. It actually is less. But the data are far less probable on not cheating. The structure of the problem is what is making the shared errors (low probability data) so informative. You go to the low probability data because of the way the problem is structured. So in the evolution case, as I said, it is an indicator of the metaphysics at work.
CH: But the species do not fall into a pattern like this. With HGT you could account for some genes showing up in odd places, but that is not the observed pattern. It is not as though there is a CA pattern with some odd genes here and there.
DeleteLet's stick with the issue of contrastive reasoning for now. I was not making a general case for (or against) CA in these posts.
CH:What you are calling fallacious is fundamental to evolutionary reasoning. It runs all through the literature, from Darwin to Coyne and everywhere in between. All the proofs for evolution spring from it, in one way or another. That’s why Sober gives it so much attention.
No, it is not fundamental to "evolutionary reasoning". Contrastive reasoning is fundamental to hypothesis testing, and the construction of the correct null is important. If you do not construct your null correctly you get an invalid inference. Not constructing the correct null is not fundamental to "evolutionary reasoning". It does tend, however, to be fundamental to ID reasoning. Which is why I keep asking you about Dembski. However, what a Bayesian approach does is allows you to do is to make your priors explicit. That may be what you don't like. Certainly evolutionists set their priors on the evolutionary framework being broadly correct high.
CH:Well my point is that contrastive reasoning as used in evolutionary thought is metaphysical. If there is a way to use contrastive reasoning without the metaphysics, then great. I’m not making a universal claim that all contrastive reasoning necessarily is metaphysical. That’s not my point. I’m focusing on evolutionary thought.
I still don't know what you mean by "metaphysical", but I'm taking it to mean something to do with not considering enough alternatives (excluding a middle). But I'm glad you agree that the problem doesn't lie in contrastive reasoning per se.
CH: Yes, you have objected to this several times, which seems strange since Sober and the math make this quite clear. With the CA and SA models, the smaller the p(data|CA), the better the case for CA. I still don’t understand why you think this is false.
It isn't false. What is misleading, and from your responses seems to have misled you, is the idea that someone because the probability of the thing on the numerator is small, somehow both hypotheses are unlikely, but the one on the bottom is even more so. As long as you are absolutely clear that this is not the case, I'll stop banging on about it. But if you are clear about that, then don't keep implying that it's only because the denominator is "worse" that the numerator, though bad, it is OK.
continued...
DeleteCH:No, this is where you are still wrong, and I think I’m seeing why, so I think I can explain. Imagine a case where this likelihood ratio approach (or your Bayesian calculation) was scientific. That is, the two theories were known from science to be the only two possibilities. For instance, let’s say that the CA and SA models were known to be the only alternatives. Then there would be no metaphysical assumptions at play, and the low p(data|CA) would be no problem. I suspect this is the way you are thinking of it.
Yes, I'm glad you "suspect" it because I stated it explicitly, with a specific example!
Cool.
But even in this case, CA wins not with data which has high p(data|CA), but rather low p(data|CA). And the lower the better.
Yes, but what is that "even" doing in there? Why "even in this case"? Why the "But". It's just: "CA wins not with...."etc.
You still seem drawn to the idea that there is some kind of "trick" (your word), about the strength of the case for CA going up as p(data|CA) goes down.
The reason for this is that as p(data|CA) reduces, p(data|SA) reduces at an even faster rate. So by finding those unlikely data, you drive up the case for CA dramatically. You can see this to be true if you just look at the models, and the math.
Indeed. It's why I keep presenting the math :)
Continued....
DeleteYou need to understand the point that the reason why the low p(data|CA) is used is due to how the problem is set up. It is due to the structure of the CA and SA models, and the fact that they are the only alternatives.
Yes, and I do understand that. That's why I keep on banging on about p(A)=1-p(B) (where A and B are alternate hypothesis If this is valid (there is no excluded middle), your posterior probability for A is valid. If it isn't, it isn't.
CHBut now switch back to the actual case. The assumption that these particular two models, CA and SA, are the only alternatives, is a metaphysical assumption.
Well, it's an invalid assumption. As I pointed out.
So reason why the low p(data|CA) is used, and drives up the case for CA so well, is metaphysical.
Aaaaaarrrrrgghhhhhh! No!!!!!
Sheesh, Cornelius, you are so close, but you keep going back to this totally irrelevant point!
There is no "So..." there. Whether or not it was true that p(A)=1-p(B) you will still get a larger posterior probability for A if p(data|A) is small. The "reason why the p(data|CA) is used" has nothing to do with wheter p(A)=1-p(B) or not. The size of p(data|CA) (or the original "p") is totally orthogonal to the issue of whether the assumption that p(A)=1-p(B) is valid or not.
You should be able to see this easily in the math.
Yes, and I have laid out the math myself several times!
I'm not disputing the math, I'm simply pointing out that the problem you are laying (erroneously in my view, but let that pass for now) at the door of "evolutionary reasoning" has NOTHING to do with the fact that "contrastive reasoning" is used NOR has it to do with the fact that the data have low probability. It simply has to do with whether or not the hypothesis testing set up has an excluded middle. No more, no less.
EL: Now, legitimately, p(NM)=1-p(IM).
CH: No, you are still making metaphysical assumptions. Science does not know that bad design X arose either from NM or from IM.
I didn't say it did!!!! All I'm saying is that if you define one hypothesis as being all hypotheses not covered by the other (after all, a thing can't be independent and non-independent at the same time) then, necessarily p(Independent Mutations)=1-p(Non-independent Mutations). Clearly under the "Non-independent Mutations" covers both CA and ID, as well as HGT, and probabily some we haven't thought of. But I have made NO "metaphysical assumptions" here - I have simply ensured that there is no excluded middle. A thing cannot be A and not-A, right? Wasn't that what a whole bunch of people were expelled for UD for querying?
CH:This is just more evolutionary metaphysics at work.
No. It. Is. Not.
Please read my post again carefully. What I have demonstrated is that IF we carefully operationalise our hypotheses so that there is NO excluded middle (in other words "NM" in effect becomes the Null hypothesis), THEN we can use exactly the same contrastive reasoning and we can compute exactly the same likelihood ratio and exactly the same posterior probability, where the posterior probability approaches 1 as "p" goes down, and yet we have committed NO fallacy, nor does the fact that p(data|NM) is tiny have any bearing on the fact that the evidence for it is overwhelming.
The difference of course, is that we can no longer conclude "common ancestry". All we can do is say that we have falsified Independent Mutations.
There may be a great many alternative hypotheses that also involve non-independence, including, as I said CA, HGT and ID. Also Craig Venter.
EL:
DeleteCH:What you are calling fallacious is fundamental to evolutionary reasoning. It runs all through the literature, from Darwin to Coyne and everywhere in between. All the proofs for evolution spring from it, in one way or another. That’s why Sober gives it so much attention.
EL: No, it is not fundamental to "evolutionary reasoning".
It is what makes evolution compelling. It is what makes evolution a fact, or the evidence overwhelming, or however you want to put it. Without it evolution is unlikely and so would have been dropped a long time ago (actually never hypothesized in the first place). I call that pretty fundamental.
I still don't know what you mean by "metaphysical", but I'm taking it to mean something to do with not considering enough alternatives (excluding a middle).
Well I have explained this. The metaphysics, in this example we are discussing, lies in the premise that we know the set of possible explanations. That doesn’t come from science.
What is misleading, and from your responses seems to have misled you, is the idea that someone because the probability of the thing on the numerator is small, somehow both hypotheses are unlikely, but the one on the bottom is even more so. As long as you are absolutely clear that this is not the case, I'll stop banging on about it.
Promise? After all, I never said that.
CH: But now switch back to the actual case. The assumption that these particular two models, CA and SA, are the only alternatives, is a metaphysical assumption.
EL: Well, it's an invalid assumption. As I pointed out.
Well actually it is not “invalid.” It is “metaphysical.” Invalid means there is a fallacy in the logic. There is no fallacy in the assumption that CA and SA are the only two alternatives. It simply requires the belief that that is true, which is a premise that does not come from science. Given that premise, there is no fallacy.
Continued …
Continued …
DeleteCH: So reason why the low p(data|CA) is used, and drives up the case for CA so well, is metaphysical.
EL: Aaaaaarrrrrgghhhhhh! No!!!!! Sheesh, Cornelius, you are so close, but you keep going back to this totally irrelevant point!
There is no "So..." there. Whether or not it was true that p(A)=1-p(B) you will still get a larger posterior probability for A if p(data|A) is small. The "reason why the p(data|CA) is used" has nothing to do with wheter p(A)=1-p(B) or not. The size of p(data|CA) (or the original "p") is totally orthogonal to the issue of whether the assumption that p(A)=1-p(B) is valid or not.
