8/31/2007

VOX DEI-BATE #10: NO 'BACK-TESTING' HERE


Vox’s characteristic skepticism regarding TENS’s ability to lead to testable predictions has been well-noted, but in this post he actually sticks his neck out with respect to the Miller presentation that I posted about earlier here. As you may recall, Miller presented the evidence that human chromosome #2 was produced by the fusion of two separate chromosomes, and argued that the existence of this telomeric event was a testable prediction. Vox, who wants to characterize such claims as ad hoc (or, as he puts it, ‘backtesting’) is not merely skeptical, but incredulous:

"Scott's example is the work done by Dr. Ken Miller, who posits the rather bizarre notion that evolution is not only busy creating new species by natural selection, but also making testable predictions. This prediction - for which he shows no evidence of having been made prior to the test - is that "the common ancestor had 48 chromosomes (24 pairs) and humans carry a fused chromosome; or ancestor had 23 pairs and apes carry a split chromosome."

Well, Vox, ask and ye shall receive. It turns out that the original prediction of interest was not based on genomic data (which did not exist back in the 1970's) but on perceived similarities in banding patterns between primate chromosomes. It was easy to determine significant homology in many of the chromosomes compared on the basis of those banding patterns alone. However, this alleged similarity did not address the fact that humans and the other great apes differed in the number of chromosomes.

(Hmmmm---what scenario could explain the difference in chromosome number, and yet be consistent with that working hypothesis of common descent?)

To the best of my knowledge, the first suggestion that human chromosome #2 might’ve been produced by the fusion of two chromosomes which exist separately in chimps was made in a paper "Comparaison de la structure fine des chromatides d'Homo sapiens et de Pan troglodytes" by LeJeune, Dutrrilaux, Rethore and Prieur which appeared in 1973. I don’t have the original, but the abstract for that paper can be seen here.*

Now, what should be mentioned to the non-specialist is that this was a fairly daring proposal back in 1973. Relatively few examples of what were called Robertsonian translocations were known at the time, and virtually none involving telomeres----whose well-known role as a buffer against copying mistakes during cell division had led to their being regarded as ultra-stable structures, the sort of chromosomal chunk least likely to recombine or fuse with another. As Holmquist and Dansis (1979) remarked, events like telomeric fusion were considered "discordant with classical cytogenetic theories, which assume all chromosome rearrangements to require at least two breaks, and consider centromeres and telomeres as immutable structures rather than structures determined by mutable DNA sequences."

In that context, we can appreciate the boldness of the 1973 paper. The scientists who made it were specialists in the effects of such events on human chromosome, and they were justifiably confident in their ability to interpret banding patterns: the lead scientist on that paper, Jerome LeJeune, was the discoverer of trisomy-21 (Down’s Syndrome). Nevertheless, as this pointed critique of an ID enthusiast notes, the inference of an event of telomeric fusion was not in itself evidence for common descent: "It’s the sequence, not the fusion, that tells us of our relatedness."

Quite so—and it was ultimately the completion of that sequence in the Chimpanzee Genome Project that provided us with the high level of detail needed to either falsify or confirm the original speculative hypothesis, which (if true) predicted that telomeric signatures would be identified. The identification of the telomeres within human chromosome #2 (the fulfillment of said prediction) was made in 1991 and summarized here.

Note the beautiful convergence over time of completely separate research programs!Cytogeneticists, prompted by evolutionary theory, provided the scientific community with a testable hypothesis. Molecular biologists put that hypothesis to test with genomic data, which not only confirmed the hypothesis (the fusion event) but provided powerful evidence for a recent common ancestor for Homo and Pan.

* Yunis and Prakash’s 1982 article in Science is probably a more accessible discussion of the research program at that time, as well as the logic behind the hypothesis of telomeric fusion.

16 comments:

R. Moore said...

Scott --
I have become completely lost as to where the Vox debate is going. First there is the confusion between a model (eg macroeconomics) and an algorithm (TENS). Strictly speaking, only models can expect to predict, as they incorporate both data and algorithms.

Then the back-testing complaint -- am I to object to Newton's theory of gravity because it explained all past behavior but made no future predictions until Einstein?

Now we are expected to believed that there is little predictive power in TENS. Ok, here goes:

I predict no new species will ever be naturally created, except from an earlier species. This claim cannot be back-tested, since the fossil record is not complete enough. If some disagrees with this, I would love to see their reasoning.

I predict every future issue of Science magazine will continue to be full of articles on evolution, confirming predictions made in previous articles. (It is hoped further work in this area will show...)

I am purposefully generalizing, because I am afraid Vox's next argument will be "Ok, just once something happened that looks like an evolutionary prediction, but this could just be a random occurrence).

Anonymous said...

Scott Hatfield:
It turns out that the original prediction of interest was not based on genomic data (which did not exist back in the 1970's) but on perceived similarities in banding patterns between primate chromosomes.

Sincerely, Scott, where is the TENS prediction?

