Re: Part 1 (of 3): What are major aspects of evolutionary theory?
- From: John Harshman <jharshman.diespamdie@xxxxxxxxxxx>
- Date: Mon, 02 Jan 2006 06:09:01 GMT
anon1@xxxxxxx wrote:
>>I'm not talking in circles.
>
>
> I ask you how to root the tree, you answer by using directed links.
> I ask you how to decide which direction a link goes, you answer by
> using the already-rooted tree.
> I say that is talking in circles, you say it isn't.
That's because you have garbled the explanation. The "directed links"
are different from the "links" that are directed by using already-rooted
trees.
>>Perhaps I have not explained well enough.
>
> In a few areas you have explained well, but your median is crappy.
I will persevere.
>>But you have no reason to be smug in your belief that all the
>>scientists do everything wrong.
>
> I have no such belief. I merely claim that the way scientists refine
> their models of evolution, after already being 100% convinced the
> prevailing theory of evolution is true, is no good at all for the task
> of how to provide a concise summary of evidence for evolution.
Not true. You are claiming that the ways in which trees have been
rooted, historically and in the present, are invalid.
>>>In theory, it's impossible for a duplication to occur unless
>>>some form of living thing existed prior to the duplication. So it is
>>>*always* presumed that the true tree of life includes both the parallel
>>>post-dup branches and the single set of pre-dup branches.
>>
>>No, it's not presumed. It never has to be presumed because it's
>>irrelevant. Why bring it up at al
>>(Yes, it's true that such branches exist or must at one time have
>>existed. So what?)
>
> So you claim it's true but all the professional scientists deny it?
No. I'm claiming that talking about the pre-duplication condition is
useless and irrelevant to using dpulicated genes to root a tree.
> You seem to be agreeing with me, while disliking how I worded it.
> So how would you express the view that professional scientists take
> toward the model that life must have existed before the duplication?
> If somebody were to ask them whether they accept it as true, how would
> they answer?? Would they refuse to take a stand on the issue because
> it's not their job to express such an opinion?
No. They would ask you why you are bringing up irrelevancies. Perhaps
they would tell you not to be such a tool.
>>>>Classical systematists formed lots of groups based on distinctness.
>>>>Some of them are monophyletic and some are paraphyletic.
>>>
>>>Which puts us at square zero with no reliable data, at that point, as
>>>to which of those are true clades (monophyletic) and which aren't.
>>
>>No, it doesn't.
>
> How, in your opinion, does a mix of guesses, some true and some untrue,
> with no way to distinguish the true from the untrue, give us any valid
> starting point?
It does if all the guesses are independent and each has some reasonable
probability of being true, and if you can concatenate them. All these
are, in my opinion, correct.
>>But if it did, then we would have no way of rooting a tree and you,
>>by your own logic, should be a creationist.
>
> With nothing whatsoever to go on, except a bunch of ancient opinions
> half of which are wrong, indeed there'd be no way to favor evolution
> over creationism. Whenever I suggest some way to show evolution really
> did happen, you say no that evidence is invalid, and the only thing you
> have to replace it is circular arguments (use already-rooted tree to
> establish direction along a link and use that to root the tree), so per
> your criterion we have no evidence for evolution whatsoever.
No, that's your criterion. I'm merely asking you to face the
implications of your opinions.
>>If so, then everything we know about phylogeny is built on sand.
>
> No, everything we might have guessed about phylogeny prior to the first
> fossil datings was built on sand. Now we have fossil datings, which
> gives some starting point for phylogeny, but you deny that fossil
> datings are of any value.
That may be too strong a statement. They may be of some value. However,
the fact is that we do not use and have not used fossil datings to root
trees. So if that's the only valid method, we really will have to start
from scratch, and everything we think we know about phylogeny must be
redone completely. Perhaps you dispute this, and think that fossil dates
are commonly used to root trees. If so, perhaps you can show me a
scientific paper in which this method is used. I don't know of any, and
I'm pretty familiar with the literature.
>>For all we know, the root of the tree lies between Homo sapiens and
>>all other species, or even within Homo sapiens. Hey, why not?
>
> Fossil datings refute that hypothesis.
You may think so, but I don't know of any case in which this method has
been applied, even if it's considered valid.
>>Most groups are not in question, and nobody is looking.