Wow, this is good. We’re getting somewhere. Your statement here is not right, and I think it reveals the source of difference.
Contrary to what you claim, p(CA)=1-p(SA) is crucial. First, without p(CA)=1-p(SA) you couldn’t even formulate your equation. That Bayesian calculation you did wouldn’t even be possible. You have to have mutually exclusive and exhaustive hypotheses to even use that approach you used. So right off the bat we have a problem with your claim that “Whether or not it was true that p(A)=1-p(B) you will still get a larger posterior probability for A if p(data|A) is small.”
Now perhaps you were referring to the possibility of more hypotheses, such as a third hypothesis, C. In that case, yes, of course you could formulate and compute your posterior probability for A. But in that case, you still need a model for C, and the posterior probability for A will hinge on that model for C. You will *not* obtain a larger posterior probability for A, when p(data|A) is small, if, for example, p(data|C) is large. In that case your posterior probability for A drops precipitously.
It all depends on the structure of your problem, and that is where the metaphysics comes in.
I didn't say it did!!!! All I'm saying is that if you define one hypothesis as being all hypotheses not covered by the other (after all, a thing can't be independent and non-independent at the same time) then, necessarily p(Independent Mutations)=1-p(Non-independent Mutations).
On the one hand, you want to claim your hypotheses are mutually exclusive and exhaustive, but on the other hand you want to use hypotheses (Independent Mutations and Non-independent Mutations) which do not satisfy this without metaphysics.
EL:
DeletePlease read my post again carefully. What I have demonstrated is that IF we carefully operationalise our hypotheses so that there is NO excluded middle (in other words "NM" in effect becomes the Null hypothesis), THEN we can use exactly the same contrastive reasoning and we can compute exactly the same likelihood ratio and exactly the same posterior probability, where the posterior probability approaches 1 as "p" goes down, and yet we have committed NO fallacy …
Gosh no, not at all, you’re way deep in the metaphysics. For evolution, this is not possible within science. If you define your hypotheses as mutually exclusive and exhaustive then you cannot formulate models for them (without metaphysical assumptions). Conversely, if you formulate models, then they are not mutually exclusive and exhaustive, again without metaphysics.
Cornelius, you really can't be reading my posts!
DeleteIt was *I* who pointed out that: "p(CA)=1-p(SA) is crucial".
I've pointed it out in almost every post! Indeed I think I was the first of the two of us to type it!
And yet you say:
Contrary to what you claim, p(CA)=1-p(SA) is crucial.
wtf?
Yes, it's crucial. It's the only thing that is crucial. Contrastive reasoning isn't crucial (it's all we've got). Low p(data|H) isn't crucial, it doesn't matter how low it is.
All that matters is that we don't exclude a middle.
What is it about me saying that that makes you think I'm saying the opposite? It's not as though I haven't said it a gazillion times!
And, if you define one of your two alternate hypotheses as the Null of the other (as I have said, again a gazillion times) then you can set up your Bayesian equation just fine, as I laid out in my example. It's just that then your conclusions are much more limited. It would, in my example, allow us to conclude that the probability of Independent Mutations was vanishingly low, it just wouldn't allow us to distinguish between Non-independent Mutation hypotheses.
CH:Gosh no, not at all, you’re way deep in the metaphysics. For evolution, this is not possible within science. If you define your hypotheses as mutually exclusive and exhaustive then you cannot formulate models for them (without metaphysical assumptions).
Of course you can. We do it all the time. That's the whole principle of the Null hypothesis. Nothing "metaphysical" about it.
CH:Conversely, if you formulate models, then they are not mutually exclusive and exhaustive, again without metaphysics.
If what you were saying were true, all hypothesis testing would be impossible, Fisherian, Bayesian, Likelihoodian.
As I demonstrated in my example, although we cannot give a precise estimate for p(data|NM), we can assume it is larger than p(data|IM). Therefore, if we observe [data] then we have evidence for NM.
Of course it is perfectly true that if we are obliged to include ID models, and if ID models are completely non-predictive, then of course we can conclude nothing.
That's where priors come in (and metaphysics!)
But I am perfectly happy to add, as a caveat, to any evolutionary conclusion: barring miracles, or Last Thursdayism, the data suggest X.
We simply cannot, ever, rule out a hypothesis that can account for any data.
But that's ID's problem, not ours. As soon as ID can produce a predictive model, we can plug it into our analyses.
In the mean time: how about Dembski's excluded middle?
Dr Hunter;
DeleteThe assumption that these particular two models, CA and SA, are the only alternatives, is a metaphysical assumption.
No. It's just a test of two models. Can't someone test two models contrastively without assuming completeness?
Without a competing alternative model, there's really nothing further to discuss.
Pedant:
DeleteCH: The assumption that these particular two models, CA and SA, are the only alternatives, is a metaphysical assumption.
P: No. It's just a test of two models.
If you assume CA and SA are the only alternatives, then that is not merely a test of two models.
Can't someone test two models contrastively without assuming completeness?
Sure, but then you can’t make the conclusions evolutionists make. You lose your fact status, evolution is then exposed to the science, and would be seen for what it is.
EL:
DeleteCornelius, you really can't be reading my posts!
It was *I* who pointed out that: "p(CA)=1-p(SA) is crucial".
I've pointed it out in almost every post! Indeed I think I was the first of the two of us to type it!
And yet you say:
Contrary to what you claim, p(CA)=1-p(SA) is crucial.
As I pointed out, I was responding to your claim that “Whether or not it was true that p(A)=1-p(B) you will still get a larger posterior probability for A if p(data|A) is small.” That claim is false.
CH:Gosh no, not at all, you’re way deep in the metaphysics. For evolution, this is not possible within science. If you define your hypotheses as mutually exclusive and exhaustive then you cannot formulate models for them (without metaphysical assumptions).
EL: Of course you can. We do it all the time. That's the whole principle of the Null hypothesis. Nothing "metaphysical" about it.
Yes, I certainly agree that evolutionists do it all the time. Unfortunately that doesn’t mean that assuming “it’s either CA or SA” isn’t a metaphysical claim. Of course it is. To claim that the species arose via CA or SA, or IM or NM, is prima facie metaphysical. These simply do not come from science, no matter how many times you “do it.”
If what you were saying were true, all hypothesis testing would be impossible, Fisherian, Bayesian, Likelihoodian.
No, we’re talking about evolution, not “all hypothesis testing.”
As I demonstrated in my example, although we cannot give a precise estimate for p(data|NM), we can assume it is larger than p(data|IM). Therefore, if we observe [data] then we have evidence for NM.
Your argument was that NM and IM are mutually exclusive and exhaustive, which is a metaphysical claim.
But I am perfectly happy to add, as a caveat, to any evolutionary conclusion: barring miracles, or Last Thursdayism, the data suggest X.
We simply cannot, ever, rule out a hypothesis that can account for any data.
I’m afraid you’re not helping your case. The fact that you’re deep in the metaphysics isn’t remedied by any “barring miracles” caveat. Scientism doesn’t obviate metaphysics, it just adds more fuel to the fire.
CH:Sure, but then you can’t make the conclusions evolutionists make. You lose your fact status, evolution is then exposed to the science, and would be seen for what it is.
DeleteBut it's only you, Cornelius, who keeps insisting that scientists regard evolution as a "fact" in a "metaphysical" sense.
I don't, for one. For that matter, I don't think any scientific model is a "fact" in a "metaphysical sense". Ultimately, the test of a scientific model is its predictive power. When a model is enormously predictive, some people call it a "fact". I don't, but I don't accuse those who do doing metaphysics where no metaphysics should be done. Just of shorthand.
So I'd like you to accusing me of being "deep in metaphysics". Which appears to be code for "closed-minded". I'm not, and nor are any scientists that I know. If they were, they would be unable to do science.
CH:Your argument was that NM and IM are mutually exclusive and exhaustive, which is a metaphysical claim.
No, it is not.
Unless "A thing cannot be both A and not-A" is a metaphysical claim". Do you think it is?
That should of course read "I'd like you to stop accusing me of...."
DeleteCH:Your argument was that NM and IM are mutually exclusive and exhaustive, which is a metaphysical claim.
DeleteEL: No, it is not.
Unless "A thing cannot be both A and not-A" is a metaphysical claim". Do you think it is?
Prezactly. Common ancestry and separate ancestry are mutually exclusive and exhaustive. That’s why DNA testing is used in paternity disputes. As Elizabeth noted earlier, that will hold until they repeal the law of excluded middle.
"As Elizabeth noted earlier, that will hold until they repeal the law of excluded middle."
DeleteAt which point, the gates of Uncommon Descent will be reopened ;)
Elizabeth, were you going to present a scientific test for evolution with clear and measurable objectives? You pick the test.
DeleteNeal, evolution is not one single hypothesis. It is a broad explanatory framework that encompasses many sub-theories, from which we derive testable hypotheses.