You've shown the description of the phenomena (HSA2 appearing as a telomeric fusion of chimp chromosomes 12 & 13) originally observed in comparative banding studies in 1973, and argued this was the result of evolution from chimp to human, and with successively more detailed genomic sequencing analyses in 1982 and again in 1991 and again in 2005 confirming that HSA2 is indeed a telomeric fusion of two chromosomes also found in chimps.

But where is the TENS "prediction" (a specific statement made in advance of the observation or experiment) before 1973 that some human chromosomes would be fusions of chimpanzee chromosomes?

Indeed, the genomic sequencing studies confirmed the banding study observations, but the banding study did not predict the HSA2 fusion from chimp chromosomes.

As recently noted here: As an example of the differences between backtesting or curvefitting and prediction, consider that many decades prior to Penzias and Wilson the theoretical temperature of space was computed and predicted variously by Guillaume, Eddington, Regener and Nernst, McKellar and Herzberg, Finlay-Freundlich and Max Born as 1.9 - 6.0 deg K, which prediction was finally observed (accidently, unexpectedly) by Penzias and Wilson and measured as 2.7 deg K.

Arguing that a telomeric fusion of HSA2 "is explained by TENS" does not equate to saying it "was predicted by TENS"

A prediction is a theoretical expectation in advance of observation or experimental outcome. The data observed may be old or new, but the findings expected are described by the theory in advance of and subsequently match the observations.

Further, saying that TENS or common descent "predicts" shared ancestry and that this prediction is confirmed by the HSA2 fusion has two problems:
1) that humans vertically inherited HSA2 sequences from chimpanzees or a common ancestor is not proven. Other mechanisms such as endosymbiotic gene transfer, lateral gene transfer and non-random recombination need to be conclusively ruled out. And while genome diagnostic techniques are quite well understood, the mechanisms for mutation, expression, regulation and conservation etc., even what a gene is vs a psuedo-gene, are not. Nor are the databases of ancient genomes sufficient for conclusive comparisons.
2) the prediction itself needs specificity to be credible. While common descent and shared ancestry are arguably "predictions" made long ago by Darwin they are essentially tautologies: that being, common descent is proven by descendants having common genes. That prediction is always true because its premise of vertical inheritence is tacitly presumed true. Lacking in that prediction is any specificity as to mechanism of gene acquisition and gene expression in phenotypic differences.

Richard above draws a comparison to Newton's theory of gravity. Had Newton predicted that "gravity is a downward force that will cause stuff to fall down" then the comparison to TENS would be valid. That is a qualitative tautology. But Newton made an exact quantitative prediction on the behavior of two objects of different mass on each other at different distances: F=G(m1*m2)/r**2 That is a prediction in advance of observing a trajectory and in advance of knowing distances and masses. Newton's theory of gravitation further refined and explained Kepler's 3rd law. While neo-Darwinism (TENS?) purports to incorporate Mendelian genetics, the two have yet to be reconciled in actuality (unlike Newton's correct refinement of Kepler's 3rd law).

A predictive theory is much more than explanations of particular events in hindsight. It also explains in comparable detail future results of an entire general class of phenomena. Newton didn't just explain why a particular apple fell (past tense). He predicted (with accuracy still relevant to modern astronomy, cosmology, ballistics, etc.) the future general behavior of any two masses on each other.

While you may lament an overly rigorous expectation on my part, words have meaning and "prediction" does not mean "observation" or "explanation." or "confirmation".

Scott Hatfield . . . . said...

Starwind:

You are just spinning your wheels here by attempting to impose (1) conditions that do not obtain in biology, and in the process you are obscuring (2) what really happened in this research program.

(1) Prediction does not require a mathematical formalism in advance, nor does it need the level of specification you seem to imply. Biology involves phenomena that are far more complex than those studied in Newtonian mechanics.

(2) Lejeune et al saw a pattern in the banding patterns of chromosomes that they thought could explain the difference in chromosome number between humans and other great apes in a way consistent with TENS-based scenarios of human evolution. Their proposed explanation, telomeric fusion, was a radical explanation in 1973, for reasons already given. And this was a falsifiable prediction: the sequencing data could've failed to confirm the existence of the embedded telomeres, or the presence of an inactivated centromere. The patterns that LeJeune et al inferred could've simply been artifacts of other processes, and the alleged pattern of telomere-telomere fusion an illusion, like the 'face' in an electrical outlet or the 'canals' of Mars.

How, then, is this not a falsifiable hypothesis? How, then, is the detailed exposition of sequence data confirming that hypothesis not, in effect, a fulfillment of the prediction that the sequencing data would, in fact, confirm same? Because you can bet your sweet bippy if it didn't the scientists would be eager to publish said result!

Anonymous said...

Scott Hatfield:
You are just spinning your wheels here by attempting to impose (1) conditions that do not obtain in biology, and in the process you are obscuring (2) what really happened in this research program.