>
> Per the purpose of this thread, you're completely wrong here.
> *every* philogeny is in question until we establish what evidence shows
> any evolution happened at all.
Only to you and creationists. To scientists, it's not in question. You
may have a tiny point in that every scientific conclusion is provisional
and open to question. However, some are better supported than others, to
the degree that nobody wastes their time questioning them. The monophyly
of living mammals relative to all other living groups is one of those,
for example.
> You've proposed the argument that mere unrooted-tree distribution of
> genotypes and/or characters is evidence for evolution. I showed a
> washout model where unrooted trees don't correspond to any sense of
> different times at different nodes. You didn't give any reason to
> reject my model, so my model stands as a refutation to your claim that
> every unrooted tree implies it has a root representing flow of time
> from that point forward.
My justification for rejecting your model is that it doesn't even
produce unrooted trees, and its connection to any biological phenomenon
is clear as mud.
>>I, however, am willing to believe that if all protists fall into one
>>group on an unrooted tree, and all metazoans fall into another, that
>>this means metazoans are monophyletic.
>
> Given an unrooted tree Protist--Metazoa, there are several possible
> interpretations:
> - Root within protist, evolved over time to make metazoa.
> - Root within metazoa, evolved over time to make protist.
> - Root exactly between metazoa and protist.
> - There's no root because there's no such thing as evolution.
> I'm looking for evidence against that last one. Do you know any?
Yes, but you won't accept it. Show me a process other than evolution
that produces a tree, rooted or unrooted. By this I mean a nested
hierarchy, not just a tree shape. If you don't think an unrooted tree
should be called a nested hierarchy, give it some other name. But I'm
talking about a set of compatible bipartitions of a set of objects. By
compatible I mean that each bipartition divides the set into two subsets
such that each subset has one of two relationships to each other subset:
either one of the subsets includes the other, or they are disjunct.
There are no partially overlapping subsets.
>>You aren't willing to make that choice, which means that for you,
>>Mammals are not a clade.
>
> That's not correct. Fossils of mammals don't exist prior to about 150
> million years ago. Other parts of the unrooted tree existed as long as
> 500 million years ago. Therefore the root, the UCA, if any, can't be
> within the mammals.
Whether or not this method would be valid, you must contend with the
fact that it's not the method that anyone actually has used. If mammals
are monophyletic, this has so far been untested in the scientific
literature (by your criterion). And that's true of every single group we
know of. So perhaps mammals are monophyletic for you, but in your
opinion this would have to be unknown to anyone but you.
>>we may have some clue as to which character states are likely to be
>>derived, because we know something about evolution.
>
> For the purpose of this thread, showing evolution isn't just a joke, it
> really happened, that's not a valid argument.
Not necessarily. One could be consistent by supposing that if evolution
happened, it would have happened in such and such a way. Therefore it's
not necessary to assume evolution actually happened, a priori, in order
to use this sort of method.
>>If, for example, we suppose that life becan in the water, for which
>>there seem reasonable arguments (even absent fossils, again),
>
> Without a single fossil, what evidence are you claiming?
> (Philosophical arguments not based on evidence are not allowed here.)
Physiological arguments, for the most part. Life on land requires all
manner of special mechanisms not needed for life in the water.
>>then there is a presumption that terrestrial groups are monophyletic;
>
> I am quite sure that terrestrial groups are *not* monophyletic.
> First of all, at least two different oceanic groups evolved toward
> living above the water line, so even at that point terrestrial groups
> are two different clades.
No, not that all terrestrial groups form a single clade, but that the
several terrestrial groups are individually monophyletic.
> Second, within at least one of those clades that got onto land, several
> sub-clades have returned to water more recently, so the clade isn't
> fully terrestrial nowadays, i.e. the terrestrial subset isn't a clade
> at all, it's only part of a clade.
All literally true but pointless. Even though whales are aquatic, they
don't prevent tetrapods from being monophyletic. Remember that we have
this unrooted tree on which tetrapods form a bipartition. The idea is
that this bipartition is a clade, not any quibbly idea about what has
fins and what has feet.
> I think you mis-worded something so badly as to turn some idea into a
> misspeak.
I think you miss the meanings amid the quibbles.
>>thus the root of life cannot lie in the midst of any terrestrial
>>group.