DeleteWhich particular theory or hypothesis did you want me to present a test for?
Elizabeth said, "evolution is not one single hypothesis. It is a broad explanatory framework that encompasses many sub-theories, from which we derive testable hypotheses"
Delete--
Yes, of course, pick the one that you are the most familar with and is a good representation.
OK, well the one of the nicest early experiments was Endler's guppie experiment, which showed experimentally that a population will adapt to its environment by natural selection. The Grants' Galapagos finch study also showed this in an observational study.
DeleteSo that demonstrates, definitively, that Darwin's mechanisms of adaptation by natural selection actually works in biological populations.
Elizabeth, Darwinism isn't the theory of how existing population demographics oscillate back and forth from the mean through natural selection. But, if it were, your test would have validity. Nothing that you mentioned gives me any reason whatsoever to believe that natural selection can produce definitively new species, new organs or body plans. It doesn't show that the variations are permanent.
DeleteCan you give me another test that would give me reason to believe that new species, new organs and body plans are produced via natural selection?
EL:
DeleteSo I'd like you to stop accusing me of being "deep in metaphysics". Which appears to be code for "closed-minded".
No, those are two different things. I noted that you’re approach relies on premises that don’t come from science. That says nothing about the truth value of your premises, being closed minded, and so forth.
CH:Your argument was that NM and IM are mutually exclusive and exhaustive, which is a metaphysical claim.
EL: No, it is not. Unless "A thing cannot be both A and not-A" is a metaphysical claim". Do you think it is? […]
Pedant: Prezactly. Common ancestry and separate ancestry are mutually exclusive and exhaustive. That’s why DNA testing is used in paternity disputes. As Elizabeth noted earlier, that will hold until they repeal the law of excluded middle.
EL: "As Elizabeth noted earlier, that will hold until they repeal the law of excluded middle."
Oh gosh no. Science cannot tell you that the two hypotheses: CA and SA, or "independent mutations" (IM) and "non-independent mutations" (NM), are mutually exclusive and exhaustive. This assumption comes from “above” the science, “meta” science, or meta physics. I can elaborate if that isn’t clear.
EL:
DeleteAs I pointed out earlier, it’s probably worth repeating that if you define two competing evolutionary hypotheses as mutually exclusive and exhaustive then you cannot formulate models for them (unless you are employing metaphysical assumptions). Conversely, if you do formulate models, then they cannot be mutually exclusive and exhaustive (again, unless you are employing metaphysical assumptions). This is a simple way of seeing that evolutionary thinking is metaphysical, such as in Sober’s elucidation of it.
Neal: Elizabeth, Darwinism isn't the theory of how existing population demographics oscillate back and forth from the mean through natural selection. But, if it were, your test would have validity. Nothing that you mentioned gives me any reason whatsoever to believe that natural selection can produce definitively new species, new organs or body plans. It doesn't show that the variations are permanent.
DeleteCan you give me another test that would give me reason to believe that new species, new organs and body plans are produced via natural selection?
This is why I asked you to specify a theory or hypothesis.
Yes, we have observed speciation, in real time.
We have also observed, in the fossil record, incremental changes that have transformed one body part (e.g. a jaw) to another (part of the mammalian ear).
There are also experiments that show what happens when you transplant hox genes from one species to a very different species.
All these studies are tests of specific evolutionary hypotheses.
But you seem to want One Big Test that will prove that Evolution is True. That's not how science is done.
CH:As I pointed out earlier, it’s probably worth repeating that if you define two competing evolutionary hypotheses as mutually exclusive and exhaustive then you cannot formulate models for them (unless you are employing metaphysical assumptions).
DeleteSo all null hypothesis testing employs "metaphysical assumptions"?
Really?
And when are you going to tackle the elephant-in-the-room which is Dembski's Explanatory Filter?
CH:Conversely, if you do formulate models, then they cannot be mutually exclusive and exhaustive (again, unless you are employing metaphysical assumptions). This is a simple way of seeing that evolutionary thinking is metaphysical, such as in Sober’s elucidation of it.
Cornelius, please go through my post carefully, in which I explained how two mutually exclusive hypothesis (Independent and Nonindependent Mutations) could both be modeled.
Please also note that the conclusion from such a test is much more nuanced than the conclusion that "Common Descent" is probable. All we can conclude is that "Non-Independent mutations" is probable. We've ruled something out (Independent mutations) but we have kept in everything else.
But I have to ask: have you never constructed a null hypothesis? Do you really want to throw out the whole of Fisherian statistics?
And a lot of Bayes as well?
What will you use instead?
CH: Oh gosh no. Science cannot tell you that the two hypotheses: CA and SA, or "independent mutations" (IM) and "non-independent mutations" (NM), are mutually exclusive and exhaustive. This assumption comes from “above” the science, “meta” science, or meta physics. I can elaborate if that isn’t clear.
DeleteYou are right: science doesn't tell us that two hypotheses are mutually exclusive. LOGIC tells us that. Common ancestry and separate ancestry are mutually exclusive because there are no other possibilities. You are either a descendant of your parents or adopted; there are no other possibilities.
If you want to claim that logic is metaphysics, you may do so, but that gets you nowhere with respect to the [Sober] argument.
Elizabeth said, "Yes, we have observed speciation, in real time."
Delete--
I asked this before and all you could come up with with incipent speciation which is not actual speciation, but speculation about what could happen in the future if one believed in evolution. Otherwise you wouldn't have to preface the term speciation. Evolutionists always need to fudge their results.
Elizabeth said, "We have also observed, in the fossil record, incremental changes that have transformed one body part (e.g. a jaw) to another (part of the mammalian ear)."
--
The fossil record is a historical interpretion. Evolutionists cherry pick fossils from the vast mosiac of life forms while under the heavy spell of confirmation bias. If you only focus on the cherries you can prove anything. It's all the other stuff that real science would take as nothing but falsification... Stasis, extinction, sudden appearance are what is found in the fossil record. Gradualism is really not there other than what we observe today... minor variation within bounds.
But you did not specifically define any tests with objective and measurable results to make me believe that natural selection is capable of producing new species, organs or body plans. You've got a few bits and pieces, but any idea can muster together such support.
As far a real science, I do not see where anything you said would qualify as following scientific methodology in supporting evolution.
Evolutionists possess no criteria for falsifying their theory objectively.
EL:
DeleteSo all null hypothesis testing employs "metaphysical assumptions"? Really?
No, I didn’t say that. You’ve responded with this twice now, in spite of my explanation. Problems about different colored balls in an urn are not the same the origin of the world. More below …
Cornelius, please go through my post carefully, in which I explained how two mutually exclusive hypothesis (Independent and Nonindependent Mutations) could both be modeled.
Please also note that the conclusion from such a test is much more nuanced than the conclusion that "Common Descent" is probable. All we can conclude is that "Non-Independent mutations" is probable. We've ruled something out (Independent mutations) but we have kept in everything else.
I did go through your post carefully. You did not merely assume that IM and NM are mutually exclusive. You also assumed they are exhaustive. That assumption does not come from science.
But I have to ask: have you never constructed a null hypothesis? Do you really want to throw out the whole of Fisherian statistics? And a lot of Bayes as well?
No, I said no such thing. You’re confusing statistical approaches with the assumptions that you insert into them. When you tell the method that IM and NM are mutually exclusive and exhaustive, then it will give you an answer. There’s nothing wrong with the method, but metaphysics-in, metaphysics-out.
What will you use instead?
I wouldn’t make metaphysical assumptions and then deny it, or justify it because “We do it all the time.” Yes, you “do it all the time,” and that is important to understand. That is why evolutionary thought is metaphysical.
Pedant:
DeleteYou are right: science doesn't tell us that two hypotheses are mutually exclusive. LOGIC tells us that.
Did you forget your claim they are exhaustive, or did the goal posts just move?
Common ancestry and separate ancestry are mutually exclusive because there are no other possibilities.
No, that is false. If there are “no other possibilities” then that would mean CA and SA are exhaustive.
Of course, they are both mutually exclusive and exhaustive.
DeleteFeel free to challenge the logic by suggesting other possibilities.
Pedant:
DeleteOf course, they are both mutually exclusive and exhaustive. Feel free to challenge the logic by suggesting other possibilities.
Your claim is that CA and SA as explanations for the origin of species, or IM and NM, are logically mutually exclusive and exhaustive. That there are no metaphysical assumptions involved. If that were true you could make a logical argument to back it up.
But of course that is not true. Unless you know of some logical argument no one else knows of. The problem is that, unlike you, science does not know everything. So, in particular, science does not know the set of all possible explanations for how the world arose. You can’t just pick out your favorite two explanations and decide with no justification that they are mutually exclusive and exhaustive.