Ok then, What was the research program expected to demonstrate, going in, and what conditions do you think should obtain in TENS (TENS predictability is the issue, not biology per se, but TENS) to be on par with other testable, predictive theories? I'll grant you as examples that super string theory lacks predictability, meteorology is near-term predictable, and relativity is highly predictable.

Prediction does not require a mathematical formalism in advance, nor does it need the level of specification you seem to imply.

Were we discussing the theory of flight I'm sure you'd have a different viewpoint. I'll grant you prediction need not be mathematical, but if it isn't formalized in advance with some specificity, then the "prediction" becomes whatever you or someone asserts it to have been to elicit agreement, or opponents assert to elicit disagreement. A case in point is the citing of HSA2 fusion as evidence of the prediction that has yet to be predicted.

Biology involves phenomena that are far more complex than those studied in Newtonian mechanics.

I went with Newton as an example because Robert had, and it is far more complicated than the classical gravitation formula would imply. Newton's application of his gravitation theory to Kepler's resulted in Newton's three laws of motion which application further required Newton's newly invented calculus (only Newton called it "fluxions"). If you're looking for comparable complexity, as previously mentioned consider meterology or fluid mechanics.

Lejeune et al saw a pattern in the banding patterns of chromosomes that they thought could explain the difference in chromosome number between humans and other great apes in a way consistent with TENS-based scenarios of human evolution.

Explaining the difference in chromosome number and predicting that difference (or showing why that difference is required by theory) in chromosome number, are two different concepts. Explaining is not predicting.

Their proposed explanation, telomeric fusion, was a radical explanation in 1973, for reasons already given.

Ok, but however radical an explanation of the phenomena it may have been, where was that explanation predicted or required before hand by TENS?

And this was a falsifiable prediction:

The explanation of telomeric chromosome fusion was (and still is) falsifiable, yes. But a prediction or requirment of TENS, no.

How, then, is this not a falsifiable hypothesis?

What hypothesis? You keep asserting the premise of TENS having made a prediction or hypothesis (aside from the tautological common descent) but emphasize observation and verification of the phenomena, without ever addressing where that phenomena was predicted, required, or hypothesized by TENS.

Just as a point of reference then, What are the hypothesese, predictions or requirements presently made by TENS (neo-Darwinism, modern synthesis, whatev), both observed and unobserved (but sought)? I assume a theory which makes predictions has a list of those predictions, and even if I might quarrel with the definition of "prediction", what comprises that list, presently? How will we recognize when the next predicted phenomena is observed? (Those are not a rhetorical questions).

Scott Hatfield . . . . said...

What hypothesis? You keep asserting the premise of TENS having made a prediction or hypothesis (aside from the tautological common descent) but emphasize observation and verification of the phenomena, without ever addressing where that phenomena was predicted, required, or hypothesized by TENS.

Starwind: Perhaps we are talking past each other, as the confusion referenced by Richard on this thread suggests. First, let me address (1)the question of what sort of testable prediction could be made from the 1973 paper, and then (2) the relationship of such a prediction to TENS itself.

(1) Now, any number of hypotheses could be advanced here as testable prediction based upon the 1973 paper in question. Here's one which Miller clearly alludes to as significant: that, when the chromosome banding patterns under discussion were understood at the level of the DNA sequence, that a region corresponding to an inactivated centromere would be identified.

Which it was, in the article "Generation and Annotation of the DNA sequences of human chromosomes 2 and 4" by Hillier et al which appeared in Nature in 2005, and which can be read in its entirety here.

The relevant passage reads as follows: "A search of genome sequence for the presence of vestigial centromere and pericentromeric sequences identified a 2.6-Mb region in 2q21.1-2q21.2 that is enriched for pericentromeric duplications in chromosomes 1, 7, 9, 10, 13, 14, 15, 18, 21 and 22 as well as a variety of centromeric satellite repeat sequence motifs....Within this 2.6-Mb interval, we identified a relatively large tract of satellite sequence ...which likely demarcates the location of the ancestral centromere."

Well? How can you argue that the model of telomere-telomere fusion doesn't implicitly carry the prediction that one of the ancestral chromosome's centromeres would've had to be inactivated, and that a signature of the inactivated centromere should still be present?

(2) Having said that, keep in mind that in this example what is being tested is not TENS as a theory per se, but a particular application of TENS within a research program. Remember, TENS is already well-demonstrated in specific cases where natural selection has been observed to lead to evolution, so the validity of TENS in those cases is not the issue. As Richard notes, TENS functions like an algorithm in this case, generating predictions like the above, which can then be tested.

If you understand that (and I suspect you do), aren't you just splitting hairs on the question of what exactly is doing the predicting? Surely you realize that LeJeune and his colleagues didn't propose the hypothesis of telomeric fusion as evidence for an act of special creation! They clearly did so in response to a model derived from TENS---right?

Quit avoiding the real issue! Doesn't this work demonstrate common descent between humans and great apes? Is there any way to square these results with the hypothesis of special creation?

Anonymous said...

Scott,

Two questions, one related to your debate, another perhaps off topic. At the outset I will claim complete ignorance regarding evolution and natural selection.

