Barry seems to have noticed TSZ again, and so I will take this opportunity of inviting him over here, where he can post freely, and will not be banned unless he posts porn or malware or outs someone, which I expect he can manage not to do.
And he responds to my post, Lawyers and Scientists. He does so in two parts, so I will devote two posts to them. Here is my response to his first part. Barry writes:
PART 1
First Liddle writes that I have
. . . confused the assumption of common descent with the conclusion of common descent, and thus detected circular reasoning where there is none.
Where did I do such a thing? Boiling that paragraph down I made the following claims:
- Common descent is not necessarily false.
- But Cladistics does not establish common descent one way or the other.
- Instead, cladograms are constructed ASSUMING common descent.
- It is circular reasoning to conclude that a technique establishes that which it assumes in the first place.
- Therefore, anyone who says that cladistics establishes the fact of common descent has used faulty reasoning and is mistaken.
- There are in fact people who make that mistake.
To establish beyond doubt point 6, Glen Davidson kindly jumps into Liddle’s own combox with this:
Barry: “This is not to say that common descent is necessarily false; only cladistics does not establish the matter one way or the other.”
Glen: “Of course it does. What a ridiculously ignorant dweeb.”
All six assertions seem to me to be on solid ground. Not only are they true, they are not even controversial. But for Liddle’s charge to be correct, at least one of the points I made must be false. OK Liddle, which of the six totally non-controversial points I have made do you disagree with? If the answer is “none,” then the only gracious thing to do is to withdraw your claim.
The short answer is that I disagree with 2-6, for the reasons I gave in my first post: the answer lies in null hypothesis testing. Far from “assuming a tree”, both linear correlations and tree distributions are tested by FITTING a slope/tree, and testing whether the best fit is a better fit than would be expected under the NULL of no linear relationship/no underlying tree structure. If, having fitted the slope/tree, the fit is no better than would be expected under the null of no linear relationship/no underlying nested hierarchy, then you RETAIN THE NULL. If it is better, i.e. if a fit as good as that observed is UNLIKELY under the null, you reject the null and consider your hypothesis (linear fit; common descent pattern) supported. Of course there could reasons other than common descent that could explain the tree – but the tree can be established as an OBSERVATION to be EXPLAINED. Which Linnaeus did before Darwin. And it was that clear tree that Darwin sought to explain by, firstly, Common Descent, and, secondly, by a mechanism that would explain adaptive change-over-time.
If Barry cannot understand that testing a NULL HYPOTHESIS is the OPPOSITE of assuming that your model is true, then perhaps he could shoot an email to the former owner of his site.
It is of course true that null hypothesis testing is counter-intuitive and doesn’t do what many of its practitioners think it does, but it’s still an excellent workhorse, and what’s more, is the beating heart of ID’s very own eleP(T|H)ant.
[My response to the second part will have to wait – I have some null hypotheses to test first….]
ETA: Looks like this response deals with Pt II at as well.
The equation for CSI, for one.
GlenDavidson,
OK, so it’s an openly shameless rehash. I’d do an impression of Lloyd Bentsen cutting Skippy Quayle down to size, except I don’t know Phillip Johnson, and he probably wouldn’t be a friend of mine if I did. So here’s a NOVA interview with Johnson after the Dover trial. I defy Arrington to identify a post that puts him anywhere close to Johnson (who’d suffered several strokes) in eloquence, acuteness, and astuteness. It was Johnson, not Michael Behe, who came up with the idea of pushing irreducible complexity (known long before as interlocking complexity) as evidence of design. I cannot imagine that Arrington was even the source of FIASCO.
When Barry is finally overwhelmed by conniptions, UD will no doubt assemble its Little Wedgie to produce a festschrift, Liddle’s Nemesis: Barry Arrington and the Intelligent Design Blog. There will be Winnie the Pooh Bear to edit, and former senator Michelle Bachmann to supply a foreword.
The downside for ID trying to pretend that the genetic code is like a language, so supposedly designed, is, of course, that languages are not designed but evolve. A linguistic cladogram. Intelligence is involved in language evolution, naturally, and some design may occur, but the constraints of the past typically hold sway in languages to the degree that cladistics actually works with languages.
Interestingly, there are artifacts that do (or at least could) fit properly into cladograms, unsurprisingly, the ones that involve derivative copying. I know, I know, I bring up an exception that the IDiots will run with if they see it, ignoring the exceptional nature of such cases. But so what, they’re always ranting some stupidity anyhow:
Here’s where cladistic analysis was used to find out if Thai Buddha statues in fact evolved.