If you define IM and NM to be mutually exclusive, and if you say that the species had to have arisen either by IM or NM, and so therefore IM and NM are both mutually exclusive and exhaustive, then yes, you are correct. But that’s not science. You can’t just dictate with no justification the set of all possible explanations for how the world arose.
This is by no means a minor issue. The problem of unconceived alternatives has plagued science for centuries. Here the problem has become rather acute as evolutionists mandate their metaphysics, insist they did no such thing as EL is doing, and proclaim that their scientifically absurd idea is a fact.
Elizabeth Liddle
DeleteBut you seem to want One Big Test that will prove that Evolution is True. That's not how science is done.
Yep. Tedford is demonstrating exactly what I described above. Ignorant Creationists always demand to see THE test, or THE piece of evidence. It's like their tiny closed minds can't deal with more than one simple thought at a time.
ToE doesn't rest on one piece of evidence, or even ten pieces. It's based on millions of independent observations. As such, falsifying any one piece doesn't falsify all the others.
CH:I did go through your post carefully. You did not merely assume that IM and NM are mutually exclusive. You also assumed they are exhaustive. That assumption does not come from science.
DeleteAre you going to tell Barry Arrington that you do not accept the Law of Non-Contradiction? That a thing can be A and not-A? That A and not-A is not "exhaustive"?
The assumption certainly does not come from science. It comes from classical logic.
I would tend to agree with you that classical logic does not always hold sway in science (in quantum physics for example), but the reason UD is so quiet these days is that Barry banned a whole swathe of people for making that point. Should he have banned you too?
This is why I keep bringing up the concept of the null hypothesis, and why I'd like you to tackle Dembski's (fallacious) use of it.
A null hypothesis is the complement of your study hypothesis, and in traditional null hypothesis testing, your support for your study hypothesis is always indirect, in that you "reject the null". It does not tell you that your study hypothesis is probably true, it merely tells you that the null is probably false (in that p(data|null) is small. But by rejecting the null, you leave the way open for all possible hypotheses other than the null.
That is why, in science, as soon as we "reject a null" we start looking for alternative hypotheses to our study hypothesis that could also account for the data.
In Bayesian inference, if our alternative hypothesis is also a null hypothesis (i.e. complementary to the study hypothesis) then the same applies.
However, a Bayesian approach also allows us to compare two non-complementary hypotheses, in which case, as you point out, the probability of both may be small, but one may be substantially more likely than the other. This is where our priors come in useful, but your posterior probabilities are of course only as good as your priors!
But then your beef should be, not with "contrastive reasoning" but with Bayesian inference and the concept of priors. And you may have a point. Perhaps Bayesian priors are what you call "metaphysical" (though I would strongly argue that they are not).
To return to Dembski: Demski infers "Design" because he has rejected a null. This is fallacious, for reasons I have given above, and for reasons with which you seem to agree. Dembski's null and "Design" are not exhaustive.
This is not because of an intrinsic problem with null hypothesis testing (although there is an intrinsic problem with Fisherian hypothesis testing, which he uses, but that's not the primary problem here). It's because he has not specified his null in a way that its rejection permits him to reject evolutionary processes.
In other words he has excluded a middle. He has regarded "equiprobable independent events" as the complement of "Design" when it clearly is not. p(equiprobable independent events) does not = 1-p(Design).
So you seem to have conflicts to resolve with both Barry Arrington and William Dembski :)
Good luck.
CH to me: Your claim is that CA and SA as explanations for the origin of species, or IM and NM, are logically mutually exclusive and exhaustive. That there are no metaphysical assumptions involved. If that were true you could make a logical argument to back it up.
DeleteBut of course that is not true. Unless you know of some logical argument no one else knows of.
For the umpteenth time: THE LAW OF NONCONTRADICTION (You can look it up.) As Elizabeth just said:
EL to CH: Are you going to tell Barry Arrington that you do not accept the Law of Non-Contradiction? That a thing can be A and not-A? That A and not-A is not "exhaustive"?
The assumption certainly does not come from science. It comes from classical logic.
EL and I are not making an empirical claim, despite your misapprehension. It's a claim about logical alternatives.
I'm looking forward to your response to her request for the resolution of your conflict with Arrington and Dembski.
EL:
DeleteAre you going to tell Barry Arrington that you do not accept the Law of Non-Contradiction? That a thing can be A and not-A? That A and not-A is not "exhaustive"?
Lizzie I can’t respond to a non sequitur. Furthermore, this is ironic because, if you look at what you have been saying, you actually are not following the Law of Non-Contradiction. Earlier you agreed that the origin of species by (i) CA and by (i) SA are not exhaustive hypotheses. But now you are saying that the origin of species by (i) independent mutations (IM) and by (i) non-Independent mutations (NM) are exhaustive hypotheses.
It would help if you would try to provide an explanation for this, for then you would see the fallacy. Instead, as evolutionists always do, you simply make the bare assertion. This is important because this is a key argument, as Sober elucidates, for evolution.
With this bare assertion that you know the two and only two explanations that could possibly describe how the species arose, you can easily make the evidence for evolution to be overwhelming.
However, a Bayesian approach also allows us to compare two non-complementary hypotheses, in which case, as you point out, the probability of both may be small
But NM and IM are not complementary. Are you now warming up to this, or are you merely speaking in generalities here?
but one may be substantially more likely than the other. This is where our priors come in useful, but your posterior probabilities are of course only as good as your priors!
Of course. As I’ve pointed out, because NM and IM are not exhaustive your entire approach goes out the window.
But then your beef should be, not with "contrastive reasoning" but with Bayesian inference and the concept of priors. And you may have a point. Perhaps Bayesian priors are what you call "metaphysical" (though I would strongly argue that they are not).
Sigh, I guess not. Evolutionists cannot see the metaphysical spectacles they are looking through. In this case it actually nicely boils down to some very specific math and logic, and still the evolutionist cannot see it. They mandate that evolution is a fact, or overwhelming, in spite of scientific problems, because the metaphysics demands it. And then they turn around and deny any such metaphysics.
Pedant:
DeleteCH: Your claim is that CA and SA as explanations for the origin of species, or IM and NM, are logically mutually exclusive and exhaustive. That there are no metaphysical assumptions involved. If that were true you could make a logical argument to back it up. But of course that is not true. Unless you know of some logical argument no one else knows of.
Pedant: For the umpteenth time: THE LAW OF NONCONTRADICTION (You can look it up.)
I did look it up. The Law of Non-Contradiction does not say that common ancestry (CA) and separate ancestry (SA), or IM and NM, are mutually exclusive and exhaustive. If you are so certain that it does, then please, by all means, let us see how this works. We won’t hold our breath.
It's a claim about logical alternatives.
Yes, it is a claim about logical alternatives. I’m glad we’ve made it this far. CA and SA (and IM and NM) are very specific models for the origin of species. They provide two alternatives.
But these alternatives are not exhaustive. You used the word “logical.” This is precisely where you are failing. If you were to say CA and NOT CA, then you would have two mutually exclusive and exhaustive alternatives. But of course that is not what you are saying.
CH:Lizzie I can’t respond to a non sequitur. Furthermore, this is ironic because, if you look at what you have been saying, you actually are not following the Law of Non-Contradiction. Earlier you agreed that the origin of species by (i) CA and by (i) SA are not exhaustive hypotheses. But now you are saying that the origin of species by (i) independent mutations (IM) and by (i) non-Independent mutations (NM) are exhaustive hypotheses.
DeleteCA and SA are not exhaustive as posed by Sober, who computes the probability p(data|SA) on the principle that under SA the only way that M could appear in both populations is by Independent Mutations, with probability p, and while under CA, M could appear in both populations not only if mutated spontaneously twice but also if it appeared in a common ancestor.
This excludes several scenarios:
HGT: the sequence was inserted by the same virus
ID: a common designer used the same trick twice.
That is why I redid the hypotheses as follows:
Independent mutations, with probability of happening twice (i.e. in two populations independently) = p^N
Non-independent mutations, with probability of being observed >p^N.
That second hypothesis subsumes all hypotheses that include non-independence, and is the complement to IM. Hence "NM".
p(IM) in other words = 1-p(NM).
This makes a difference, because, as p(data|NM) >p(data|(IM), then the smaller p, the greater the confidence we can have in rejecting IM.
However, our conclusions are very different. We are not rejecting Separate Ancestry, and we are not inferring Common Ancestry; we are simply concluding that the appearance of the same mutation in both populations was probably due to a common factor.
Do you now see what I am saying, and why Dembski's claim that because he has rejected Equiprobable Independent Events he can confidently infer Design?
By both your reasoning and mine, he cannot, and nor can we infer CA from rejecting IM (or SA). All we can claim is support for it. However, by rejecting IM, we can claim support for the case that some common factor caused M in these two populations
Note also, Cornelius, as I have reminded you several times, that you are attempting to generalise from a specific hypothetical to a broad claim.