>
> Truthful conclusion, flaky argument getting there.
>
>
>>it's reasonable that multicellularity is not primitive for life.
>
>
> Agreed, and irrelevant.
> Consider this farfetched but possible scenerio, if fossil evidence ignored:
> Simple replicator -> complex replicator -> simple pre-cells -> true
> cells -> multi-cellular -> some clades revert to single cells.
> Now suppose the first four stages have all gone extinct, replaced by
> the next-to-last stage, then later evolution produced that last stage,
> and only those last two stages are observed today, with LCA of all
> extant life multi-cellular.
How probable would that be? Yes, it's a logical possibility. But this
isn't mathematics. It's science.
>>Second, we may propose that too great a departure from a molecular
>>clock is unlikely, so that if we are required to accept evolution
>>proceeding thousands of times faster in some mammals than in others, we
>>might perhaps reject roots that do so.
>
> If we allow the genetic clock to vary by a factor of a thousand, but no
> further than that, we may be able to eliminate root within mammals, but
> the root could still be within other metazoa, such as sponges.
No problem if what we're trying to do is root mammals. So we've
established is that you can root at least some parts of the tree by this
method. That's a start.
> Farfetched mechanism to support the farfetched sponge-root hypothesis:
[snip bizarre fantasy]
>Without any fossil
> datings, how can you refute that?
How can you refute it with fossil datings? It's highly unparsimonious,
certainly. And in fact most of the action would have happened in a way
invisible to most fossils, a billion or so years ago.
>>>That seems to be an argument for my position of discarding the
>>>traditional taxonomy completely and starting from scratch using
>>>inherent character polarity (SINEs DUPs etc.) to establish portions of
>>>the tree as true clades and leave the root as unknown location within
>>>the remaining part of the tree.
>>
>>If you want to do that, feel free. Do you really believe that?
>
> Yes. I believe we don't yet know precisely where the root was, or even
> whether there was a single root, although we have it (or them) narrowed
> down quite a bit. But I haven't yet seen the solid evidence to support
> all those narrowings, and you haven't helped much, so at this point I
> can't show anti-evolution folks anything to convince them my belief is
> better than theirs.
I remain mystified about why you actually believe in evolution.
>>Have I converted you into an evolutionary agnostic?
>
> Huh?? I'm strongly in favor of evolution. I'm just root-agnostic.
Why are you in favor of evolution? If I believed your arguments, I would
be forced to reject most of the evidence for evolution.
>>Hundreds of sequences exist for many genes.
>
> I'm not sure what you mean by that sentence.
> Do you mean:
> - There are many genes, such that for each such gene hundreds of
> sequences (i.e. homologous for that gene in hundreds of species) exist?
Yes
> - There are hundreds of sequences, distributed among many genes,
> perhaps only one or two different species per gene, for example 50
> different genes, which is "many genes", each sequenced in 6 different
> species, would yield 300 total sequences which is "hundreds of
> sequences"?
No.
>>Is it sensible that life began with Metazoa, and that all other
>>groups have emerged from within it?
>
>
> No, but it's sensible tht life began with something other than Metazoa,
> evolved to Metazoa, then degraded to Protozoa along some branches, just
> as parasites have evolved degraded from fully-functional lifeforms.
> Evolution isn't a one-way climb up the latter of complexity. It must
> have originally started very simple, and then evolved to become more
> complex, but there's no rule it can't evolve "backwards" to greater
> simplicity, not following the same path in reverse, but nevertheless
> following some other path that takes complexity back down many notches.
Is it sensible that all life prior to Metazoa would have gone extinct?
Since metazoans require autotrophs, that means that all prior autotrophs
would have to have gone extinct after the evolution of metazoans, while
being replaced by metazoan-descended autotrophs, all the while
maintaining a viable biota. And of course that would require
chloroplasts to have evolved from metazoans and have been incorporated
into metazoan-derived plants.
> I bet that among the tens of millions of different species today,
> there's at least one where an obligate symbiont is "just barely"
> obligate, where it it tried to live outside its host it'd die, not
> because it can't perform basic life functions, but simply because life
> is too harsh outside the host, just as a baby kangaroo can survive just
> fine outside mama's pouch, but it's just too dangerous there, so all
> kangaroos at that age stay in the pouch for a few more months. When we
> find such a species, I bet in the lab we can extract the symbiont and
> raise it in a suitable media where it can survive and replicate
> indefinitely, and then by varying the conditions we can force it to
> evolve to become robust enough that it could survive out in the real
> world. The very earliest mitochondria were like that, and if some
> global catastrophe had happened just then to make it better for some
> mitochondria to live outside the host, that could have happened.