How well or poorly done that was isn’t the point, it’s how cladistics is used to test whether or not things have evolved with enough constraint to yield a proper cladogram.
So yes, of course cladistics can be and is used to figure out whether something has evolved. We all know that you can stick artifacts onto a tree regardless of its having evolved–it’s the nature of space–but cladistics only works properly for evolved things, like organisms, languages, and maybe Thai Buddhas.
Glen Davidson
I know three equations for CSI. I’m not sure what you mean who you say that Dembski doesn’t know what the terms mean.
I appreciate the information. Now, can someone tell me what the point of cladistics is? I mean, other than arranging biological data into groups and patterns. Are those patterns then used for something?
I suppose I should make a halfway substantive remark before ducking out. ERVs (endogenous retroviruses, not Abbie Smiths) are much easier for most people to understand than phylogenetic inference. You don’t have to generate a tree to give clear evidence of common descent.
I usually check Wikipedia before demanding an explanation.
Well, I don’t think he knows (or perhaps he does but doesn’t say) how you would set about computing p(T|H) for a given pattern. When I say he doesn’t know what they mean, I mean he doesn’t seem to have any idea about how you populate his equations with real-world measurements.
I don’t know this null stuff but I do know that these trees do so presume what they are testing. They do not allow for other option(s). You can
‘t say it fits a tree for these connections if it better fits connections by other options.
A creator with a common blueprint, in major concepts for bodies, would also have a tree outlook even though there was no biological descent relationship.
Then special cases of minor mechanism issues.
So Yes we look like primates but no reason to say we are related because of looks, just a option hunch, but yes we are related amongst races, so called, as a special mechanism.
Evolutionism simplistically assumes evolution is true when it makes its trees to prove its true.
Sure they do.
Creationists got a good point here.
What information do you appreciate? I find it hard to consider your questions here other than disingenuous. You should know that cladistics (by which I think you may mean phylogenetic analysis, for which “cladistics” is not used as a synonym) is used to determine the phylogenetic relationships among species. Or are you asking what you can do if you know phylogenetic relationships? To answer that latter question, there’s a whole field known as comparative biology.
No, I don’t think anyone could explain it to you, no.
The reason you think that “these trees do so presume what they are testing” is that you “don’t know this null stuff”.
They are used to discover the patterns and groupings in biological data, not to “arrange biological data into groups and patterns”.
Elizabeth,
Yeah, well I don’t understand what you’re talking about but I know you’re wrong!
Imagine the tree as a template. How good a tree is a tree? A bush? A Stick? A circle?
I don’t know you well enough to tell if that was a joke. I hope it was.
A bush is just a tree with short branches. A stick is, presumably, a tree with most branches very short and a few very long. Both these are tested by the standard methods in use. A circle isn’t something one generally tests, but data would have to be exceedingly peculiar to fit one.
John Harshman,
Hi John. When I make the universal sarcasm symbol I’ll be Rich(er). I understand phylogenetics a little – we use oglomerative hierarchical clustering at work, 😉
Richardthughes,
Well, that’s a relief.
It’s good to ask whether Dembski knows what he means. I seem to be trolling, and pulling you off-topic. You aren’t aiming for bulletproof-ness with me. But it’s important to distinguish calculation and measurement carefully, because he is careless about stuff like that. CSI is supposed to be out there in the real world. It’s supposed to be something that scientists can measure objectively, sort of like specific gravity. Pun intended. I don’t know what it means to measure the chance of a specified event physically. It seems that Dembski took probability “measure” in math, an analogy, literally. Was he really so stupid? We have all the evidence we could ever want that he is not at all stupid. The default position, looking at his claims, should be that they are nefarious. He went out of his way to earn it. Phillip Johnson recruited Dembski to his “Wedge of Truth” 25 years ago. There comes a point… Scratch. There came a point long ago when it was stupid to give Dembski the benefit of the doubt. He sought to win debates and court cases by any and all means. He was a Psychopath for God.
I am not so sure that anyone should respond seriously to Dembski. This thread has expert scientists responding seriously to a paltry Johnson wannabe on a virtual soapbox. He’s to be derided as ethically dubious political hack, playing accounting games in the back room when he’s not propagandizing in the open. He should not be empowered. Pretending that he is something other than what he is, in order to have fun intellectualistic arguments, is ethically dubious. What’s at stake, here in the U.S., is the education of schoolchildren. Beyond that, constitutional separation of church and state. Banny believes that government should be kept out of religion, not that religion should be kept out of government. Majoritarian religious views rule. You may not feel that you need to take this stuff into consideration. But I wish that you and the other Europeans would.