It is perfectly reasonable to conclude that if you find similar sequences in similar populations that they probably share a common ancestor, just as we conclude that a close match on an DNA test probably implies paternity.
This is quite different from claiming common ancestry of all species on the basis that two species share a rare mutation! Which of course nobody does.
There is no slam-dunk test of common ancestry of all species. We have, I have been assuming (from Sober's text) that we are talking about a hypothetical pair of species in which a low probability M is observed.
Please try to shed this obsession with a few populists who express their confidence (which I share) in the theory of evolution in terms that refer to it as a "fact". It's an imprecise term, and it's not a scientific statement.
Nor does it reflect how science (evolutionary or otherwise is done), which requires the recognition that all conclusions are provisional and that all models can in principle be falsified, not necessarily radically, but almost always in part.
CH:I did look it up. The Law of Non-Contradiction does not say that common ancestry (CA) and separate ancestry (SA), or IM and NM, are mutually exclusive and exhaustive. If you are so certain that it does, then please, by all means, let us see how this works. We won’t hold our breath.
DeleteClearly a population either has a common ancestor or it doesn't. There is no excluded middle.
Where there is an excluded middle is in whether some factor other than CA and IM could account for the appearance of M in two populations.
The answer to this is yes.
That's why I reformulated the hypotheses as IM and NM. Clearly mutations cannot be both independent and non-independent. Again, there is no excluded middle.
CH: We won’t hold our breath.
DeleteEL: Clearly a population either has a common ancestor or it doesn't. There is no excluded middle.
But we trust that you've excluded the muddle.
I do hope so.
DeleteEL:
DeleteThis excludes several scenarios:
HGT: the sequence was inserted by the same virus
ID: a common designer used the same trick twice.
That is why I redid the hypotheses as follows:
Independent mutations, with probability of happening twice (i.e. in two populations independently) = p^N
Non-independent mutations, with probability of being observed >p^N.
That second hypothesis subsumes all hypotheses that include non-independence, and is the complement to IM. Hence "NM".
p(IM) in other words = 1-p(NM).
This makes a difference, because, as p(data|NM) >p(data|(IM), then the smaller p, the greater the confidence we can have in rejecting IM.
OK, we’re making a bit more progress. In attempting to justify your reasoning more of the confusion is revealed. What you are saying here makes no sense from a scientific perspective. First, you are simply defining origins by Non-independent mutations (NM) to be all possibilities other than origins by Independent mutations (IM). This of course makes no sense because the label “Non-independent mutations” does not describe the set of all possibilities other than Independent mutations. So you are greatly confused about The Law of Non-Contradiction. You can’t simply decide that one specific hypothesis constitutes all possible hypotheses other than another specific hypothesis.
Evolutionists are never aware of their nonsensical position.
What you could do is use IM and NOT IM as your two hypotheses. Then you would have your two mutually exclusive and exhaustive alternatives. But of course, as I have tried to point out to you, you then would not have a model for “NOT IM.” You have tried to save your math by simply defining NM to be all possibilities other than IM. What you don’t realize is that in that case you lose your math because your nice neat model for NM no longer applies. On the other hand, you also save your math by defining NM as your nice neat model. That allows you to do your math. But in that case, you lose your math because IM and NM are no longer mutually exclusive and exhaustive.
The bottom line is, none of this is science. Your claims are strictly non scientific, and its important to be clear about that.
Please try to shed this obsession with a few populists who express their confidence (which I share) in the theory of evolution in terms that refer to it as a "fact". It's an imprecise term, and it's not a scientific statement.
No, that is the consensus position. I asked you to provide a reference for an influential evolutionist who believes evolution is not a fact, as much as gravity for example. As we saw in the Emory University case, once again the biology professors made the claim that evolution is as sure as gravity. And you said the evidence is “overwhelming” and that “last Thursday-ism” anti-realism would be the only other option. So again you are greatly confused about this. You’re not even reckoning with your own view, much less that of the consensus.
CH:OK, we’re making a bit more progress. In attempting to justify your reasoning more of the confusion is revealed. What you are saying here makes no sense from a scientific perspective.
Deleteum, yes, it does.
First, you are simply defining origins by Non-independent mutations (NM) to be all possibilities other than origins by Independent mutations (IM).
No. I. Am. Not.
Please read my posts, and please do not skim for button-pushing words and ignore the ones in betwee.
No I am not "simply defining origins by Non-independent mutations (NM) to be all possibilities other than origins by Independent mutations (IM)."
I am simply defining two hypotheses to account for the appearance of a rare mutation in two separate populations.
One hypothesis says: They appeared independently in each population with probability p (i.e. p^2 for it happening in both).
The other says: they didn't.
There is no excluded middle because one is simply the complement (not-A) of the other (A).
I said nothing about defining "origins". I didn't mention "origins". I simply provided two complementary hypotheses to account for the same rare mutation in two populations.
This of course makes no sense because the label “Non-independent mutations” does not describe the set of all possibilities other than Independent mutations.
Yes, it does.
Oh - hang on - is it "mutation" you are bothered about? Call it "sequence" then.
So you are greatly confused about The Law of Non-Contradiction. You can’t simply decide that one specific hypothesis constitutes all possible hypotheses other than another specific hypothesis.
Yes I can. I can say: this was caused by A, or it was not caused by A. And I just did.
CH:Evolutionists are never aware of their nonsensical position.
Possibly because the nonsensicality is completely illusory, arising for the fact that the anti-evolutionist hasn't actually understood the position. Or even read it properly.
CH:What you could do is use IM and NOT IM as your two hypotheses. Then you would have your two mutually exclusive and exhaustive alternatives.
YESSSSSS!!!!!
Which is why I called them "Independent Mutations and NON-independent Mutations"!!!!!!
("IM" and "NM" for short.)
Blimey.
CH:But of course, as I have tried to point out to you, you then would not have a model for “NOT IM.” You have tried to save your math by simply defining NM to be all possibilities other than IM. What you don’t realize is that in that case you lose your math because your nice neat model for NM no longer applies. On the other hand, you also save your math by defining NM as your nice neat model. That allows you to do your math. But in that case, you lose your math because IM and NM are no longer mutually exclusive and exhaustive.
Yes, as I said, first several posts ago, and most recently in my last post, the math is not so neat. However, you still have the prediction that (data|NM)>p(data|IM). So you still have support for NM (or against IM) if you see the same mutation/sequence in two populations and p is low.
continued...
DeleteThis isn't rocket science Cornelius. All this fancy math is saying is that if you see the same sequence in two populations, and you know from other data that that sequence has very low probability of occurring spontaneously, that the cause of that sequence in the two populations probably wasn't independent mutations.
We do this all the time. It's at the core of Dembski's entire thesis - if you see a pattern that is unlikely to have arisen from some kind of independent random draw, some kind of non-independent process was probably at work.
Except that we don't normally make Dembski's mistake (the one you are accusing me of, but should be accusing Dembski of) that the only non-independent process has to be intelligence. He excluded a middle. I'm not.
The bottom line is, none of this is science. Your claims are strictly non scientific, and its important to be clear about that.
I would agree if it were what I was doing, but it isn't. It is what Dembski is doing, however, wouldn't you agree?
CH:No, that is the consensus position. I asked you to provide a reference for an influential evolutionist who believes evolution is not a fact, as much as gravity for example. As we saw in the Emory University case, once again the biology professors made the claim that evolution is as sure as gravity.
Right, and I agree with those professors. I just think that "fact" is the wrong word to describe an established theory, so I don't use it. Some do. I don't consider the theory of gravity "a fact" either. We actually don't even have an explanatory theory for gravity, just a set of laws. Evolutionary theory is ahead of gravity on that score.
And you said the evidence is “overwhelming” and that “last Thursday-ism” anti-realism would be the only other option.
You totally missed my point, Cornelius, as, unfortunately, you have done repeatedly.
Unless a theory makes predictions we cannot compare it with theories that do. We can only add, as a caveat: "of course something quite different that we haven't thought of and haven't modeled might be responsible".
ID is an appeal to miraculous intervention, and thus it makes no predictions. We can add the caveat that it might be true, but we cannot possible predict what data would look like under ID, nor under Omphalism.
Science proceeds by falsification only in the sense that we keep ruling stuff out. We can't ever rule stuff in. But if what we are left with is Evolution OR ID OR something else we haven't thought of, and there is no way of putting Evolution head to head with ID, then the scientific approach is to stick provisionally with Evolution until someone comes up with an actual predictive theory that can be compared with it.
Frontloading is a possibility, but I still haven't seen a testable version.
CH:So again you are greatly confused about this. You’re not even reckoning with your own view, much less that of the consensus.
I profoundly disagree.
