Panderichthys rhombolepis

One quick point: The problems run deeper than the professor suggests. By divorcing coelacanths from lungfish, the molecules contest not only the branchings within the trichotomy, but the trichotomy's very existence. This overturns the central predictions of evo biology. If the science can't even get the basics correct, why trust the details?

Common Descent may rest on many evidences, but defining a "good" tree by how robustly it supports evolutionary relationships, and then using that tree as independent evidence for evolutionary relationships, is circular reasoning at its most transparent. What if "star" phylogenies represent reality better? How would a Darwinist know? Answer: he wouldn't, because a tree full of long branches is tossed in the wastebasket.

— Ghost of Paley

Guts to gametes, Mr P?

If the science can't even get the basics correct, why trust the details?

One quick point: The problems run deeper than the professor suggests. By divorcing coelacanths from lungfish, the molecules contest not only the branchings within the trichotomy, but the trichotomy's very existence. This overturns the central predictions of evo biology. If the science can't even get the basics correct, why trust the details?

Please, if you're going to use the term 'trichotomy' in the context of systematics, please use it properly. 'Trichotomoy' usually refers to an unresolved phylogenetic tree (that is, a dendrogram with three branches to a common node, rather than two to a node). What is this 'trichotomy' that you keep talking about?

Most molecular analyses I'm familiar with support a sarcopterygian status for coelacanths. Why should we conclude that the data used to generate any result are to be the basis of any conclusion?

Certainly, this is a question I must answer as well as you. However, you seem wont to to use phylogenetic analysis in making conclusions with much more certainty than I ever would. You simply cherry-pick the phylogenetic result you want and use it as the basis of an argument claiming that evolutionary biologists are committing that very same fallacy.

GOP,
Why do you never clearly state your position, so that even somebody as ignorant as I am can see it?

You always seem to want to play things out as a game.

Why not simply give your hypothesis and back it up with evidence in a single post?

Why not simply give your hypothesis and back it up with evidence in a single post?

Posted by 'Rev Dr' Lenny Flank on January 25, 2006 10:11 PM (e) Because he can't. (shrug) ID simply has nothing scientific to offer. And he knows it just as well as we do.

The reason more molecular studies don't show Stuart's result is because most molecular studies avoid including ray fins with both lungfish and coelacanths. Stuart et al didn't, and got the surprise of a lifetime.

— Ghost of Paley

The resolution of this particular analysis may be limited by the relatively small number of independent characteristics (64---80 copep motifs) extracted from the 13 gene families of mitochondria. Many more independent characteristics would likely be available from larger data sets containing whole bacterial or nuclear genomes.

— Stuart et al. 2002

I prefer molecular studies because they alone can rigorously define a character and subject it to independent statistical tests. Perhaps this explains why so few of them support evo wishful thinking. And as I've mentioned before, the Stuart study was the most comprehensive analysis available. No cherry-picking on my part. And why do you keep questioning my character anyway? The studies stand or fall independent of who cites them, or so I've been led to believe. More later.

While criticisms of the NJ method may have merit, other tree-sorting methods don't seem to fix the problem; oddly enough, more powerful techniques often make things worse. Although one molecular synapomorphy unites the coelacanth with his putative lungfish cousin, Maximum and log likelihood analyses argue that "Old four-legs" is a basal skate. For some reason, some scientists omit the coelacanth from their studies, relying solely on lungfishes as their representative sarcopterygian. I predict that a heat map analysis would recover a star phylogeny for fish, and that a comprehensive study of all shared insertions would muddy the phylogenetic waters. More later.

Please remove the question mark from the web address on the second link if you want this paper to show up.

While criticisms of the NJ method may have merit, other tree-sorting methods don't seem to fix the problem;

oddly enough, more powerful techniques often make things worse.

Although one molecular synapomorphy unites the coelacanth with his putative lungfish cousin,

Maximum and log likelihood analyses argue that "Old four-legs" is a basal skate.

For some reason, some scientists omit the coelacanth from their studies, relying solely on lungfishes as their representative sarcopterygian.

I predict that a heat map analysis would recover a star phylogeny for fish, and that a comprehensive study of all shared insertions would muddy the phylogenetic waters. More later.

Again, Martin, the rest of us here at PT have very much appreciated your willingness to post and your doggedness in responding with accurate information, even to a dogged troll--definitely a free education for us all (though PaleThing will be smarting from his "education").

Nah, GoP will continue to be oblivious. It was fun watching him get casually smacked around, though.

Nah, GoP will continue to be oblivious. It was fun watching him get casually smacked around, though.

— P.Z. Myers

Good ol' Arnason et al. Two major problems are apparent here. For one, it's clear that you don't know how to read trees (this explains alot). Their result has coelacanths, chondrichthyans, and teleost fishes as an unresolved polytomy. The coelacanth is not resolved as a basal skate

— 'Ripper' Brazeau

Similarly, the crown position of Chondrichthyes in the piscine tree, rather than as the sister group of the bony fishes, negates the strict cladistic definition of the Osteichthyes as including all extant gnathostomous vertebrates except Chondrichthyes. The use of the name Sarcopterygii to describe both lobefinned fishes and terrestrial vertebrates is also unjustified [I'll say! G.o.P.], as the jawed fishes and the amniotes form separate branches in the gnathostome tree (Rasmussen et al. 1998).

However, this does not imply that tetrapods do not have piscine ancestors. On the contrary, it is probable that extinct piscine taxa (such as Elpistostegalians, 'Osteolepiforms' and Rhizodontiforms) constitute the forerunners of the tetrapods, with the two groups forming the Tetrapodomorpha (see Janvier 1996; fig. 5.3, branch 5). This interpretation is consistent with Jarvik's (1980) theory that the 'osteolepiform' Eusthenopteron was anatomically close to the hypothetical ancestral gnathostome.

More importantly and perhaps deeply embarrassing for you: A similar and more recent study by Brinkmann et al. 2005 (which still stands as a more recent study) does not find this. Unlike Arnason et al., Brinkmann et al. don't completely ignore tetrapods! You know, in order to test the question of who is more closely related to tetrapods, actual tetrapods are likely to be an important source of data and should be included in the analysis!

— 'Ripper' Brazeau

The reason more molecular studies don't show Stuart's result is because most molecular studies avoid including ray fins with both lungfish and coelacanths. Stuart et al didn't, and got the surprise of a lifetime.

— Ghost of Paley

This is quite simply false. In fact, it's either incredibly ignorant or a lie. Any analysis of sarcopterygian interrelationships generally includes ray-finned fishes. Usually Polypterus and or Polyodon where sequences are available. Certainly always Fugu and/or Danio.