This doesn't fit the unrooted tree, which has mitochondria joining the
tree at a separate point from the rest of the eukaryotes.
>>guesses at the polarities of many characters, ..., multicellularity
>>terrestriality, wings, etc. Even if these have been lost from time to
>>time (and they all have) or acquired independently from time to time
>>(and some of them have), they can still be used to root portions of the
>>tree of life.
>
> Sure, all those sound reasonble, just as the pig bulding a house out of
> straw seemed reasonable, until the big bad woof showed it to be a
> mistake. Yeah, that's just a fictional story for children, but it has a
> good moral point. Just because something seems reasonable at the time,
> doesn't show it's actually correct. I want to present a checklist of
> evidence and progressive logic to show evolution is very strongly
> supported, not just that it seems reasonble as a guess. If I want to
> establish a particular tree of life, which *really* happened, we are
> very sure of that particular tree, and then as a side remark "obviously
> evolution happened, not only did it happen but it happened in that
> particular way I demonstrated", then I need solid evidence, not guesses
> that seem as reasonable to us as a straw house seemed to the mythical pig.
This conversation shows that you are not equipped to show anything to a
creationist.
>>>Does the unrooted tree have all protists in one
>>>branch and all metazoa in another branch, with a single link between
>>>those two pseudo-clades, i.e. P--M in short form?
>>
>>Yes.
>
> Good, then we have four choices (root in P, root in M, root between P
> and M, no such thing as evolution in the first place so no root at
> all). Do you have any evidence to eliminate that fourth choice? Do you
> have any solid evidence to eliminate the second and third choices too?
Not that you would accept. Therefore you should reject evolution.
>>>Are you saying that among those few which *have* already been
>>>sequencing, they fit into the simple P--M model, and you're predicting
>>>the 90% not yet sampled will likewise fit that same model?
>>
>>I'm predicting that most of them will. There may be somewhere, somehow,
>>and unknown "protist" that is really a degenerate metazoan, but even if
>>there is, I don't see the problem.
>
> In that case we'd have to re-define the word "Protist" not to mean any
> single-celled eukaryote, or abandon the word totally, neither of which
> is a problem in my opinion. If we invent new taxa (names) only for
> firmly established clades, not any more "type names" as with the old
> system, such as "Pisces" which is known very well not to be a clade,
> I'll be satisfied.
So why bring it up?
>>Given the data we have, relationships among the taxa we know about
>>are clear in this case. Likewise, tetrapods are still monophyletic even
>>if we know about snakes.
>
>
> It depends on your definition of "tetrapod" of course. Actually I'll
> use the technical name "Tetrapoda" instead.
> <http://tolweb.org/notes/?note_id=471>
> There is currently a debate concerning the definition of the taxon
> Tetrapoda, which is part of a wider debate dealing with the definiton
> of phylogenetic groups. ...
> then it goes on to propose two cladistic definitions, the crown group
> (smallest clade containing currently-living Amphibia and Amniota),
> which is the least-inclusive reasonble definition, and the stem-based
> definition (largest clade containing that crown group but not
> containing sister groups such as lungfishes or coelacanths) which is
> the most-inclusive reasonable definition. Then it goes on to show
> problems with either definition, namely the least-inclusive definition
> excludes many clear cases of land tetrapods with all the modern
> characteristics definining such, whereas the most-inclusive definition
> includes many clear cases of animals that are in no sense really
> four-legged land animals.
>
> IMO, the taxon "Tetrapoda" should be abandoned, just as "Aves" is being
> abandoned, for a similar reason, replaced by new terms for each of the
> crown group and the stem group.
Did you have a point?
[snip extraordinarily pointless bit]
> <http://www.uky.edu/KGS/education/Devonian.htm>
> Recent Findings Fishes with Legs, Devonian Times. Good summary of the
> fossil evidence for the evolution of fish to amphibians. A table is
> provided that includes the names of the known "transitional" tetrapods
> (four-legged animals) with links to summaries and Web sites for those
> animals. ...