First, I'm trying to understand the basic assertions you're making here. Perhaps it's too simplistic to boil it down this much but I get the gist you're saying that:

1) In 1973, LeJeune et al made a prediction that human chromosome #2 was produced by "the fusion of two chromosomes which exist separately in chimps".

2) LeJeune et al made this prediction based on their understanding of evolutionary theory.

3) In 2005, Hillier et al provided evidence, newly available with the "complete DNA sequence of ... the chimpanzee", that indeed human chromosome #2 was produced by the fusion predicted by LeJeune et al.

And that you also agree that,

4) A fair definition of "prediction" is "a theoretical expectation in advance of observation or experimental outcome".

Is this an accurate summary of the example you've given? If not can you correct my misunderstanding but still keep it in the form of a fairly straightforward outline of the case you're making?



Second (and I admit this is off topic regarding your debate so I'm happy to take it to email if you prefer), something you said earlier prompted some thoughts regarding the distinction between natural and supernatural,

I don't rule out the possibility of some other agency at work, but as a scientist I confine myself to natural causes as a formal matter.
[...]
We all routinely assume natural causes for almost all sorts of activity or phenomena; science merely codifies this assumption as a boundary condition for doing science.


Is there any necessary reason for differentiating between natural and supernatural - particularly with regards to science? Whether an existing phenomenon is natural or supernatural, we can say that it does at least exist as part of reality. If we assume that there are so-called 'supernatural' phenomena, such as God, and we assume that so-called supernatural entities can affect the 'natural world', then don't the supernatural entities become natural by connection?

For a supernatural entity to affect a natural entity requires at minimum some connective mechanism between the two. And we ought not rule out that such a connective mechanism, and the supernatural entity itself, are observable, even measurable. In fact, shouldn't it be theoretically possible to observe and measure such things, if they do exist?

In other words, if supernatural entities exist, and if they exert a force on the natural - then why would we continue to call them supernatural instead of simply including them in the realm of the natural? It seems to me that in practice natural is synonymous with known and supernatural is synonymous with unknown. I hear this idea of a boundary in science between natural and supernatural all the time, but is such a boundary really needed? There are many unknowns, but science doesn't deem all unknowns as supernatural, does it?

To put it yet another way, there could be things that exist (e.g. angels) but which are currently unknown - why don't we assume that if they do exist, they are natural and therefore theoretically observable and measurable (just as with neutrinos, formerly only theorized but later observed)? And just let science proceed as it normally does without worrying about some kind of boundary?

Scott Hatfield . . . . said...

zroxx: Thanks for leaving a comment! Welcome!

You ask good questions.

First, I think your summary is pretty much the case as I see it. The 1973 paper proposed telomeric fusion as an explanation for human chromosome #2's banding pattern; implicit with this was a prediction that an inactivated centromere region from one of the ancestral chromosomes would be identified; the 2005 paper in Nature (Hillier, et al) satisfied this prediction, in my view.

Your second question is more difficult. It seems less a question than a rumination on what conditions must obtain in order for scientists to investigate the supernatural. You posit a connection of some sort between the natural and the supernatural, as if this would eliminate the conceptual barrier between the two. Is the connection you propose 50% natural, 50% supernatural? I would say the moment a causal connection to something natural is established, it would have to be natural as well---and so the conceptual barrier remains.

Bottom line: I think the domain of science and the domain of supernatural claims overlap, but I don't see how I can ever do science directly on a supernatural claim. I think I can only do science on some natural object/event/process, on some alleged or real consequence of this or that claim. Just how do you propose to bridge this gap?

Anonymous said...

Scott,

Thank you! Re: my first question, to further clarify, you're asserting that it is possible that the 1973 expectation was false - that we might have discovered that human chromosome #2 exhibited the trait for some other reason than telomeric fusion. And because of this, in your opinion it qualifies as a testable prediction? The capability to test the expectation was not available in 1973, but once someone was actually able to observe the complete DNA sequence of the chimpanzee, it became clear that the 1973 expectation was supported by the new observations rather than falsified. To a layperson like myself it seems you've provided an adequate example of a theoretical expectation in advance of observation or experimental outcome - I think if your opposition disagrees on that point they would best be served by focusing on the specific assertions you've made and building a case as to why one or more of them are incorrect. Nevertheless it would be interesting to read about other similar examples, if you have any.



Re: the second, yes perhaps it's more of a rumination. I have this mental picture of a rectangle divided between "natural" and "supernatural" - and science says we're not going to go near the supernatural, it's irrelevant to us. I see an almost identical picture of a rectangle divided into "known" and "unknown", but in that case science seems to say, we're going to expand, we're going to make the unknown into the known, and uncover everything we can. From a practical standpoint I don't see any real difference between "supernatural" and "unknown".

Now you've stated that you do believe in God, but when you operate in the realm of science you designate those beliefs as part of the supernatural and kind of partition that off. And I guess I'm wondering if there's really a logical reason for doing that. And I think you'd give an interesting perspective.