I think I'm done here.

The wingnuts are always at their most amusing when they blithely display not a clue as to how fast their little wings are spinning...

I second the rather hopeless desire that Ghost make his position clear. I'm trying very hard, as a nonspecialist, to follow the discussion, but I'm having a lot of trouble extracting the underlying positions. Brazeau seems to be saying that different techniques produce different possible cladistic relationships, that these techniques each have strengths, dangers, and drawbacks, and that "best-fit" phylogenies are thus hard to pin down; there's no slam-dunk surefire history.

Ghost, on the other hand, seems to basically agree that there are debates, ambiguities, and difficulties here. But rather than argue in favor of one particular interpretation of insufficient/ambiguous evidence over another, I guess Ghost is saying that if we don't know what actually happened, somehow - what? -*nothing* happened? Magic happened? We shouldn't bother to make the effort to resolve these questions? If we don't know exactly what branched from what when, then branches didn't happen?

Beats me. Is Ghost trying to use unresolved issues at the bleeding edge of biology as support for the position that evolution never happened at all? Or what?

Is Ghost trying to use unresolved issues at the bleeding edge of biology as support for the position that evolution never happened at all?

"What's being pushed is to have Darwinism critiqued, to teach there's a controversy. Intelligent design itself does not have any content."

More importantly, the coelacanths occupy a crown position with skates.

— For some bizarre reason, Ghost of Paley

Especially when it has such rich taxon sampling!

— Ghost of Paley

True enough - but this doesn't help you at all. The study clearly groups coelacanths, teleosts, and cartilaginous fishes in a tight trichotomy while moving lungfishes to a suspiciously basal position. No tetrapods needed, since coelacanths obviously wouldn't cooperate in a mitochondrial analysis anyway. More importantly, the coelacanths occupy a crown position with skates. Could it get any more embarrassing for our poor little evos?

I thought Santa Claus was real until I was 17 until some mean person told me he didn't exist. Paris Hilton.

And here, again:

Especially when it has such rich taxon sampling!

— Anton Mates

while linking to this, evidently intending to imply that the Brinkmann et al. paper didn't look at enough taxa. Of course he somehow overlooked that this was only one of three trees the paper contains, and that the biggest one, which also supports their conclusion re: Everything Else(Coelacanths(Lungfish Tetrapods)), contains 22 species---as many species, in fact, as the Arnason et al. paper GoP was previously touting, and drawn from a far richer array of vertebrate taxa. Including, as Brazeau said, actual tetrapods, which is kinda helpful.

Steviepinhead:

Sorry for the delay in responding. I dont think we need to worry that this is off topic, since this thread is way off the front page anyway. If it wasn't for GoP's pointless bickering with Brazeau, this thread would have died long ago.

Anyway, I basically agree with your appraisal of the Enrico piece. Congrats on having made your way thru it -- it's extremely technical. If you dont mind my asking, what is your degree in? Have you had some ling classes? I was basically able to follow it, with the qualification that I don't know much at all about Pacific Northwest languages.

My basic feeling about the Greenberg & Ruhlen show is that they seem to believe that the old 'garbage in/garbage out' rule is overridden if you put ENOUGH garbage in. That is, if you posit 1,000 etymologies, it doesn't matter if 999 are shit, so long as one is valid -- the one valid etymology proves the whole thing, and refutes all your opponents. It's not too different from the ID modus operandi.

Another lesson from the Enrico article is just how vastly improved the results of the research are when the researcher has expert knowledge of at least one of the languages involved. (As Bloomfield once said, "if you are going to compare two languages, it helps to know one of them".) This lack of deep expert knowledge in, well, anything, is another fatal flaw in the Greenberg/Ruhlen approach. And again, the analogies between that and the ID movement (indeed, elsewhere within this very thread!) aren't hard to spot.

That said, my personal feeling is that Enrico has correctly revived the debate over the relationship between Haida and NaDene, but I still don't see how several of his conclusions follow. He posits numerous historical scenarios to account for the current linguistic situation seen in the Alaska/NW British Columbia area, such as positing that the Tlingits were a 'high status group' of the Haidas once upon a time, but he doesn't offer evidence for these notions. I think if you want to use historical explanation to buttress a linguistic argument, you have to have something more than linguistic evidence for it. So I feel that Enrico sort of slips a bit when he tries to be a historian.

w

For a start, I'm not looking for help. If you think you're holding my on trial here, think again. Each post you make reveals your deep ignorance masked in a veneer of freshly dredged references. Branch statistics are only as good as the data distributions they represent. Change/add taxa (especially large, well-sampled groups) and the distributions will change. I should not have to tell you this

— Martin 'The Rippa' Brazeau

I basically agree with your appraisal of the Enrico piece. Congrats on having made your way thru it --- it's extremely technical. If you dont mind my asking, what is your degree in? Have you had some ling classes? I was basically able to follow it, with the qualification that I don't know much at all about Pacific Northwest languages.

My basic feeling about the Greenberg & Ruhlen show is that they seem to believe that the old 'garbage in/garbage out' rule is overridden if you put ENOUGH garbage in. That is, if you posit 1,000 etymologies, it doesn't matter if 999 are shit, so long as one is valid --- the one valid etymology proves the whole thing, and refutes all your opponents. It's not too different from the ID modus operandi.

Another lesson from the Enrico article is just how vastly improved the results of the research are when the researcher has expert knowledge of at least one of the languages involved. (As Bloomfield once said, "if you are going to compare two languages, it helps to know one of them".) This lack of deep expert knowledge in, well, anything, is another fatal flaw in the Greenberg/Ruhlen approach. And again, the analogies between that and the ID movement (indeed, elsewhere within this very thread!) aren't hard to spot.

That said, my personal feeling is that Enrico has correctly revived the debate over the relationship between Haida and NaDene, but I still don't see how several of his conclusions follow. He posits numerous historical scenarios to account for the current linguistic situation seen in the Alaska/NW British Columbia area, such as positing that the Tlingits were a 'high status group' of the Haidas once upon a time, but he doesn't offer evidence for these notions. I think if you want to use historical explanation to buttress a linguistic argument, you have to have something more than linguistic evidence for it. So I feel that Enrico sort of slips a bit when he tries to be a historian.

ID makes very few (if any) novel assertions...

I think I have it now. I live in a town of perhaps 200,000 people. 30 years ago, it was a town of 60,000 people or so. Growth has been pretty explosive, and it's not easy to find people who were born and raised here. I certainly wasn't. Now let's imagine a study trying to determine how those not born here actually got here. We start at the physical level...