Back to probability. The probabilities arising in tree inference do not reflect indeterminacy in the real world. They are not chances. Not what Dembski means by “chance.” The talk about IDers misunderstanding probability is too loose for me.
And now Erasmus Wiffball scurries back under the baseboard, to scrub his carapace, and scuttle on to his place at the Errington family table. Cousin Banny doesn’t like me much. But we are blood kin.
Why should I know that? I don’t know anything about cladistics or phylogeny, other than looking those terms up.
Okay. I looked up comparative biology, and that’s a pretty wide-ranging assortment of applications. I guess my question is a lot more narrow or specific.
You can catalogue individual taxa all day and night without even attempting to arrange them into relationships. IOW, you’d have a large amount of data where this set is similar to that set, or even identical. If we threw out common descent, you’d still have the data, and I’m sure that data would still be useful in a lot of ways.
But, in what specific way is that data only useful if it is arranged into patterns based on common descent? Apparently, the use of such CD-specific patterns is widespread in biology (I gather, from the definition of “comparative biology”), but I’d like to understand how the CD pattern arrangement is specifically useful, even if just in one particular example.
I’m not arguing against common descent here. I have no skin in that debate. I’m really just asking a basic question. I don’t understand what such arrangements of CD relationship patterns brings to biology that data simply arranged by similarities doesn’t. Why is conceptualizing data sets of taxa in terms of greater or lesser similarity not enough to get the same work done as conceptualizing them as parts of a universal branching tree of life? Is there a specific example anyone can give me?
My apologies if the question sounds stupid. I’m not a scientist. I’m just trying to understand, at least in one specific case, what value “cladistics” would bring to any specific biological issue other than simply arranging patterns of lineage for the sake of arranging such patterns.
It’s emergence, William. When Linnaeus published his Systema Naturae he arranged animals and plants by morphological similarity. The hierarchy indicating common descent emerges from the pattern of morphology, the evidence.
The pattern emerges from the data, it is not imposed on the data. Hopefully many things will go “click” now you are starting to realise this. But I doubt it.
Isn’t that a just value judgement? I don’t see any compelling reason to say that the hierarchy indicates common decent.
You can also arrange geometric shapes into a nested hierarchical schema. But does that mean that squares and pentagons are descended from triangles.
Or you can arrange integers in the same manner but that does not mean that the even numbers are descended from two and the odd from one.
Like WJM I’m not arguing against common descent here.
This seems to be a deeper question about the subjective nature of interpreting patterns. We see a pattern but it’s meaning is supplied by us according to our own subjective experience and desires.
peace
Same with elements, we could throw out the periodic table of elements and we’d still use iron like we do, carbon like we do, etc. And still, potassium and sodium react similarly, wouldn’t it be nice to know why? So eventually Mendeleev came up with the periodic table, and discoveries flowed from that. We could hang the elements from a tree, but that would yield an arbitrary classification. The table groups similar elements in useful ways. I have never seen rubidium that I know of, but I still have a pretty good idea of what kind of metal it is (very soft and very reactive) just because it is in the first column under potassium and above cesium.
Likewise, we could know a lot about German and Dutch, but if we recognize that they both descend from a common Germanic language we can understand the similarities and differences between the two in a comprehensible and causal manner–and thus know more. A lot of people won’t care, but for linguists a hierarchical “tree” reflecting common descent of the two languages is the only non-arbitrary way of categorizing them. Horizontal transfers of French words into German become obvious by comparing “related” and “non-related” languages, like German and Dutch as related, and French and Dutch as non-related, or actually, as rather distantly related.
Likewise, we have mammals exhibiting hair and mammary glands. Whale and bat–not a lot alike, but they both have hair and mammary glands. We could classify whale as big and wet, and the bat as small, flying, and dry, but that doesn’t get us anywhere in understanding why they both have hair and mammary glands. Nor does it tell us why testing drugs on mice typically gives us a pretty good idea of what the drugs might do in humans. Putting the two together, and recognizing the non-arbitrariness of shared hair and mammary glands (common descent) both classifies these usefully and explains similarities and differences. Putting zebrafish, humans, and mice into the Chordate phylum classifies all of these according to similarities, but because zebrafish are not mammals we understand that tests on them are likely to be less relevant to human physiology than are tests on mice.