Let me try and give an example re: the neutrino. I assume what happened is that scientists had accounted for the known set of particles and forces to date and come up with what they thought should be a sound equation(s) or theory as to how those particles and forces work. And when they took that theory and compared it with prior observation and data, they found that it didn't quite fit. And they realized that clearly there was something they missed, some unknown particle or force that they hadn't accounted for. Now I would guess they then made many theoretical predictions that could explain this unknown, and one of those predictions was the existence of a currently 'invisible' particle. After some time, new obervations supported that predicition and falsified others until it became clear that there must be an existing particle, which was called a neutrino. Maybe I've butchered physics, but you get the idea.

Prior to any predicitions by these scientists, what was the actual difference between assigning the label 'supernatural' versus 'unknown' to the mystery they faced? The distinction seems meaningless, aside from our notion that supernatural, by definition, seems to imply that it is impossible to ever observe. So when you talk about how to bridge the gap, it seems to be a matter of simply saying, there is no gap. So X is unknown; whether someone's predicition involves a force exerted by God or a force exerted by some new particle, we'll deem any process of hypothesis and falsifiable test related to X as "valid science" and let the chips fall where they may.

Take intercessory prayer. If it works, then there must be some "unknown" force or particles that are operating in the natural world that we are currently unable to observe. Why should science say that's not a valid area of research because it's 'supernatural'? Or more specifically, why shouldn't persons who make theoretical expectations about forces/particles that are involved in intercessory prayer and then attempt to make new observations or experiments that falsify or support those expectations be considered scientists who are practicing valid science? Such exploration could lead us to discover the mechanism by which God exerts a change in the matter of ailing patients. Or perhaps it could lead us to discover some kind of mechanism by which our thoughts translate into a material change in a patient. Either one seems possible, right? Why set up an arbitrary boundary that causes people to shun scientific inquiry into such areas?

Excuse me if I rambled, but this is at least tangentially related to your debate... In the example of the human chromosome #2, is there any logical reason why you would designate someone as not doing valid science who predicts that, for example, God exerted a change in the material composition of the chimpanzee chromosomes in order to create Adam and Eve, and then conducts some tests in an attempt to support or falsify that prediction? Or someone who predicts that God simply created homo sapiens and chimpanzees with the chomosome compositions we see today, and then conducts some tests in an attempt to support or falsify that prediction? Of course I don't know if anyone is actually doing that, if the intelligent design people are doing that - but as a layperson it seems to me that it is possible that someone could make such predictions and try to conduct such tests. Regardless of whether you believe they'd be successful or that their predictions would be supported, is it not possible for them to practice "valid science" in an attempt to support their expectation?

Scott Hatfield . . . . said...

From a practical standpoint I don't see any real difference between "supernatural" and "unknown".

I think one of those philosopher of science types would say two things:

1) That the moment an 'unknown' becomes a 'known', it has effectively been naturalized; and

2) That you're not allowed to invoke any sort of 'unknown' as an explanation. A hypothesis, to be of use in science, must be falsifiable or have falsifiable consequences. There are many things, such as subjective impressions or supernatural claims, that are not falsifiable, or which have no falsifiable consequences. These things are excluded from science by definition.

And I guess I'm wondering if there's really a logical reason for doing that.

If I read you right, I think you're wondering about whether or not I wear a seamless garment of reason at all times, and if not, how this can be justified. Well: I don't, and I'm not sure that it can by appeal to reason alone.

The scientist who privately harbors religious sentiments which are outside the domain of science, if he or she is honest, must admit that they are compartmentalizing. Belief in the supernatural, faith in God etc. can not be justified through reason alone, but instead requires an intuitive leap. In a very real sense, faith is unjustified belief, and an element of fantasy (some would say self-deception) must always be present.

Now, that admission doesn't bother me personally, since I think it is actually normal to compartmentalize somewhat. It does bother others, and admire those who have attempted (I'm not sure that they've succeeded) to really put on that seamless garment. I just don't feel compelled to do that with every aspect of my life.

Now, both in the example of the neutrino or the chimpanzee chromosome, you suggest that the unknowns at that time (call them 'X1' and 'X2') could be called either 'unknown' or 'supernatural' with no effect on the conduct of future research. I would think that this only true to the extent that the workers involved do, in fact, compartmentalize.

If the privately-religious scientist is scrupulous on insisting upon falsifiable explanations, then you are right, it does not matter what they privately believe about the supernatural. If they are not, on the other hand, they are likely to go astray. Here's why: no matter what we discover about neutrinos (X1) or chimp chromosomes (X2), we are never ruling out supernatural intervention or involvement(Yx1x2). Conversely, the fact that aspects of X1 or X2 are unknown does not demonstrate the reality of either Yx1 or Yx2!

Anonymous said...

Appreciate your patience, this helps me understand where you're coming from.

That you're not allowed to invoke any sort of 'unknown' as an explanation. A hypothesis, to be of use in science, must be falsifiable or have falsifiable consequences. There are many things, such as subjective impressions or supernatural claims, that are not falsifiable, or which have no falsifiable consequences. These things are excluded from science by definition.