We look at Joe Blow. He doesn't have much money, he doesn't own a car. No passenger trains stop here. So we speculate that he arrived either by bus, or as a passenger in someone else's car. It's not a sure thing, but it seems most likely.

But wait! Here comes Ghost, pointing out that our speculations are dubious in some respects. Our basic problem, says Ghost, is that we are assuming Joe got here by some form of transportation, and we are presuming the existence of transportation because Joe is here. Us poor 'transpos' are using circular reasoning.

Now Martin points out that our analysis is the best we can do based on limited data. Yes, Joe *might* have flown in, he *might* have previously owned a car but has it no longer, but when we're looking at tens of thousands of Joe Blows, we can make fairly accurate probabilistic analysis.

Ghost now counters that he has examined the alleged bus station and airport for milliseconds on end and seen no sign of the purported "transportation". Which he finds entirely to his expectations. Instead, he has found a study showing that individuals are actually capable of walking great distances.

Martin replies that yes, walking is indeed possible and perhaps someone in the current population actually used that method, but studies examining the surrounding roads and countryside have found that pedestrians are nearly nonexistent. Certainly the are not present in anywhere near the numbers necessary to explain the population growth. These same studies find vast multitudes of highway traffic. Surely that traffic is strongly indicative of how people got here.

Ghost responds that he has found multiple studies showing that this can't be the case. He has studies of vehicular reliability showing millions of failures, studies showing a ferocious traffic accident rate, studies of drunks and teenagers, all carefully linked and cited. So clearly the "highway traffic" tale is at the very least subject to multiple powerful countervailing evidence. Which of course 'transpos' choose to ignore, because it doesn't fit their doctrine.

Martin now tries another tack. An increase in population by a factor of 3 over 30 years exceeds any rational human breeding rate. Studies of specific characteristics of individuals, especially speech characteristics, strongly suggests that only a small minority of the population could have been raised in Alabama.

Ghost replies with studies showing that speech is an unreliable indicator of origin. And another study showing breeding and population growth rates indeed DO show a tripling in 30 years in many places. It's entirely unexceptional.

Martin says, well, speech is only one characteristic. We all have reliable data about age, dress, income, skill sets, etc.

Ghost produces a study showing the wide range of human variability.

Martin produces actual testimony from a sampling of individuals, who say they came by car (some by plane), and where they came from.

Ghost produces studies showing that people's testimony is unreliable, along with a great many documented cases of mendacious claims. Taking testimony is evidence not of history, but of tester's gullibility. But of course, gullibility is guaranteed when the alternative is for poor little dogmatic "transpos" to admit error.

So the battle goes back and forth. And it's clear that Martin is trying to figure out how people got here. What Ghost is trying to say isn't clear at all.

What's funny about this complaint is that I was originally going to link to the first figure until I saw that it was based on only one gene. Fearing that some would accuse me of ignoring the tree drawn from the concatenated RAG 1/RAG 2 data, I chose the more comprehensive phylogeny instead, and still got flamed.

But in any case, my point stands: the paper did not include many fish species, and this compromises its conclusions.

And what about the other paper?

Which is precisely my point. The lability of phylogenetic branches bedevils Darwinists, since even the sturdiest of branches may collapse at a taxon's notice.

— Ghost of Paley

That's why the Arnason study prickles at the conscience:

what would happen to Brinkmann's delicate Bonsai if he allowed the addition of a few ray fins?

No one knows, and few wish to try.

The only recourse is to rely on the dreaded Neighbor-Joining algorithm, or narrow the study's scope. One thing is certain: Brinkmann needs more genes to support his position.

Time is limited, so this post will be a little crude.

Anton Mates:
You still miss my point. Brinkmann's 22 species were distributed across a large phylogenetic space. This is entirely different from a more focused approach, because adding more taxon within a family often "breaks up" monophyletic groups and pushes some members of the old clades towards others. Also, his approach leaves him vulnerable to the charge of cherry-picking "well behaved" species that exhibit "tree-like" behavior, instead of including all sequenced organisms and letting the chips fall where they may.

Even slowly evolving genes have their problems, so no, one must always use whole genomes or multiple genes when testing a phylogeny. Read Theobald if you don't believe me. If you wish, I'll also throw some references your way.

Mr. Brazeau:
Sorry, but I'm still skeptical. If lungfish and coelacanths share a close relationship, this should be robust over different analyses, and shouldn't crucially depend on including tetrapods. I plan on supporting this claim a little later.

More to come....

You still miss my point. Brinkmann's 22 species were distributed across a large phylogenetic space. This is entirely different from a more focused approach, because adding more taxon within a family often "breaks up" monophyletic groups and pushes some members of the old clades towards others.

Also, his approach leaves him vulnerable to the charge of cherry-picking "well behaved" species that exhibit "tree-like" behavior, instead of including all sequenced organisms and letting the chips fall where they may.

Even slowly evolving genes have their problems, so no, one must always use whole genomes or multiple genes when testing a phylogeny. Read Theobald if you don't believe me. If you wish, I'll also throw some references your way.

More to come....

Sorry, but I'm still skeptical. If lungfish and coelacanths share a close relationship, this should be robust over different analyses, and shouldn't crucially depend on including tetrapods. I plan on supporting this claim a little later.

Make sure you have an exit strategy here Martin. Don't wait for "Ghost of Paley" to understand he's wrong and concede.

1) Tetrapods are lacking. A source of synapomorphies for either of the sarcopterygians included is lacking. This has already been discussed. It is far and beyond the most important point; by itself totally invalidating the results of Arnason et al. for this discussion. Nevertheless, we soldier on. 2) The 'problematic' taxa represent very disparate and undiverse lineages. They are thus unlikely to provide a rich source of 'good synapomorphies' linking them closely with any taxon. Notice how very diverse clades with lots of closely related sub-lineages (such as the teleosts and chondrichthyans) clump together rather nicely and form the monophyletic groups we expect. The taxa that always seem to 'fly around' from one analysis to the next tend to be these lonely and disparate taxa.

— 'Ripper' Brazeau

3) There is a large, unresolved clade of coelacanth, chondrichthyans (sharks, rays, and chimaeroids), and teleosts (advanced ray-fins). This suggests that the information in the preceding plesions (lungfish and bichir - what I'll call the 'immediate outgroup') and the outgroup are insufficient to polarize this ingroup. There are many conflicting signals that are resulting in all three possible topologies emerging in the analysis. This is a problem that would be expected to emerge if, say, you had an inappropriate immediate outgroup! The lungfish and coelacanth needn't clump together because, in the absence of tetrapods, symplesiomorphies are probably being reconstructed as synapomorphies. With tetrapods, a 'bushier' branch is provided that could pull over coelacanths and lungfishes and reveal their synapomorphies which are currently getting swamped by a erroneous optimization.