HIV testing was done on chimpanzees. If we classified animals as quite hairy and rather hairless, we might have tested HIV in blind mole rats, at least until we realized that they can’t contract AIDS. But why can’t they? They’re rather distantly related to us, while chimps are closely related.
Just “similarity” is arbitrary, both with respect to the characteristic as well as across the various types of organisms. What do we pick, size, color, shape? Why are snakes classified differently from legless lizards? Because snakes share a similar body type, lizards another. Snakes have long bodies and short tails, legless lizards have rather shorter bodies (per size) and long tails (how do you know? From where the cloaca is–the tail is post-cloacal). We could classify them together as similar, but what about the differences?
Such a non-arbitrary classification tells you what the possibilities are within the various taxa. You will not find mammary glands among the birds, nor feathers on bats, thus you categorize them differently. Cladistics and similar phylogeny-based taxonomy methods are used to reduce arbitrariness, both in our understanding and according to what is expected in the various clades (related to taxa).
Glen Davidson
The legitimate claim is not that hierarchy indicates common descent, it is that certain otherwise-arbitrary traits that do actually appear according to a “nested hierarchy” (we construct it, but not arbitrarily–it is a found non-arbitrary pattern) that indicates common descent.
Of course you can, but the relationships are purely geometric, and really your nested hierarchy is rather arbitrary. A table might work better.
Do mammary glands and hair lump together arbitrarily? Or notochords and vertebrae? No? We find patterns of traits appearing together, and these form a non-forced nested hierarchy that, if non-arbitrary, has only one explanation–common descent.
Or you could arrange them quite differently, no matter. How else could you arrange life except as a nested hierarchy–at least if you’re striving for meaning?
The meaning is supplied to us through empiricism–experience. We see a word that is similar in Greek and in Latin–are we to think that this is entirely arbitrary? If it’s just one word, who knows? Maybe it is just a strange coincidence. But hey, there are a lot of cognates between the two, pater for “father” in Latin, and pater (with some added marks) for “father” in Greek, gignosko for “knowing” in Greek and nosco for “knowing” in Latin, etc., and we don’t get that sort of similarity between, say, Greek and Hebrew. Well, maybe they borrowed, after all, “camel” in English seems to derive from some Semitic language (if through Latin). Well, no, the differences are rather great, and Greeks and Romans lived fairly far apart (there may have been some borrowings early on, even so), plus newer words are not so much related (can be related through similar derivation, but could be borrowed from other languages or derived from very different words). What can it be, then?
Like comes from like, for a rule of thumb, so the answer seems to be that Greek and Latin came from a “common ancestor.” Otherwise, the similarities are just arbitrary. But hold on, Latin and Old English also have a number of cognates, albeit fewer, so did Greek, Latin, and English all just come from one language? Weird thing, though, Old English and German have a lot more cognates than to Old English and Latin, so they seem more closely related. So in the end, to make sense of it all, we take it that German and Old English separated from each other later than did Latin and “the Germanic languages,” and we have to put the languages into a nested hierarchy if we want to reflect their various relatedness and unrelatedness. You could classify them differently, as a periodic table or some such thing, it’s just that the periodic table wouldn’t group together similarities meaningfully like a nested hierarchy will.
Of course one may classify things arbitrarily, but we don’t classify elements as nested hierarchies because it makes no sense of elemental similarities and differences, and we do classify life according to nested hierarchies because those do make sense of organismal similarities and differences.
Glen Davidson
William J. Murray,
You ask about the uses of phylogenies. The short answer is that they’re necessary any time we want to investigate evolution over a long period. The long answer would involve dozens of methods. I’ll try a medium answer.
For example, suppose you want to know if an increase in characteristic X is evolutionarily correlated with an increase in character Y. Just plotting X against Y for a bunch of species fails to take into account any correlations due purely to closeness of relationship, and you could easily get a significant result purely for that reason. Taking phylogeny into account removes that problem. There are many methods, but probably the most popular was introduced by Joe Felsenstein in 1985 (http://www.jstor.org/stable/2461605?seq=1#page_scan_tab_contents). Here’s the first application of the method I ran into by googling: http://www.amjbot.org/content/86/9/1272.short.
Or suppose you’re trying to find out what, genetically, makes us human; you would want to know what mutations have become fixed in the human lineage specifically, and separate that from things that happened on various prior branches of the tree. To do that you would have to compare the genomes of various apes and monkeys in the context of a phylogenetic tree. Without a tree all you see would be a collection of differences from various species without any clue as to whether the differences between, say, you and an orangutan happened because of changes in the human lineage or the orangutan lineage.