Let me rephrase what you're saying in the context of the neutrino - it would not have been valid science for someone to predict that, "I believe there is something I call 'g-force' that we can't see that explains why these particles aren't behaving the way our current equations predict, and that's that. Case closed.". It would be valid science for someone to predict that, ""I believe there is something I call 'g-force' that we can't see that explains why these particles aren't behaving the way our current equations predict, and I have an experiment that will isolate what I call g-force and allow us to observe and/or measure it to show that it exists and affects these other particles", that would be valid science (assuming the guy is telling the truth and the experiment isn't a hoax, etc).

In other words the differentiation is not whether the prediction is about something that is currently unknown (or 'supernatural') - it's about whether someone is actually following up their prediction with a valid test that relates to supporting or falsifying that prediction (about the unknown)?

no matter what we discover about neutrinos (X1) or chimp chromosomes (X2), we are never ruling out supernatural intervention or involvement(Yx1x2). Conversely, the fact that aspects of X1 or X2 are unknown does not demonstrate the reality of either Yx1 or Yx2!

That really clarifies it and makes sense to me. Basically, if a 'genuine' scientist wants to explore Y, then they must make predictions specific to Y and which test things about Y itself - and not merely assert something about Y and claim 'correctness'. Otherwise there's no (scientific) point in debating whether there is or isn't Y.

Thanks, best regards...

Anonymous said...

Scott Hatfield:
Doesn't this work demonstrate common descent between humans and great apes? Is there any way to square these results with the hypothesis of special creation?

As previously noted, tautologies are always conveniently true. Common descent is always demonstrated by common traits among presumed descendents. The issue you keep dancing away from is the lack of proof that those traits were conclusively acquired by vertical inheritence, alone. That is your a-priori premise, and it has colored your inferences and narrowed even your selection of the evidence from which you're willing to make those inferences, becoming your own evolutionary stable strategy.

Other kinds of mechanisms that may account for HSA2 aquisition are endosymbiotic gene transfer (see Endosymbiotic gene transfer - organelle genomes forge eukaryotic chromosomes) and lateral gene transfer (see Lateral gene transfers and the evolution of eukaryotes: theories and data). LGT has been shown (experimentally) in fungi to be possible for entire chromosomes (see Role of horizontal gene transfer in the evolution of fungi). One or more of these kinds of mechanisms may have been facilitated by zoonotic transmission from apes. There are numerous cross-species viral infections such as HIV/SIV and H5N1. Lastly, mitochondrial DNA studies have determined mutation rates (1/33 generations) that date the human MRCA to 6,500 y.a. (see A high observed substitution rate in the human mitochondrial DNA control region). Slightly slower mutation rates (1/40 generations) have also been observed (see How Rapidly Does the Human Mitochondrial Genome Evolve?). However disconcerting to evolutionists, followup studies have verified that such faster mtDNA mutation rates may be more correct than phylogenetically estimated rates (see Mutation Rate in the Human mtDNA Control Region).

I touched on these possibilities earlier, perhaps they were overlooked.

The HSA2 fusion may have been lateral gene transfer of the entire chimpanzee chromosomes 12 & 13 facilitated by a cross-species viral infection after the emergence/divergence of homo sapiens, without vertical inheritance from chimpanzees. As previously noted, the date of the HSA2 fusion event may be significant to reconciling its origin, and is not irrelevant.

Perhaps we are talking past each other, as the confusion referenced by Richard on this thread suggests. First, let me address (1)the question of what sort of testable prediction could be made from the 1973 paper, and then (2) the relationship of such a prediction to TENS itself.

Indeed we are talking past each other. I'm focusing on what specific theoretical predictions are made in advance by TENS, and you seemingly refocus on what has been explained by TENS after the fact. For you to cite Le Jejune findings as a truely testable prediction of TENS, a prediction that is not "backtesting" (as you assert) I would expect Le Jejune to have cited a prior TENS prediction related to human chromosomes being acquired from chimpanzees. But they didn't cite any such prior prediction. They expected that future studies by more detailed methods (gene sequencing) would confirm their banding analysis of the theretofore undiscovered HSA2 fusion. But you have relabled this expectation of re-confirmation as a new "prediction" of TENS. Nowhere has TENS made the specific testible prediction that an HSA2 fusion of chimp chromosomes 12 & 13 was expected or required by TENS. TENS is given credit for this "prediction" only after being discovered unexpectedly, an explanation masquerading as "prediction".

If TENS predicts that human chromosomes were acquired from chimpanzees, then what of the human chromosomes that have no discernable homologs with chimpanzees? Is that a failure of TENS? Whereas if TENS only generally predicted that some (as yet unspecified) human chromosomes (possibly) were acquired from chimpanzees, but then where was that prediction first published and was it published before the HSA2 fusion was discovered by Le Jejune et. al.?