— 'Ripper' Brazeau

I understand. I would find this explanation more convincing if there was a way to quantify this effect beforehand. Otherwise, the researchers can just fiddle with the outgroup and taxon density until they achieve the results they wish (not that I'm implying conscious cheating).

Sell it to the hillbillies at Florida State, whose intrafamily squirrel tree makes do with a mountain beaver outgroup. Seems mighty undiverse to me.

— Ghost of Paley

If this is what you're saying, OK, I understand, but what point are you trying to make? If different results can be defended, and a "currently considered most likely" result can be selected, you still can't conclude that the selected result was wished for ahead of time. My reading of the history of science is that there is *always* debate about new material, difference of opinion, plenty of scope for further testing and hypothesizing, and some particular "most likely" always exists, even if it's in rapid flux.

I understand. I would find this explanation more convincing if there was a way to quantify this effect beforehand. Otherwise, the researchers can just fiddle with the outgroup and taxon density until they achieve the results they wish (not that I'm implying conscious cheating). But I'll comment more on this later.

— Ghost of Paley

How about a scientific theory of design, that isn't empty of content, and can be tested using the scientific method. Is that "to come"? Or are y'all, uh, still working on that ...? When Luskin told us that there is a "positive case for design" that does NOT depend merely on "negative arguments against evolution", was he just BS'ing us?

I really shouldn't be doing this now, but as usual, Ghost isn't actually reading the papers he cites. That study used eleven outgroup species from the Muridae (true mice and rats), Dipodidae (jumping mice) and Aplodontidae (mountain beaver). The only outgroup they bothered to show on their tree was Aplodontia, because it turned out to be the most closely related:

— Anton Mates

I sure picked the wrong study to skim, didn't I? Thanks for catching the mistake. My essential point stands: the authors stuck within one order (or even one suborder, by some rodent classification schemes). According to Mr. Brazeau, this should result in a funky tree.

Re "(sound of crickets chirping)"

Yeah, but they're still crickets!

(Ducking for cover.)

How so? The FSU group had a decent outgroup size and few disparate ingroup taxa with deep independent lineages.

— Martin Brazeau

I believe he's basically making a claim about convergence. Biologists---and of course scientists in general---expect their interpretations to converge over time as new data is found and old data analyzed more rigorously. As you say, there's always debate, but one hopes that the bigger claims will one by one be settled, so that only the fine detail has to be resolved. GoP is arguing, I think, that the addition of new data has not and will not assist the convergence of biologists' common descent models, because no such model is actually valid. Hence his prediction of star phylogenies---eventually biologists will be left saying, "Well, we don't see any real relationships between any of these guys, and we can't fit their similarities and differences into a hierarchical pattern, but we still assume they all descended from something."

— Anton Mates

Of course even if that happened it would not provide positive evidence for the specific creation model provided by (interpretation X of translation Y of) the Bible, or whatever variant of that GoP supports. And the data he's presented to support that claim are frequently misinterpreted or simply false. But to his credit, I think he is making a testable claim.

Still waiting, Ghost .... (sound of crickets chirping)

— Reverend Jim

I predict that a heat map analysis would recover a star phylogeny for fish, and that a comprehensive study of all shared insertions would muddy the phylogenetic waters.

So Arnason's study (which used all three fish classes) was less diverse than the intraordinal squirrel study? How do you figure that?

— Ghost of Paley

Or are you saying that the number of outgroups is the determining factor, even if the outgroups are closely related? If Arnason had sampled more lungfish, then would everything have been OK?

But wouldn't star phylogenies provide positive evidence for multiple special creations?

So Arnason's study (which used all three fish classes) was less diverse than the intraordinal squirrel study? How do you figure that? Or are you saying that the number of outgroups is the determining factor, even if the outgroups are closely related?

— Ghost of Paley

If Arnason had sampled more lungfish, then would everything have been OK?

Interesting but not surprising, Anton and I wrote pretty heavily overlapping responses.

By the way, I wanted to say this before, please call me Martin. "Mr. Brazeau" is way too formal.

Interesting but not surprising, Anton and I wrote pretty heavily overlapping responses.

— Martin Brazeau

What's a star phylogeny? Is that where several (more than two) lineages appear to have split off from their latest common ancestor at so close to the same time that we can't distinguish the order of separation?

Henry

? The lungfish weren't the outgroup in the Arnason et al. study. Four species were---sea lamprey, lancelet, starfish and sea urchin.

— Anton Mates

For those for whom this thread has gotten a little over-technical, or for those who would like some idea of what all the fuss is about, the December '05 issue of Scientific American had a good article on the fish ==> tetrapod transitional fossils, and the current phylogenies.

This appears to be a very active area of research, as well it should be. Scientists are learning more all the time about which environments were conducive to this kind of evolution, as well as refining the information available from already-retrieved specimens (several interesting fossils have been "discovered" in museum drawers, where they were languishing).

Both processes help bootstrap further research and fieldwork (knowing the kinds of environments these animals lived in helps to tighten up the not-yet-sampled locations and formations that might yield fossils; likewise, discovering a previously-unrecognized transitional fossil furnishes another clue, as the formation that produced it may be worth revisiting).

A mass of info has been uncovered in only the last decade or two--in the last year or two!--and there's every reason to expect more exciting news about these critical critters!

To see my latest response (it's well worth the effort!), please go to the LUCA thread on "After the Bar Closes" (it's on page 46). The software wouldn't let me post here, so I had to transfer it. Sorry for the inconvenience, and obviously, leave all replies here.

What's a star phylogeny? Is that where several (more than two) lineages appear to have split off from their latest common ancestor at so close to the same time that we can't distinguish the order of separation?

— Henry J

Phylogenetic trees are constructed under lots of assumptions

— Anton Mates

I've got a test to write on Thursday. Will be back by the weekend. Sorry for the hiatus.

In the meantime...

Paley, explain for us why you are so confident in particular analyses. I don't want to know why you prefer them. I want to know why you are so confident in the validity of their results.

For the record, please don't synonymize my views on molecular phylogeny with Brinkmann et al. I don't assert the same level of confidence in single analyses as you do.

Paley, explain for us why you are so confident in particular analyses. I don't want to know why you prefer them. I want to know why you are so confident in the validity of their results.