> Would you dismiss that evidence as you generally dismiss all fossil evidence?
I don't dismiss fossil evidence. Do you have a point?
> The Fish That Was Not a Fish, Discover. A news report on the finding
> of an apparent tetrapod fossil, Pederpes, which is older than
> Acanthostega and Icthyostega, its two more famous cousins. More
> information about this intriguing fossil and its relation to other
> Devonian tetrapods can be found at the Evolution of Tetrapods and the
> Closing of Romer's Gap.
> -> <http://hometown.aol.com/darwinpage/tetrapods.htm>
> ... A good series of
> specimens permit scientists to distinguish between ancestral and
> derived characters, ...
> That sounds like the kind of evidence you always dismiss as unreliable.
Yes or no, depending on what the writer (whoever it was) actually meant.
A good series of specimens does allow scientists to distinguish between
ancestral and derived characters if you can put the taxa into a tree and
then root that tree. You will note that there is no mention in that
passage about using the ages of fossils to polarize characters or root a
tree.
> Hmm, of more general interest about methodology:
> <http://www.pnas.org/cgi/content/full/98/19/10751>
> A major issue in all data collection for molecular phylogenetics is
> taxon sampling, which refers to the use of data from only^ a small
> representative set of species for inferring higher-level^ evolutionary
> history. Insufficient taxon sampling is often cited^ as a significant
> source of error in phylogenetic studies, and^ consequently,
> acquisition of large data sets is advocated. To^ test this assertion,
> we have conducted computer simulation studies^ by using natural
> collections of evolutionary parameters --- rates^ of evolution,
> species sampling, and gene lengths --- determined from^ data available
> in genomic databases. A comparison of the true^ tree with trees
> constructed by using taxa subsamples and trees^ constructed by using
> all taxa shows that the amount of phylogenetic^ error per internal
> branch is similar; a result that holds true^ for the neighbor-joining,
> minimum evolution, maximum parsimony,^ and maximum likelihood methods.
> Furthermore, our results show^ that even though trees inferred by
> using progressively larger^ taxa subsamples of a real data set become
> increasingly similar^ to trees inferred by using the full sample, all
> inferred trees^ are equidistant from the true tree in terms of
> phylogenetic error^ per internal branch. Our results suggest that
> longer sequences,^ rather than extensive sampling, will better improve
> the accuracy^ of phylogenetic^ inference.
> Any comment, honorable expert in the field?
No. I don't disagree with that result as far as it goes. What do you
think it has to do with anything we're talking about here?
>>>If we restrict the data going into construction of our unrooted tree to
>>>*neutral* sequences only, then I'd buy the clock hypothesis as a
>>>reasonable way to calculate the odds. Unfortunately, prior to
>>>whole-genone sequencing, most of the known sequences were genes that
>>>code for polypeptides which have important biological functions,
>>>correct?
>>
>>No. Certainly they are represented in a highere proportion than their
>>proportion in the genome, but I wouldn't say "most".
>
> Oh, so if I were to reject all cladograms based on expressed characters
> that are probably selected, and insist only presumed-neutral DNA be
> used, so that the contamination from selected sequences be reduced to
> just that small part which is believed neutral because it isn't known
> to code for anything but which unbeknownst to us actually is selected,
> such restriction wouldn't throw out a majority of the data, so it would
> be sufficient to generate meaningful cladograms? Has such an attempt
> been made, any that you know of?
Have you considered making your sentences shorter, so as to be more
readable? If you are asking whether cladograms have ever been produced
strictly from non-protein-coding DNA, then yes, that happens all the
time. You understand that sequences that are not under selection evolve
too fast to be useful for the most ancient divergences, right?
> (Breaking reply here, will reply to rest of your article later...)
> .
>
.
- Follow-Ups:
- References:
- Prev by Date: IF THERE IS A GOD OR GODS WHAT EXACTLY ARE WE?
- Next by Date: Re: IF THERE IS A GOD OR GODS WHAT EXACTLY ARE WE?
- Previous by thread: Re: Part 1 (of 3): What are major aspects of evolutionary theory?
- Next by thread: Re: Part 1 (of 3): What are major aspects of evolutionary theory?
- Index(es):
Relevant Pages
|
Loading