We're told "biologic phenomena are far more complex than" phenomena predicted by other sciences. So complex in fact that what genetically evolves (species, genera, etc) are difficult to even define precisely and consistently, but regardless genetic evolution itself is ostensibly well understood.

Previously, you stated several times that TENS (the "Theory") made "testable predictions". Yet when asked where are those predictions, when were those predictions made, how specific were they, we're told its "algorithmic", not a "mathematical formalism in advance", doesn't "need the level of specification [implied by other scientific theories].

Arguably TENS is in the same category as porn - you may not be able to describe a TENS prediction in advance, but you'll know it when you see it.

The authors of superstring theory, the leading candidate for grand unification of relativity and quantum mechanics, (i.e. "the best theory we have", to borrow the Neo-Darwinist retort) have the intellectual honesty to admit it is neither predictive nor testable, presently, and may never be. Further, researchers of the Standard ("Big Bang") Cosmological Model as well have the intellectual honesty to even investigate "a beginning" while leaving the issue of a Beginner to theologians.

When TENS demonstrates comparable intellectual honesty and stops cherry picking its "predictions" and test cases, makes definitive specific testable claims in advance, addresses itself to other known mechanisms of genetic transfer, possible uncommon beginnings for humans and apes, and resolves or reconciles conflicting phylogenetic evidence such as molecular clocks (all of which is what puts the "hard" in the hard sciences), it'll be more like science and less like a political agenda.

Scott Hatfield . . . . said...

The issue you keep dancing away from is the lack of proof that those traits were conclusively acquired by vertical inheritence, alone.

I'm not dancing away from it. I was under the impression that you would like very much to undermine the case for any evolutionary model that would show that modern humans and chimpanzees were related. If this is not the case, you can say so.

As for endosymbiotic or lateral gene transfers? They are still examples of evolution, my friend. They are still examples which invoke common descent. The only difference is that, rather than visualizing the gene flow in terms of branching lineages, where such gene transfers occur the phenomena is more like a net. This makes analysis more difficult, but it doesn't mean that the organisms in question don't have ancestors, or that they don't share common descent with other organisms. So I fail to see the point of that line of reasoning.

I might add, further, that I am unaware of any evidence supporting a role for endosymbiosis or lateral gene transfer in the speciation event of any hominid, much less Homo sapiens. I wouldn't rule it out by any means: I don't have any a priori assumptions on this point, as you have incorrectly inferred. But 'extraordinary claims require extraordinary evidence', and I would invite you to present positive evidence that would support an alternative evolutionary hypothesis, rather than simply inveil that such examples exist and infer (incorrectly) that I have an a priori bias against same.

I might add, Starwind, that the emergence of Margulis's idea of endosymbiosis in a climate of hostility is a beautiful example of the self-corrective process of science, a process that pseudoscientists seem not only unable, but unwilling to place in context. I can remember, as an undergrad, reading textbooks that said, in effect, 'Here's an interesting idea; might be true.' Then, a few years later, graduate-level texts in microbiology were whistling a different tune: in effect, they were saying, 'This is a really cool hypothesis, with much evidence in its favor'. And now, a few years later, the text I use as a high school teacher refers to Margulis's idea as the 'theory of endosymbiosis'.

Now, as far as the mitochondrial stuff goes, I confess I don't know much about that. I'll read up on it. I can't help but notice, however, the curious 6,500-ya. date of recent human divergence. Are you seriously going to argue that beings like ourselves didn't exist before that? Are you going to argue that beings which thousands of years earlier were painting in caves, making ornamental beadwork and ritually burying their dead were not human? Because the archaeological evidence alone for human antiquity > 6,500 ya. is overwhelming. Even Ross and Rana (of RTB), who are eager to assert a recent supernatural origin for our species, have been forced in their book 'Who Was Adam?' to push the emergence date back thousands of years to account for this evidence.

Anonymous said...

Scott Hatfield:
I might add, further, that I am unaware of any evidence supporting a role for endosymbiosis or lateral gene transfer in the speciation event of any hominid, much less Homo sapiens.

Attributed to speciation specifically in primates, nor am I.

But 'extraordinary claims require extraordinary evidence'

My "claim" was that mechanisms existed that might explain the HSA2 aquisition without invoking vertical inheritance, and proffered links in evidence of that claim. My "claim" was made in response to your statement (as I copied and italicized) that [Le Jejune's HSA2 fusion work] demonstrated common descent.

My issue with "common descent" has been its tautological application as theory and evidence, and that the HSA2 fused chromsome is presumed (but not demonstrated) to have been inherited vertically, directly from great apes, and further that the timing is likewise presumed (but not demonstrated) to be irrelevant.

You then asked for ways to square Le Jejune's work with the hypothesis of special creation, to which I proffered the other unspecial mechanisms which might account for an HSA2 fused chromosome to be acquired not by "common descent" but more as a consequence of zoonotic transmission.

Like Margulis's idea of endosymbiosis, the claim may not seem so extraordinary in time.