— Martin

What Ghost is saying is, at the cutting edge of science, you would expect this sort of debate - very few of the results have come in yet, and those that have come in aren't particularly reliable, so there's lots of scope for debate. Ghost's point is that over the last decade or so, a very large amount of additional results have become available, adding a wealth of genetic and molecular analysis evidence to the existing morphological evidence. And yet these trees are NO CLOSER to resolved than they were before. Which lineages are included in the sample change the apparent relationships among other lineages. Different analytical techniques using the same data produce very different trees. Ghost's argument is that when a wealth of additional information becomes available, and when that information is a great deal more reliable, and we STILL can't build trees any more robust than ever, maybe the problem is that our assumption of trees is wrong in the first place. We can't produce good reliable trees because there are no trees to produce - the data stubbornly refuse to fit our wrong assumption.

Does this answer your question?

— Ghost of Paley

To which two counter-arguments have been presented over at PT. First, that there has indeed been a trend toward a solid consensus as more information comes in; the tree model seems to be working just fine. And second, that EVEN IF the tree model is wrong, this lends absolutely no positive support to the 'poof' model. Ghost has at this point been reduced to claiming the consensus as to phylogenetic trees isn't very solid, and for the good reason that God didn't do it that way. Instead, God created 'kinds' that have been generally milling around their Platonic centers. And therefore attempts to find that one 'kind' evolved out of another is doomed to the kinds of problems cladists are suffering; they're looking for what didn't happen. Now, if only we'd read the freepin' BIBLE, we'd have known this all along and saved ourselves all this confusion and heartburn.

— ..in continuation Flint

Selective quotation from my good friend Flint only shows that you are sinking fast GOP. There is a hole in your Googletrawler perhaps?

— The Yenta

So Paley you don't disagree with the rest of Flints quote? ....interesting

By the way, Mr. Martin, I really don't understand your question. Could you please rephrase it? Thanks.

— Ghost of Paley

You need to establish why you think a particular analysis ought to be a good estimate of phylogeny and why you think the conflicts actually amount to evidence against evolution.

— Martin

You also need to judge the appropriateness of the data. Some types of molecules are used for higher-rank (i.e. phylum level) phylogenetic estimates while others are used for lower-rank (i.e. species or family level) estimates because of their degree of polymorphism. So, when a particular analysis sets out to test higher order interrelationships and uncovers lower order conflicts (or vice versa) are you taking such factors into account? If, given a particular data set, you equate the displacement of chondrichthyans with the displacement of an individual plesion such as a coelacanth you're gravely mistaken.

"Aaiiieee!," shrieks The Thing Called Paley. "I've come to a curb! Measurements are inexact! It's probably only six inches down, but how can we ever really know? And the far side of the street is, is, another inexact distance, one virtually unfathomable, given the multiple inexactitudes of measurement, the unreliability of instrumentation, the uncertainties of temperature, pressure, and the uncertain structural integrity of the untested materials that span the predatory void between this perilous curb and that other, possibly illusory mirage of an apparent curb, far over yon! Whatever is a poor pedestrian to do?"

Our knowledge about any situation is never going to be perfect, precise, or complete, Fading Fast. That doesn't mean--to most of us--that we're not entitled to do the best we can with what we've got.

But that's okay, reality-wracked one! Stay at home with the cat ladies and the newspaper-collecting people.

While you wait for the firefighters to show up to pass you out your apartment window on a stretcher and take you away to a very safe place where you can be alone with the pristine forms that vibrate harmonically inside your head, the rest of us will just have to take our chances with that getting-somewhere-in-the-world-despite-reasonable-risks thing that we call life.

Stevepinhead
Paley is impervious to abstaction although I think he probably dreams of electric sheep. Good little robot.
I think the Grand Old Designer painted one of those bosons in the big bang with Paleys name just to remind everyonelse how lucky they were.
Oh thats right Paleys so smart that that silly boson could not have found life on earth I mean what if it ended up in a black hole ?
Soo he must have really got some dirt under his finger nails .......oh Calvin dinner now and don't forget Hobbs!

But that's okay, reality-wracked one! Stay at home with the cat ladies and the newspaper-collecting people.

— Vic

While you wait for the firefighters to show up to pass you out your apartment window on a stretcher and take you away to a very safe place where you can be alone with the pristine forms that vibrate harmonically inside your head, the rest of us will just have to take our chances with that getting-somewhere-in-the-world-despite-reasonable-risks thing that we call life.

Attaboy, Paley! Though you're still poised, paralyzed and trembling, above that perilous step, you've still got the cajones to launch your spittle at those walking by.

Who of course avert their heads, thinking, "Poor fellow! If only everyone had access to quality psychotherapy!"

Paley
How much? Oh 3 gods of the gaps ...or bosons whatever you can't imagine.
You know if you are having trouble with creation can I suggest the Kama Sutra?

Re "frequently used genes have already been certified kosher by Darwin"

And here I could've sworn that Darwin didn't know about genes... ;)

ghost wrote: I saw the cladogram, but this does not represent the majority point of view, morphology-wise. I think a careful reading of Zimmer's book will highlight the difficulty in the Lungfish/Amphibian transition; in fact, I think he makes the same point about lungfish skulls that my source does. In any case, check out the evo predictions prior to the molecular studies - it's a classic case of an ad hoc adjustment to avoid falsification.

frequently used genes have already been certified kosher by Darwin

I had no idea Darwin was Jewish...

I had no idea Darwin was Jewish...

To catch people up: my latest reply cites a fresh Arnason paper that sequences more lungfish in an effort to increase the Sarcopterygian diversity and thereby remove the spurious long branch attractions in the older study. He also includes plenty of tetrapods to create a potential source of synapomorphies for the modern sarcopterygians, while taking advantage of the phylogenetically friendly characteristics of entire mitochondrial DNA genomes. He even tosses in a few nuclear genes! Yet the results, as predicted, support Dembski's ID model.

Let's review some complaints of the original Arnason study: 1) No tetrapods 2) Not enough lungfish/coelacanth species 3) Bad root, partly due to 1) [...] To address the first two points, let's examine fresh studies. After making the very adjustments that Martin et al. demand, our dear Aranson finds that the results don't improve

— Ghost of Paley

On the other hand, its nuclear DNA behaves better under the evolutionary whip. So should this organism be used? Depends on the researcher's presuppositions.

By the way, Bichirs and ropefish were the basal piscines, leading to the oddly creationist mantra: "a tetrapod is a tetrapod, and a fish, a fish".

As you know, scientists screen genes for their reliability as phylogenetic markers before using them in a particular analysis;

frequently used genes have already been certified kosher by Darwin.

Researchers also use several stategies to combat varied mutation rates, hidden paralogy, and other problems. To no avail.