Continued …
DeleteCH: What you could do is use IM and NOT IM as your two hypotheses. Then you would have your two mutually exclusive and exhaustive alternatives.
EL: YESSSSSS!!!!! Which is why I called them "Independent Mutations and NON-independent Mutations"!!!!!! ("IM" and "NM" for short.)
No, you are still confused about this. Independent Mutations and NON-independent Mutations are not mutually exclusive and exhaustive. NOT IM is not the same as NON-independent Mutations. NOT IM is a huge set of explanations. Your NON-independent Mutations is a very specific model (in your own math!), reflecting a very narrow explanation.
CH: No, you are still confused about this. Independent Mutations and NON-independent Mutations are not mutually exclusive and exhaustive. NOT IM is not the same as NON-independent Mutations. NOT IM is a huge set of explanations. Your NON-independent Mutations is a very specific model (in your own math!), reflecting a very narrow explanation.
DeleteThen it should be easy for you to list a couple of the explanations in that huge set.
No, you are still confused about this. Independent Mutations and NON-independent Mutations are not mutually exclusive and exhaustive. NOT IM is not the same as NON-independent Mutations. NOT IM is a huge set of explanations. Your NON-independent Mutations is a very specific model (in your own math!), reflecting a very narrow explanation.
DeleteExplain, please. I meant mutations aka sequences (in case you were hung up on the word "Mutations" as implying some specific process) that emerged from processes that were NOT indepedent of each other. That's why I called them "NON-independent Mutations". Meaning they did NOT arise from independent processes.
If you are happier with "Not Independently-Generated Sequences" I am happy to go with NIGS. Same math applies.
If not, please show why not.
EL:
Delete(edited for clarity)
CH: First, you are simply defining origins by Non-independent mutations (NM) to be all possibilities other than origins by Independent mutations (IM).
EL: No. I. Am. Not.
Yes you are. When you say p(NM) + p(IM) = 1, that is claiming that your two explanations for the origins of species are exhaustive. It’s in the math. You have no justification for that, it is a bare assertion.
I am simply defining two hypotheses to account for the appearance of a rare mutation in two separate populations. One hypothesis says: They appeared independently in each population with probability p (i.e. p^2 for it happening in both).
Well to be clear, mutations are causes. We’re talking about observations. Sober uses example such as tail bones and gill slits. You’re confusing cause and effect. If you presuppose mutations as your cause, then it is obvious that your explanations are not exhaustive. This is just silly.
Oh - hang on - is it "mutation" you are bothered about? Call it "sequence" then.
At least this is a move in the right direction. A sequence at least is not a cause. It is an observation. So that’s an improvement. A better word would be “structure” since we’re talking about various biological structures in different species. But now you have no scientific justification for your claim that your two models are exhaustive. Elsewhere you even admitted that other explanations exist, so now you’re caught even in your own words.
You’re just jumping back and forth from frying pan to fire to try to justify your metaphysics. Your NM and IM alternatives are not exhaustive alternatives for the observed structures, sequences, or whatever you want to call what we see in different species. You insist that they are and therefore that NM easily beats IM. None of this is science.
Continued …
Sorry looked at your second post, first.
DeleteYes, it seems as though the word "mutations" has been a problem. I simply meant "sequence" .
I hope it is now clear to you that my hypotheses are in fact mutually exclusive. To make things (I hope) crystal clear:
Data: same sequence in two different populations.
Hypothesis 1: Sequence arose independently in the two populations .
Hypothesis 2: Sequence did not arise independently in the two populations.
p(data|Hypothesis 1)=p^2
p(data|Hypothesis 2)>p^2.
Ergo, the smaller p, the greater the confidence with which we can reject Hypothesis 1, and seek a common factor that led to the observed sequence in both populations.
Now, will you please tackle Dembski's problem, which is that he claims that if we reject the hypothesis that pattern A was generated by a sequence of independent events we must infer design.
While my claim is that I have NOT contravened your valid objection to conclusions based on non-excluded middles, Dembski clearly has.
And it is fundamental to his entire ID project.
If not, why not?
Also, Cornelius, I'm getting a bit fed up with this "your metaphysics" and "none of this is science", mantra.
DeleteI get it that this is what you think. I disagree.
Let's concentrate on why we disagree.
That's odd- your "continued" post is above your other one!
DeleteOh well, software is weird.
CHYour two hypotheses now deal with how sequences arose, so that’s good.
DeleteThey always did, Cornelius. I've re-read my posts and I cannot understand why you did not get that. However, now you do.
Good.
Earlier I suggested the word “structure” and you should take me up on it, because that, more generally, is what we’re dealing with. Different structures may or may not be generated by different sequences. But in any case, switching to “structure” would not change your argument. And importantly, it would still be problematic.
Yes, I know that different structures may or may not be generated by different sequences.
We are talking about some hypothetical sequence M, which may or may not be a coding gene, and which appears in two different populations, are we not?
And, we are assuming that from other data we know the probability, "p", of that sequence arising by chance.
It's a toy example, because of course we do not ever know "p", and, of course, "p" is itself dependent on what preceded M (assuming something did).
So if you want to go into details, let's take a real-life example - your bat and dolphin sonar, for intances.
But not here. Right now, let's stick to your original point, from Sober, that the smaller "p" the better the evidence for any hypothesis that includes non-independent causes of M (for Sober, Common Ancestry for the two populations; I am widening this to include any non-independent cause).
Continued …
DeleteFor your two alternatives A and B, evolutionists have two mathematical models for the probability of a homologous structure appearing in two species. For A, the structure spontaneously arises once, and for B it arises twice. Sober includes terms to account for the probability of the structure arising, and being passed on. That’s fine as far as it goes, but such mathematical models do not account for all possibilities. They are modeling a very narrow, particular version of “Not Independently-Generated Structures” and of “Independently-Generated Structures.”
What is important here is not whether the labels “Not Independently-Generated Structures” and “Independently-Generated Structures” are exhaustive. They certainly appear to be exhaustive. What is important is whether or not the underlying math models, which hold the real meaning being poured into each label, is exhaustive. We can play semantic games and fool ourselves that we have p(A) + p(B) = 1. But it’s just a big equivocation if the two math models don’t have this property.
In other words, the math models reveal that these two choices, A and B, do not provide completeness. You do not have p(A) + p(B) = 1. And if you make the bare assertion that they are exhaustive, then that invalidates the math models.
Look at my response to Pedant to see that there are all kinds of theoretical variations which evolutions can, and have, come up with, which invalidate those two narrow models.
And this is simply the tip of the iceberg. You can imagine a conversation where I would erroneously be agreeing with you that those two narrow math models are indeed exhaustive. I would agree because I hadn’t realized there were more possible alternatives than the two math models. We would then agree and be done with it. The problem is we would have been unaware that there were unconceived alternatives. In fact, evolution is a great example of a broad, malleable theory where there is a wide range of explanations. There are always more alternatives to handle the next contradictory finding.
So it would be very difficult to define exhaustive alternatives.
Continued …
CH:Look at my response to Pedant to see why that is a problem If they cover all possibilities then you cannot model them.
DeleteLook at my response to you to see why you can.
And, Cornelius, can you please explain why you don't think this is a problem for Dembski?
Unless you do?
If so, will you say so?
Dambski: "In eliminating chance and inferring design, specified complexity is not party to an argument from ignorance. Rather, it is underwriting an eliminative induction. Eliminative inductions argue for the truth of a proposition by actively refuting its competitors (and not, as in arguments from ignorance, by noting that the proposition has yet to be refuted). Provided that the proposition along with its competitors form a mutually exclusive and exhaustive class, eliminating all the competitors entails that the proposition is true. (Recall Sherlock Holmes’s famous dictum:“When you have eliminated the impossible, whatever remains, however improbable, must be the truth.”) This is the ideal case, in which eliminative inductions in fact become deductions. But eliminative inductions can be convincing without knocking down every conceivable alternative, a point John Earman has argued effectively. Earman has shown that eliminative inductions are not just widely employed in the sciences but also indispensable to science. Suffice it to say, by refusing the eliminative inductions by which specified complexity eliminates chance, one artificially props up chance explanations and (let the irony not be missed) eliminates design explanations whose designing intelligences don’t match up conveniently with a materialistic worldview."
So Dembski is happy, to "eliminate chance explanations", as I did when I observed that if p is small, the chance explanation that gives a probability of the data of p^2 is extremely unlikely.
However, unlike Dembski, I am prepared merely to infer that some non-independent process was involved. Dembski insists that we can infer design.
He assumes that "independent chance events" and "design" are mutually exclusive.
He is clearly wrong, as you must surely agree.
EL:
DeleteWe are talking about some hypothetical sequence M, which may or may not be a coding gene, and which appears in two different populations, are we not?