When we deliberately genetically modify an organism, we don't consider the modification itself to have "descended" from a common ancestor nor do we proffer the modifcation itself as evidence that the entire organism descended from a common ancestor.

Likewise if an organism acquires a modification "laterally", that modification was not inherited and would not be evidence of common descent. Conclusive evidence of "common descent" requires more than the presence of common traits, especially in view of other possible mechanisms for those traits to have been acquired and conflicting (at present) evidence on their timing.

The HSA2 fused chromosome is a state or event, it is not a mechanism or process. Conclusive proof of "common descent" (an "algorithm" to borrow your terminology) requires that the process be demonstrated unambiguously and distinctively, and not merely pointing at the final state(s) or pattern from which the process or algorithm is inferred. While the evidence is consistent with the inferred process of common descent (being charitable about its meaning) that evidence pertains to the pattern or states and not how those states or pattern was generated.

And yes, that is likely more rigorous than you'd prefer.

As for endosymbiotic or lateral gene transfers? They are still examples of evolution, my friend. They are still examples which invoke common descent.

More precisely, they are mechanisms by which genetic change (evolution in its purest meaning) can occur without vertical inheritance. They are mechanisms by which traits can be commonly present across species which are otherwise reproductively isolated and did not decend from a common ancestor. I'll reiterate for your satisfaction such mechanisms have yet to be observed in primate speciation, though I'll note the search can't really begin without primate genomes (and ancient specimens to the extent they exist) being mapped, and useful genotypic and phenotypic defintions of species and genera, etc.

and I would invite you to present positive evidence that would support an alternative evolutionary hypothesis,

I've suggested mechanisms, I've cited problems in timing to be reconciled, and I've stated the degree of rigor which I expect from a testable hypothesis. I'm personally in no position to research and publish any substantiating findings - but then you knew that - so you may as well stop flogging that deceased eohippus.

the emergence of Margulis's idea of endosymbiosis in a climate of hostility is a beautiful example of the self-corrective process of science, a process that pseudoscientists seem not only unable, but unwilling to place in context.

Where "science" is concerned, you're preaching to the choir. Regarding "neo-Darwinism" (or its proxies) the process is indeed corrective but more liked dragged kicking and screaming. And I've noted there are obstinate pseudoscientists on both sides of the aisle, but no, I don't equate "rigorous" with "obstinate" :)

Are you seriously going to argue that beings like ourselves didn't exist before [curious 6,500-ya. date of recent human divergence]?

Hardly, but then that depends in part on what "beings like ourselves" actually means phylogenetically, genetically or phenotypically and whose mtDNA is used to measure. But I am seriously arguing the timing of the HSA2 fusion event is not as irrelevant as presumed, that there is a wide range of conflicting date techniques, and those generally perceived to be more accurate regarding homo sapiens sapiens produce sharply more recent time ranges from 100,000 y.a. to 40,000 to 6,500 y.a.

R. Moore said...

starwind--

Your attempt to mine the publications to find alternatives just demonstrates that you do not know what you are talking about. The articles you cite are not alternatives mechanisms to the subject of discussion, they are different mechanisms which lead to fundamentally different results. Nowhere do the authors of the work you mention suggest that that their results are alternative mechanisms to chromosome fusion. No scientist working in the field has suggested the research you mention offers an alternative mechanism to chromosome fusion. You are asking me to consider the hypothesis that the entire scientific community has been mistaken , and you, with a quick pub search, have stumbled upon a fundamental error in millions of man years of scientific research.

...a prior TENS prediction related to human chromosomes being acquired from chimpanzees...


You keep saying humans chromosomes were acquired from chimpanzees. It this your real understanding? Humans and chimpanzees shared a common ancestor. A variant of this ancestor (our ancestor) fused 2 chromosomes. Siblings of this common ancestor (the chimpanzees' ancestor) did not fuse the chromosomes.

R. Moore said...

zroxx

As Arthur C. Clarke said --

"Any technology sufficiently advanced is indistinguishable for magic"

His point is is that it is not magic (or supernatural).

You suggest a cause and effect (connection) between a supernatural entity and the natural world, making the supernatural natural. Hmm. What caused the supernatural to create this connection. Something super-supernatural? And what caused...? You get my point. The natural world has boundaries. Anything supernatural exceeds the boundaries, and cannot be connected. Enjoy thinking about, but remember, "it's monkeys all the way down!"

Anonymous said...

Richard:
Nowhere do the authors of the work you mention suggest that that their results are alternative mechanisms to chromosome fusion.

Nor did I claim so. I have never questioned the mechanisms of chromosome fusion, nor HSA2's homologs to chimp 12 & 13. In fact, I wrote earlier:
"Yes, the evidence for the HSA2 fusion is unarguable (and wasn't being argued). What is questioned is the premise that those chromosomes could only have been inherited from a common ancestor (as opposed to many families or orders having had a common complement of chromosomes to begin with). Almost all researchers on this subject are rigorously careful to state they "infer" a history, rather than demonstrate it. They seem less certain than you."