To catch people up: my latest reply cites a fresh Arnason paper that sequences more lungfish in an effort to increase the Sarcopterygian diversity and thereby remove the spurious long branch attractions in the older study. He also includes plenty of tetrapods to create a potential source of synapomorphies for the modern sarcopterygians, while taking advantage of the phylogenetically friendly characteristics of entire mitochondrial DNA genomes.

Comment #77678 Posted by The Ghost of Paley on February 5, 2006 03:33 PM (e) Yet the results, as predicted, support Dembski's ID model.

http://www.dinofish.com/

martin wrote:
Without a time machine, you cannot correct #2. There are no more lungfishes or coelacanths.

Yet the results, as predicted, support Dembski's ID model.

BWE,

I meant in addition to the ones we've already got. I work on sarcopterygian fishes, I'm quite aware of what does/doesn't live today. ;)

Though I suspect GoP is a prankster playing the role of creationist-provocateur, I'll bite. Where can I find "Dembski's ID model" in sufficient detail to make any predictions at all?

— Russell

Watch the Wizard skewer the jello of natural selection to the wall of disproof, leaving independent creation as the only viable alternative.

So Paley that ID Model of Dembski's is the one where all of evolution and common descent is accepted, right? You should really catch up.

Dembski's blog is saying that all living things are descended from living parents and no one has seen a living thing created from a non living thing.

Are you saying you have seen a living thing created from nothing ?
Or are you saying, you imagine a living thing created from nothing ?

Dembski supports neither of those arguments.

Mind the mathematical thorns in this thicket and you'll emerge with a firm grasp of Darwinism's problems. Watch the Wizard skewer the jello of natural selection to the wall of disproof, leaving independent creation as the only viable alternative.

To appreciate the significance of the No Free Lunch Regress in this latter sense, consider the case of evolutionary biology. Evolutionary biology holds that various (stochastic) evolutionary mechanisms operating in nature facilitate the formation of biological structures and functions. These include preeminently the Darwinian mechanism of natural selection and random variation, but also others (e.g., genetic drift, lateral gene transfer, and symbiogenesis). There is a growing debate whether the mechanisms currently proposed by evolutionary biology are adequate to account for biological structures and functions (see, for example, Depew and Weber 1995, Behe 1996, and Dembski and Ruse 2004). Suppose they are. Suppose the evolutionary searches taking place in the biological world are highly effective assisted searches qua stochastic mechanisms that successfully locate biological structures and functions. Regardless, that success says nothing about whether stochastic mechanisms are in turn responsible for bringing about those assisted searches. Evolving biological systems invariably reside in larger environments that subsume the search space in which those systems evolve. Moreover, these larger environments are capable of dramatically changing the probabilities associated with evolution as occurring in those search spaces. Take an evolving protein or an evolving strand of DNA. The search spaces for these are quite simple, comprising sequences that at each position select respectively from either twenty amino acids or four nucleotide bases. But these search spaces embed in incredibly complex cellular contexts. And the cells that supply these contexts themselves reside in still higher-level environments. As a consequence, the uniform probability on the search space almost never characterizes the system's evolution. Rather, according to evolutionary biology, the larger environment bestows a nonuniform probability that brings the search (i.e., an assisted search) to a successful conclusion. This, in a nutshell, was Richard Dawkins's (1996) argument in Climbing Mount Improbable: biological structures that at first blush seem vastly improbable (i.e., with respect to the uniform probability, blind search, pure randomness, call it what you will) become quite probable once the appropriate evolutionary mechanisms are factored in to reset the probabilities. Even if we accept the full efficacy of evolutionary mechanisms to evolve biological structures and functions, the challenge that displacement poses to evolutionary biology still stands. A larger environment bestows a nonuniform probability qua assisted search. Fine. Presumably this nonuniform probability, which is defined over the search space in question, splinters off from richer probabilistic structures defined over the larger environment. We can, for instance, imagine the search space being embedded in the larger environment, and such richer probabilistic structures inducing a nonuniform probability (qua assisted search) on this search space, perhaps by conditioning on a subspace or by factorizing product space. But, if the larger environment is capable of inducing such probabilities, what exactly are the structures of the larger environment that endow it with this capacity? Are any canonical probabilities defined over this larger environment (e.g., a uniform probability)? Do any of these higherlevel probabilities induce the nonuniform probability that characterizes effective search of the original search space? What stochastic mechanisms might induce such higher-level probabilities? For any interesting instances of biological evolution, we don't know the answer to these questions. But suppose we could answer these questions. As soon as we could, the No Free Lunch Regress would kick in, applying to the larger environment once its probabilistic structure becomes evident. And so, this probabilistic structure would itself require explanation in terms of stochastic mechanisms. On the other hand, lacking answers to these questions, we lack a stochastic mechanism to explain the nonuniform probabilities (and corresponding assisted searches) that the larger environment is supposed to induce and that makes effective search of the original space possible. In either case, the No Free Lunch Regress blocks our attempts to account for assisted searches in terms of stochastic mechanisms. Evolutionary biologists at this point sometimes object that evolutionary mechanisms like Darwinian natural selection are indeed a free lunch because they are so simple, generating, as Richard Dawkins (1987: 316) puts it, biological complexity out of "primeval simplicity." But ascribing simplicity to these mechanisms betrays wishful thinking. The information that assisted searches bring to otherwise blind searches is measurable and substantial, and discloses an underlying complexity (see section 4). Just because it's possible to describe the mechanism that assists a search in simple terms does not mean that the mechanism, as actually operating in nature and subject to countless contingencies (Michael Polanyi called them boundary conditions), is in fact simple. A final question therefore presents itself, namely, Is it even reasonable, whether in biology or elsewhere, to think that the assisted searches that successfully locate small targets in large spaces should be conceived as purely the result of stochastic mechanisms? What if, additionally, they inevitably result from a form of intelligence that is not reducible to stochastic mechanisms---a form of intelligence that transcends chance and necessity? The No Free Lunch Regress, by demonstrating the incompleteness of stochastic mechanisms to explain assisted searches, fundamentally challenges the materialist dogma that reduces all intelligence to chance and necessity.

— The Wizard

The Wizard wrote...

The Wizards wank
Blah blah all evolution is conjectural blah bla deny deny blah blah de Wizard don't like it blah blah couldn't happen because my mommy said so blah blah ....fundamentally challenges the materialist dogma that reduces all intelligence to chance and necessity.

7 out of 10 for ***sophism.
2 out of 10 for Sesquipedalian Obscurantism.: not nearly up to Denbski's OR even Sal's level.

Well Paley now we are getting somewhere.
If necessity is the mother of invention and hence intelligence and it came about by incremental change then why are you and wizard so damn dim ? You do realize sophism is just mental masturbation don't you ? Have you any idea what circular reasoning IS ?