Sober mentions tail bone and gill slits. Could also be a gene. They’re all structures.
CH:Look at my response to Pedant to see why that is a problem If they cover all possibilities then you cannot model them.
EL: Look at my response to you to see why you can.
No, your response just continues to make the same mistake, such as:
And, we are assuming that from other data we know the probability, "p", of that sequence arising by chance.
It's a toy example, because of course we do not ever know "p", and, of course, "p" is itself dependent on what preceded M (assuming something did).
So if you want to go into details, let's take a real-life example - your bat and dolphin sonar, for intances. But not here.
On the one hand you claim p(A) + p(B) = 1. On the other hand, you restrict the problem to two very specific, narrow models. Once you make this restriction you automatically lose p(A) + p(B) = 1. If you’re going to define the problem as only allowing the two narrow, specific explanations, then they are not exhaustive. This really is silly.
Think about what you are doing. You’re saying, let’s pick out two toy explanations, show that one is better than the other if we smuggle in the non scientific premise that they’re exhaustive so we can conclude that the one wins. This is incredible.
CH:For your two alternatives A and B, evolutionists have two mathematical models for the probability of a homologous structure appearing in two species. For A, the structure spontaneously arises once, and for B it arises twice. Sober includes terms to account for the probability of the structure arising, and being passed on. That’s fine as far as it goes, but such mathematical models do not account for all possibilities. They are modeling a very narrow, particular version of “Not Independently-Generated Structures” and of “Independently-Generated Structures.”
DeleteYes, I agree. Which is why I changed the model merely to p(data|A)>p^2, in order to account for all non-independent explanations. And as long as that inequality holds, the smaller p is, the stronger evidence is against IM.
CH:What is important here is not whether the labels “Not Independently-Generated Structures” and “Independently-Generated Structures” are exhaustive. They certainly appear to be exhaustive.
Yes, they do, rather :)
What is important is whether or not the underlying math models, which hold the real meaning being poured into each label, is exhaustive.
Right. Which is why I have simply expressed it as an inequality.
We can play semantic games and fool ourselves that we have p(A) + p(B) = 1. But it’s just a big equivocation if the two math models don’t have this property.
I have modeled:
p(data|A) as >p^2
and
p(data|B) as p^2
These predictions are different. There is, I guess technically, an excluded middle, in that there may be a hypothesis that predicts that
p(data|C)=< p^2.
I'm trying to think of a set of hypotheses that would have this prediction and yet be compatible with a knowable p. OK, let's say that C is the hypothesis that some process or processes unknown make M less likely if it has already occurred once - that Ms are drawn from a distribution without replacement.
We can still set up an exhaustive pair of hypotheses:
A: M, in these two populations is caused by a non-independent process, that make M more likely to recur, if it has happened once.
B: M, in these two populations is caused by independent processes with probability p.
C: M, in these two populutions, is caused by non-independent processes that make M less likely to occur, if it has occurred once.
So we have:
p(data|A)>p^2
p(data|B)=p^2
p(data|C)<p^2
But we have the data (M has been observed in two separate populations).
The smaller p, the stronger the case for A becomes over B, and for B over C.
Therefore, the smaller p, the stronger the case for A.
In other words, the math models reveal that these two choices, A and B, do not provide completeness. You do not have p(A) + p(B) = 1. And if you make the bare assertion that they are exhaustive, then that invalidates the math models.
No, see above. The math models are now exhaustive. But I concede I had not allowed for a hypothesis that involved some kind of process that makes M less likely if it has previously occurred.
Look at my response to Pedant to see that there are all kinds of theoretical variations which evolutions can, and have, come up with, which invalidate those two narrow models.
I think I have eliminated the last lurking exclusion. I submit that both my sets of hypotheses and my math models are now exhaustive.
And this is simply the tip of the iceberg. You can imagine a conversation where I would erroneously be agreeing with you that those two narrow math models are indeed exhaustive. I would agree because I hadn’t realized there were more possible alternatives than the two math models. We would then agree and be done with it. The problem is we would have been unaware that there were unconceived alternatives.
DeleteWell we still have unconceived alternatives, but we have narrowed them down, by eliminating a swathe of them, probabilistically at any rate.
In fact, evolution is a great example of a broad, malleable theory where there is a wide range of explanations. There are always more alternatives to handle the next contradictory finding.
Yes, indeed. That is evidence of its robustness as a theoretical framework. But I think there is an important misundertanding here. It's not that new data results in some "excuse" being made for the original theory. The original theory is constantly revised in the light of the new data.
That's not a weakness in the original theory - it's a strength. We have a broad framework that can accommodate a large number of new mechanisms within it. That framework is longitudinal common ancestry with various horizontal genetic transfer mechanisms, various mechanisms of variance generation, and natural selection at both within- and between-population level, as well a neutral drift.
These are elaborations of the theory, not holes in it. And each of these elaborations make new testable predictions.
CHOn the one hand you claim p(A) + p(B) = 1. On the other hand, you restrict the problem to two very specific, narrow models.
DeleteI'm not restricting "the problem" to "models". I'm simply restricting the problem to that of determining which hypothesis best explains a rare sequence, M, in two populations. The models, for that problem, are exhaustive.
Once you make this restriction you automatically lose p(A) + p(B) = 1.
No, I don't. Please read my posts, Cornelius, you are clearly skimming, because you miss really explicitly made points.
If you’re going to define the problem as only allowing the two narrow, specific explanations, then they are not exhaustive. This really is silly.
But my explanations aren't "narrow". One is A; the other is not-A.
They are therefore exhaustive. And the predictions of A and not-A are also exhaustive (now that I have allowed for an additional possibility, namely that p(data|C)< p^2).
Think about what you are doing. You’re saying, let’s pick out two toy explanations,
They are not "two toy explanations". They are now three (I do concede I had ommitted one) mutually exclusive sets of explanations.
show that one is better than the other if we smuggle in the non scientific premise that they’re exhaustive so we can conclude that the one wins. This is incredible.
It's not non-scientific premise. It's intrinsic to the way I have defined the hypotheses.
And that is precisely why, when p is low, I can conclude only that "Some non-independent process was probably responsible for the appearance of this rare sequence M in two different populations".
I am NOT concluding "Common ancestry was responsible yadda yadda".
Now, we can also consider our priors, and my priors for CA are probably going to be high, depending on other data. And maybe priors are what you call "metaphysics". But we are not loading our priors here. We are simply looking at the likelihood ratio, and, as p gets smaller, the likehoood ratio with A on the numerator increases.
And what about Dembski?
And what about the question I posed above?
DeleteCH: No, you are still confused about this. Independent Mutations and NON-independent Mutations are not mutually exclusive and exhaustive. NOT IM is not the same as NON-independent Mutations. NOT IM is a huge set of explanations. Your NON-independent Mutations is a very specific model (in your own math!), reflecting a very narrow explanation.
Me to CH: Then it should be easy for you to list a couple of the explanations in that huge set.(of NON-independent Mutations)
I'd really like to know what I'm missing.
Thanks.
Pedant:
DeleteAnd what about the question I posed above?
CH: No, you are still confused about this. Independent Mutations and NON-independent Mutations are not mutually exclusive and exhaustive. NOT IM is not the same as NON-independent Mutations. NOT IM is a huge set of explanations. Your NON-independent Mutations is a very specific model (in your own math!), reflecting a very narrow explanation.
Me to CH: Then it should be easy for you to list a couple of the explanations in that huge set.(of NON-independent Mutations)
I'd really like to know what I'm missing.
Thanks.
Sorry about that. I replied but it’s gone. Here’s my point. As more and more biology is understood, and more and more contradictory data are gathered, evolution becomes more and more complex. Basically, you can do one of two things with contradictory data: drop the theory or modify the theory to account for the new data. Of course evolutionists will never drop their theory. They modify the theory with whatever add-on is required, and with each new add-on the theory becomes more complex. At this point evolutionists have posited a great variety of explanations. Needless to say, they do not neatly fall into the “independent origin” and “non independent origin” categories.
And yet, they turn around and “prove” common ancestry by assuming these narrow, specific categories of IM and NM. So to make their truth claims they specify the options as IM and NM, and claim these are the only choices. But to account for the actual data they invoke myriad explanations. It’s just silly.
EL:
DeleteI have modeled:
p(data|A) as >p^2
and
p(data|B) as p^2
These predictions are different. There is, I guess technically, an excluded middle, in that there may be a hypothesis that predicts that
p(data|C)=< p^2.
First you were certain that p(A) + p(B) = 1. Then you said p(data|A) > p^2. That’s wrong, as p(data|A) = p. If you want to make up a new hypothesis, call it A’, for which p(data|A’) > p^2, then fine. But p(data|A) = p. You were certain that p(A’) + p(B) = 1. As before, this is a non scientific claim, but even given such a claim, you now have lost the compelling argument for A’, since the data are only epsilon more probable than on B.