There are hundreds of other possible applications, but the short answer summarizes them all: if you want to investigate evolution over long periods, you need a tree.
The real issue not being faced by creationists is that nested hierarchies and tree structures are just one of dozens of independent lines of evidence for common descent.
Now if the similarities and differences — at the molecular level — suggested a different pattern of descent from structural and fossil evidence, that would be a problem.
But if we enter the world of postdiction, then common descent postdicts the pattern found in genomes, the similarities and differences in both functional and broken genes.
Common descent requires a certain kind of model of inheritance, and given all the other evidences for common descent, it is not unsound to test the fit of actual data to models of inheritance.
This is a claim not just about trees but about any pattern whatsoever. If we take it literally, then we can know nothing about anything, as all our knowledge is inference from patterns of data. I urge you to reject epistemic nihilism. And the way to do that is to realize that some inferences are valid. One such inference is that nested hierarchy is an inevitable result of evolution on a branching tree, and of nothing else that we can think of. You can impose a hierarchy on any data, but here we’re talking about data whose structure is hierarchical and on which a tree doesn’t have to be imposed. If you can think of an explanation other than phylogeny for such data, you would be the first.
Common descent doesn’t produce nested hierarchies.
LoL! Linnaean classification has nothing to do with common descent
That’s an interesting claim. How do you justify it? If there’s a tree of descent, and changes happen at various points on that tree, how can it avoid producing a nested hierarchy?
That’s an interesting claim. How do you justify it? If there’s a tree of descent, and changes happen at various points on that tree, how can it avoid producing a nested hierarchy?
Well, this is odd. I seem to be having trouble posting, since the site thinks I’ve already posted this.
John Harshman,
And just adding that little note about trouble posting has allowed me to post. What?
Transitional forms would ruin any attempt at forming a nested hierarchy. Wagner goes over that in “Arrival or the Fittest”. Darwin goes over that in “On the Origins of Species”
A nested hierarchy has specific entailments- educate yourself to what those are
Remember you are on a short leash, Joe/Frankie. Everyone else, remember Brer Rabbit and the tar baby! 😉
Is it my fault that you guys are ignorant of nested hierarchies?
Joe’s response is here in guano.
If you can’t support that assertion, you should withdraw it.
What assertion? I can support the claim that common descent doesn’t produce nested hierarchies.
Frankie,
Indeed. Before dismissing Joe as just another internet crank I think you should all know that he is a self-declared “ID Leader” (uncorroborated by anyone else) and is “leading” in “Intelligent Design Evolution”. I presume his works are forthcoming.
How goes the research, Joe? As a leader I suppose you’re conducting lots or maybe managing a team? We could start a thread for your results.
He’s probably still occupied in his search for the largest integer. He’s such a renaissance man and polymath that Joe. 🙂
No you can’t but that is not what I meant. I am refering to this:
You need to demonstrate that ignorance or withdraw the allegation.
Where was this stupendous scientific work published?
Can the system handle all of the data?
Better to use a post-it note.
Glen Davidson
I’ve reluctantly engaged the moderation filter for further comments from Frankie. I’ll release any comments directly to the appropriate thread as soon as I spot them.
@ Frankie
Check your PMs
Check you PMS indeed!
The site already has a few glitches. The “duplicate comment” glitch is new and will be added to the list of glitches we need to fix.
You had best explain, since I, who after all am only a simple country biologist, appear not to know.
Glen Davidson said:
This is what I was asking for – a specific instance where cladistics brought something to the table other than just providing tables, diagrams and classifications or historical lineage reconstructions, or even if it is those cladistics-provided structure are then used to do something practically useful in medicine/biology – like predict what animal(s) we should be testing drugs on to see how they might affect human physiology.
However, I’m unaware of if it is because of cladistics that mice are so widely used in biological experimentation. Did cladistics predict that mice would be useful, or did the use of mice (and indeed, other animals like guinea pigs, dogs and monkeys) develop more due to similarity, availability, economics, etc? Do researchers use cladistics to choose what animals to experiment on in particular cases?
Can someone tell me of some other ways cladistics might be of practical use in biology/medicine?
It is currently being used by ecological biologists, zoologists, and agricultural scientists to try to determine why bees in particular are dying at such huge rates while other closely related insects are relatively unscathed.