Oh that's right you don't know what IS is.

***
The essential claim of sophistry is that the actual logical validity of an argument is irrelevant (if not non-existent); it is only the ruling of the audience which ultimately determine whether a conclusion is considered "true" or not. By appealing to the prejudices and emotions of the judges, one can garner favorable treatment for one's side of the argument and cause a factually false position to be ruled true.

The philosophical Sophist goes one step beyond that and claims that since it was traditionally accepted that the position ruled valid by the judges was literally true, any position ruled true by the judges must be considered literally true, even if it was arrived at by naked pandering to the judges' prejudices --- or even by bribery.

There is no hope for you Paley have you considered a career in ignorance? You have all the prerequisites.

7 out of 10 for ***sophism. 2 out of 10 for Sesquipedalian Obscurantism.: not nearly up to Denbski's OR even Sal's level.

— k.e.

Two questions: (1) why the family level? What is there in that "argument" that would permit X amount of evolution, but not 2X?

— Russell

Finally, yes, they did include tetrapods. Certainly it does not acheive a result more consonant with the accepted gnathostome phylogeny. I will concede that the exclusion of tetrapods was not solely causing the strange result. But I never claimed that to be the sole cause of the problem, but a serious technical problem and one that invalidated the analysis for the purpose of our discussion.

— Martin Brazeau

The huge environmental search space that natural selection must navigate in order to implement complex changes on multipart cellular systems. And since family-level evolution requires these multipart changes, this process is nearly impossible.

— The Ghost of Paley

Intelligence acts by changing probabilities. Equivalently, intelligence acts by generating information. For instance, a slab of marble temporarily has a high probability of remaining unchanged. Then, without warning, Michelangelo decides to sculpt David, and the probability of that marble slab taking on a new form (i.e., receiving new information) now changes dramatically.

The prototypical example of an assisted search is an Easter egg hunt in which instead of saying "yes" or "no" for each possible place where an egg might be hidden, one says "warmer" or "colder" depending on whether the distance to an egg is narrowing or widening."

Since assisted search is supposed to augment the information inherent in blind search, the information function associated with an assisted search needs to contain strictly more information than is contained in the indicator function of the corresponding blind search. This strict increase in information can be characterized as follows: An information function j´strictly augments the information in an indicator function j associated with a target T provided there is a function ϕ from Λ to {0, 1} such that ϕ ◦ j´= j and for any such ϕ there is no function ψ from {0, 1} back to Λ such that ψ ◦ ϕ ◦ j´= j

Re [Dembski's] "Intelligence acts by changing probabilities. Equivalently, intelligence acts by generating information."

Really? Maybe his acts differently than mine, but my intelligence acts by causing my body parts to move in ways that produce the desire results, such as typing this sentence into the computer.

Henry

This still leaves a question I don't think I'll ever be able to answer: To someone whose faith provides a one-size-fits-all answer to every possible question, WHY BOTHER trying to poke holes in competing explanations? Insecurity? Gamesmanship?

— Flint

Dembski seems to have missed that there is a 100% probability of things ending up as they are.

I'm actually not sure if they really did treat tetrapods properly in this later paper either. Before they worked on the data set for all taxa, they first found trees for tetrapods alone and for the paraphyletic "fish" alone; if they used those results to constrain their search on the total data set, wouldn't that tend to bias their results toward that of the earlier study in which they ignored tetrapods completely? I can't tell whether they did use those results, though. They describe the final searches on the combined data set as "exhaustive," but it seems to me a literal search of every possible tree topology and character state change placement is probably computationally out of the question here. More likely, I suspect, they swapped around large branches while keeping the internal topologies of those branches unchanged. What do you think?

— Anton Mates

The huge environmental search space that natural selection must navigate in order to implement complex changes on multipart cellular systems. And since family-level evolution requires these multipart changes, this process is nearly impossible.

— ghost

The huge environmental search space that natural selection must navigate in order to implement complex changes on multipart cellular systems. And since family-level evolution requires these multipart changes, this process is nearly impossible.

Martin Brazeau:

You shouldn't waste your time with GoP. Take a page from his book:

"Give not that which is holy unto the dogs, neither cast ye your pearls before swine, lest they trample them under their feet, and turn again and rend you." Matthew 7:6

Another look at intelligent design vs. evolution: In God's book, the Bible, he says we can know him by his creation. At least 80 percent of scientists now believe in intelligent design, and some don't even know the designer.

BWE ....and the Design Institute says ID has nothing to do with G-D ...oh no 80% of Op Ed writers with the intelligence of a bee say so, so they must be right.

Well said Wislu Plethora;
Sucky science and sucky spirituality.
2 Things you won't get from Creationists
The sermon on the mount, the plain and simple version of Christianity and plain common sense thought.
The bit they do get, and take to extreme, is the bit about 'thou shalt not insult' kind of fits right in with the Post Modernist concept of equal time for stupid ideas, not quite the same thing, but they don't consider their ideas stupid.
Pity young JC didn't have a sense of humor, but then when your on a radical and dangerous mission and so thin, there is little time for humor.

The dead hand of the irrational literalistic objectivist "The one true word of God^TM" and ass backwards thinking.

Anton you point out Creationists want science

Indeed at second glance the creationists ARE rational thinkers and more like "scientists" than we would care to admit. They realize that faith alone will never produce a live divine entity, that's irrational right? And just what did their savior say ? Faith is just too subjective for them, they can't compare it to any material thing. Faith MUST be material OR the language/law/consensus for describing reality must be changed to remove any firm significance for material nature itself. That is why they love Post Modernism.
Strange really, with the reliance feminism and pro choice had on PoMo one would think Creationism, a largely right wing political movement, would look askance at PoMo, but when one realizes that the fathers of PoMo were French Nazi collaborators who spent the rest of their lives rejecting any firm grip on reality (read conscience), it makes sense.
It used to be that 20% of the French believed in god and 60% in hell "la vie est merde", the English translators completely missed the cultural context.
Sociologists like Steve Fuller, the token lefty riding shotgun for the Design Institute and Howard Ahmanson, gets a HEAP of free publicity to sell lots of books.
Fuller has come up with something even more ridiculous to get his moment of glory, some nonsense about denying sophism explained by....wait for it....pure sophism and science reality envy, stupidity never dies..... it just rents a new tuxedo.

Orwell would not have been the least bit surprised to see his Newspeak aped so perfectly by this current crop of obscurantists particularly the press who have no more idea than Paley and his fellow travelers, indulging in "Newspeak C Vocabulary" as though it were some fantasy foreign language from another planet, where not one thing is known about it in their own minds, even the color of the sky.