Next you realized A’ and B are not exhaustive, and p(A’) + p(B) do not sum to unity. So you added C for which, conveniently, the data are even less probable: p(data|C)=< p^2. And now you’re sure that p(A’) + p(B) + p(C) = 1. Again, this not only is not scientific, but you still have lost the compelling argument for A’.
CH: In other words, the math models reveal that these two choices, A and B, do not provide completeness. You do not have p(A) + p(B) = 1. And if you make the bare assertion that they are exhaustive, then that invalidates the math models.
EL: No, see above. The math models are now exhaustive. But I concede I had not allowed for a hypothesis that involved some kind of process that makes M less likely if it has previously occurred.
CH: Look at my response to Pedant to see that there are all kinds of theoretical variations which evolutions can, and have, come up with, which invalidate those two narrow models.
EL: I think I have eliminated the last lurking exclusion. I submit that both my sets of hypotheses and my math models are now exhaustive.
No, A’, B and C are not exhaustive. p(A’) + p(B) + p(C) is not = 1. You made up an A’ for which p(data|A’) > p^2. You could have just as easily made up a B’ for which p(data|B’) > p^2, or p(data|B’) = p, or whatever. Evolutionists make up all kinds of explanations for contradictory convergences in biology. Or you could have made up a D, and an E, etc.
If you say, no, I’m *defining* A’, B and C to be exhaustive, then these become broad categories rather than specific explanations, and you lose your p(data|X) claims. They could be anything because your hypotheses are so broad.
Continued …
Continued …
DeleteWell we still have unconceived alternatives, but we have narrowed them down, by eliminating a swathe of them, probabilistically at any rate.
That’s meaningless. Narrowing a large number to a lesser number doesn’t justify the exhaustive claim.
CH: In fact, evolution is a great example of a broad, malleable theory where there is a wide range of explanations. There are always more alternatives to handle the next contradictory finding.
EL: Yes, indeed. That is evidence of its robustness as a theoretical framework. But I think there is an important misundertanding here. It's not that new data results in some "excuse" being made for the original theory. The original theory is constantly revised in the light of the new data. That's not a weakness in the original theory - it's a strength.
That speaks volumes. Contradictory data reveal how great the theory is.
But my explanations aren't "narrow". One is A; the other is not-A. They are therefore exhaustive. And the predictions of A and not-A are also exhaustive (now that I have allowed for an additional possibility, namely that p(data|C)< p^2).
There you go again. One minute you submit that your hypotheses are exhaustive, the next minute you agree you still have unconceived alternatives, and then the next you simply *define* your explanations to be exhaustive. You’re all over the map, jumping from the frying pan to the fire, and back.
As I’ve explained over and over, either way you have problems. Either you have exhaustive hypotheses, but in that case you forfeit specific math models (unless you use your metaphysics). Or you keep your math models, but you forfeit completeness (unless you use your metaphysics).
CH:First you were certain that p(A) + p(B) = 1.
DeleteThat would depend what A and B were, obviously. But if you define B as not-A, then, clearly p(A)+p(B)=1.
CH:Then you said p(data|A) > p^2. That’s wrong, as p(data|A) = p.
Well, depends what A stands for (and p). But if p (let's denote it P for clarity) is the probability of the sequence arising spontaneously, then P^2 is the probability of it arising spontaneously twice. And so, if A stands for the hypothesis that sequence M, observed in two separate populations, is the result of independent spontaneous equations, then p(data|A)=P^2.
CH:If you want to make up a new hypothesis, call it A’, for which p(data|A’) > p^2, then fine. But p(data|A) = p. You were certain that p(A’) + p(B) = 1. As before, this is a non scientific claim, but even given such a claim, you now have lost the compelling argument for A’, since the data are only epsilon more probable than on B.
This is very difficult to parse as I don't know what hypotheses of mine you are referring to as "A" and "B". But if "A" is "independent spontaneous mutations and "B" is "non-independent spontaneous mutations" (which I denoted, earlier, as IM and NM respectively) then indeed p(A)+p(B)= 1.
CH:Next you realized A’ and B are not exhaustive, and p(A’) + p(B) do not sum to unity. So you added C for which, conveniently, the data are even less probable: p(data|C)=< p^2.
No, p(IM) and p(NM) are exhaustive. What I had failed to consider was that p(NM) could include hypotheses in which the sequences had non-independent origins that made M less likely to occur spontaneously a second time, if had already occurred once. So while p(IM)+p(NM) still =1, we have to split NM into two subsets, one of which predicts that M will be more likely if it has occurred once, and one in which M is less likely if it has occurred once. I'll call them NM+ and NM-. Because real numbers can only be <0, =0 and >0, then our bottom line is still exhaustive:
p(data|NM+)+p(data|NM-)+p(data|IM)=1
Also:
p(NM+)+p(NM-)+p(IM)=1
CH:
DeleteAnd now you’re sure that p(A’) + p(B) + p(C) = 1. Again, this not only is not scientific, but you still have lost the compelling argument for A’.
See above.
CH:If you say, no, I’m *defining* A’, B and C to be exhaustive, then these become broad categories rather than specific explanations
Exactly. And this is precisely the point I have been making for several posts. I
CH:, and you lose your p(data|X) claims. They could be anything because your hypotheses are so broad.
No, they cannot "be anything". If the two occurrences of M are non-independent, p(data|NM) =/= P^2.
Moreover, p(data|NM+)>P^2, and p(data|NM-)<P^2.
And so, the smaller p is, the greater confidence we can have in rejecting both IM and NM-.
CH:That’s meaningless. Narrowing a large number to a lesser number doesn’t justify the exhaustive claim.
Of course it does. We can rule stuff out (probabilistically), leaving all the remaining stuff in. That's all I've done. In this hypothetical case, if p is low, we have support for NM+ hypotheses.
CH:
That speaks volumes. Contradictory data reveal how great the theory is.
Really Cornelius, would you please read what I have written! Contradictory data don't "reveal how great the theory is". What reveals "how great the theory is" is its capacity to be be modified to accommodate contradictory data, while remaining a coherent theoretical framework. Surely you see the difference between these two claims? If not, then I guess that "speaks volumes" ;)
CH:There you go again. One minute you submit that your hypotheses are exhaustive,
IM and NM were, and are, exhaustive hypotheses.
the next minute you agree you still have unconceived alternatives,
No, what I agreed is that NM can be two exhaustive subsets, one of which include hypotheses that predict that M will be more likely to occur if it has occurred once, and one of which include hypotheses that predict that M will be less likely to occur if it has occurred once.
and then the next you simply *define* your explanations to be exhaustive.
Yes, of course. That's what Dembski fails to do. You have to define your null as the complement of your study hypothesis. If you are going to test your hypothesis against a null, anyway.
You’re all over the map, jumping from the frying pan to the fire, and back.
Not at all :)
As I’ve explained over and over, either way you have problems. Either you have exhaustive hypotheses, but in that case you forfeit specific math models (unless you use your metaphysics). Or you keep your math models, but you forfeit completeness (unless you use your metaphysics).
No. As I have shown, it is often possible (as it is in this case) to make a clear directional prediction for one set of hypotheses relative to another, even when the sets are exhaustive. In this case, the hypotheses of independence and negative dependence predict that the frequency of M occurring in two populations will will be =<:P^2, while for the hypotheses of positive dependence, the predicted frequency will be >: P^2.
I think your data analytic skills must be rusty Cornelius :) This is really quite straightforward.
Now, are you going to tackle the Dembski issue? Or will you continue to ignore the elephant in the room, just as you ignored the Sewell elephant?
Oops, messed up the colons:
Delete"In this case, the hypotheses of independence and negative dependence predict that the frequency of M occurring in two populations will will be =<P^2, while for the hypotheses of positive dependence, the predicted frequency will be > P^2.
And I have posted a blog piece here.
ReplyDeleteI would be delighted if anyone here including Cornelius would like to drop by. All first posts are automatically held in the moderation queue to prevent spam, but I will make sure I check the queue regularly.
EL:
Delete"And I have posted a blog piece here."
Well your post is based on an abbreviation of mine, in a comment, that is in a conversation we're currently having. I explained the abbreviation, in another comment, and have posted my next comment in response to your last one in a bit. All that is to say that I don't hold the view that you represent me as holding in your post.
OK, and I am more than willing to correct what I wrote.
DeleteOr perhaps it would be even better if you commented yourself!
I am honestly trying to understand your position from what you have written.
I have a general rule,whenever I read" mound of skulls at the bottom of pyramid $chemes of Darwinian sort with their ivory tower of babble" ,I assume no factual claims are present.
ReplyDeleteI know you don't understand the use of symbolism and it seems like babble to you.
Shrug, it is what it is.