To them Myths are a LIE they just don't,can't,won't do metaphor or even nuance. They deny it but they DO have metaphor, consider the the various wars on "nouns".
And those Mythic tales echo the word wars and the people involved in them right to the present day not as a predictor of what will happen but as an insight into the mind of the actors. The Hydra is a perfect example complete with bad breath (that's propaganda kids). Note that Hercules had to cut down an entire forest to cauterize the stump of each head he chopped off but even then the damn thing only retreated to the swamp (the collective human subconsciousness) to terrorize the villagers again at some future time. Those Greeks new a thing or two about human nature.

There is no subjective comparison that they are able to make between belief systems that carries any value since their life experience is so monotonous and devoid of historical meaning. I almost feel sorry for them.

The total beauty of the Fundamentalist mind set is to completely change reality in the believers minds and fix it on "The one true word of God^TM" they can then reject ANY authority that conflicts with their reality (and if you think I kid about their reality, get your heads around this, what they believe IS as REAL to them, as an actual ice-cream is to a kid)

To make their system work they MUST create a separate entity that removes their immediate responsibility to their own conscience and more importantly makes everyone else subservient to that divine entity. And guess what? THEY get to decide what that divine entity likes and dislikes, so all the dark little prejudices, irrational fears,pride,desire for power,desire for wealth and hubris get projected. The Greeks and Egytians had all this stuff mapped out and described in their underworlds, where do you think Dante got all his ideas from ?. He even recognised his friends in his version.

GoP and his little bunch each line up to take their place under Shiva's foot AND they love it.

Each time they get rejected they repeat the age old martyrs glory, their humiliation CONFIRMS the Grand Old Designer is really there WAITING for them to give them each their eternal holiday at Club Med or 72 virgins, GUARANTEED rest and glory in the afterlife.

No different to that Easter Islander, who listening to the retelling of the one true word of whatever, hacked the last tree down to move his statue so it could watch over and ensure the return of his ancestors. A worldview completely unmoderated by any firm grasp of the consequences in the here and now because introspection is postponed, at least until your neighbor wants to eat you, because there is no wood left to make canoes for fishing.
Dembski postponed his own judgment at Dover because he knew what the outcome would be. He knows, as any sane person does, that there is nothing "over there" to recon with guilt, how can you be guilty after you are dead ? Better to have an 'imaginary after life friend' who as the good book says will redeem all those nasty doubts and inject...ah wash ...er dip you in rapture after all the dirty work is done down here.
Those guys give rapture a bad name, its something you 'get' in the here and now.
The same old side show alley, roll up roll up get your rapture here, at gun point if necessary.

Oh by the way (giggle) what is the collective noun for sophists

A pride of sophists

We have to try to base our estimates of phylogeny on careful, analytical investigation. It must be done with a healthy respect for the potential biases and assumptions (as opposed to the "More! Bigger! Better! Bust!" approach of Ghost of Paley). Then, we are most justified in selecting our trees as phylogenies when they converge on independent data sets and are well-supported. Even this, however, must be provisional.

— Martin

Without a time machine, you cannot correct #2. There are no more lungfishes or coelacanths. Three and one, that's all we've got. Polypterids (bichirs and ropefish) are represented by similarly few lineages. There's nothing you can do to add more of these, thus these long lineages will predictably have molecular sequences that are problematic for phylogenetic estimates.

— Martin

As for #3, I never said anything about a bad root. You can't have a bad root due to ingroup taxa. You're spouting nonsense and once again making up arguments which I have not made. What I have said is that the lack of ingroup resolution is consonant with a series of plesions that do not properly polarize the major ingroup clade of coelacanths, teleosts, and chondrichthyans.

Could you please define "the evolutionary whip"?

Your standard applies only if we're dealing with the optimal and most complete data set. Unfortunately, we're not dealing with such data. Even if I discount any assumptions about evolution, more taxa are extinct than are alive today. Our current sample is highly non-random. The current complement of organisms is not a random sample of all that has ever lived. It is, in fact, a heavily biased sample where selection and regional effects are involved. Then, if we even admit that evolution has happened, then we're dealing with extremely long independent evolutionary histories, variation in rates, heterogeneity, saturation, and convergence. Then there are methodological and computational biases that I needn't name here. The methodologies are coarse and, by necessity, naïve in their approach to reconstructing phylogeny. Because they conflict here and there is hardly a strong argument. You can always find one tree that disagrees with another in the placement of anywhere from one to many taxa. But to use that as evidence of disasterous conflict requires that you ignore all other places where the topologies, in fact, agree.

Sigh.

Pasty-Face, do the concepts "baby" and "bathwater" have any meaning in Bizarro World?

By now, the answer's pretty obvious to both of us (though only one of us will admit it):

That every molecular tree can't yet be confidently and consistently connected in deep time, due to the kinds of technical and evidential difficulties that Martin Brazeau has kindly explained and that EctoSpasm keeps trying to mangle and confuse, does not mean that hundreds and thousands of valid trees cannot be constructed, nor does it mean that these trees cannot be of immense assistance to us in working out any number of interesting issues in evolutionary biology.

Yeesh.

Ghost's line of argument sounds eerily familiar: It's imperfect. Therefore it's wrong. Therefore something devoid of any evidence whatsoever must be right. Which just happens to be what my religion teaches.

Still, I gotta admit that the extent Ghost is willing to go to convince himself that his evidence-free foregone conclusions are superior to 99.9% of the evidence is impressive. He digs awful hard to find a speck of verisimilitude here and there, tuning out all else, and living in a world barren of understanding, but finding it better than a world rich in *wrong* understanding. A shame some people can't sue their parents for crippling them for life.

Anton Mates: In the future, please don't confuse analogies with math. The Wizard is not interested in hand-waving.

— The Ghost of Paley

Without a time machine, you cannot correct #2. There are no more lungfishes or coelacanths. Three and one, that's all we've got. Polypterids (bichirs and ropefish) are represented by similarly few lineages. There's nothing you can do to add more of these, thus these long lineages will predictably have molecular sequences that are problematic for phylogenetic estimates.

— Martin

An untestable assertion, of course.

Contra Theobald, we must discard molecular trees as potential evidences for common descent until we work out the kinks. Too bad - it looked promising there for a while....

The main problem seems to be mtDNA's inability to derive the correct phylogeny.

— Ghost of Paley

mitochondral mutation rates can be up to an order of magnitude higher than the average nuclear rate and also exhibit base pair bias, both factors contributing to saturation

— Ghost of Paley also

Okay, that's it. Seriously. I'm done. Sorry, but I really don't have time for this any more.

— Martin