Re: Yet again, human evolution: huh?


John Harshman wrote:
> Frank Sullivan wrote:
>
> > I'm still not entirely grasping the concept of the nested hierarchy. I
> > think I've almost got it, but a complete understanding eludes me.
> >
> > And I'm still not sure why all relationships aren't being considered.
> > For instance, the first site shows a relationship between humans and
> > gorillas that is not shared amongst the other three. I know I've asked
> > this question before, and you've answered it, but something up in my
> > noggin isn't clicking.
> >
> These are both difficult questions that I would like not to answer at
> lengths, and so would like to find a short way. Let me grope around for
> a while.
>
> Nested hierarchies are hard to explain without graphics, and ASCII
> graphics aren't up to it either. The simplest way is to think of Venn
> diagrams. In a true nested hierarchy, none of the circles cross each
> other, that is there is no partial intersection of groups. So for every
> pair of circles, either one circle is completely inside another one, or
> the two circles don't intersect at all. Drawing some pictures for
> yourself that fit these criteria might help.
>
> Another way is just to talk about groups within groups, as in this short
> classification in which each group includes every group that's indented
> from it:
>
> Animals
> Insects
> Beetles
> Lady bugs
> Dung beetles
> Flies
> Mosquitoes
> Horse flies
> Mammals
> Even-toed ungulates
> Cows
> Whales
> Odd-toed ungulates
> Horses
> Tapirs
>
> I have arbitrarily decided to make each level of the hierarchy have two
> included groups, and of course I've left out most of the possible
> groups. But that's a nested hierarchy. You could easily represent it by
> a Venn diagram of the sort I've suggested previously, in which every
> named group labels a circle. And it should be obvious that this is the
> sort of structure you naturally get from a branching tree, in which the
> tips of the branches will all cluster with other tips that join to the
> same branch, and those branches with larger branches, etc.
>
> OK so far?
>
> Now let's talk about homologies. Homologies are nothing more than events
> -- changes -- that happened somewhere on the evolutionary tree. They
> provide clues to the structure of that tree. If there is no homoplasy,
> that is if every change is unique and no identical changes happen twice,
> every change defines a branch: all descendants of that branch will have
> the changed characteristic, and nothing that doesn't descend from that
> branch will have it. All we have to do to reconstruct the tree is to
> sort the organisms into bins, those with and without each change. If we
> do this, we should get bins within bins, each bin matching a particular
> branch on the tree.
>
> But of course there is homoplasy. There being only 4 possible bases at
> any site, sometimes the same change does happen twice. My argument
> relies on those extra changes being randomly distributed, while the
> changes that follow the tree are not (because they follow the tree).


Ok thanks John. I'm good so far. Sometimes, as an exercise, I imagine
how I would explain a concept to someone else. Doing that helps me to
know if I truly understand something, or if I just think I do.

So could start off with the Venn diagram explanation (which was very
helpful by the way), and this will help them to understand how, for
example, all humans are primates, but not all primates are humans; all
primates are mammals, but not all mammals are primates, and so on.

However, these are man-made labels that represent, I'm guessing,
different combinations of characteristics. So, if you have a spinal
chord, a backbone, a skull, hair, and mammary glands, then you are a
mammal. If you also have a large brain and walk upright, then you are a
human.

So, in a sense you could remove the labels and say things like "all
animals with skulls have spinal chords, but not all animals with spinal
chords have skulls."

But convergent evolution really messes this up for me, and the fact
that some characteristics disappear also messes it up for me. For
instance, would it be so far fetched for an octopus to evolve a skull?

I guess I already asked that question too, and you answered something
to the effect that phylogenies aren't pieced together with just one
characteristic, but rather with a combination of characteristics. I
guess maybe I do understand what a nested hierarchy is, however the
real world is just too complicated for me to see the nesting without
taking at least a few classes (perhaps more than that?), which I
definitely plan to do.

> And now, why I don't consider every possible set of relationships: It's
> because I'm interested in the question of whether or not there is a
> branch on the tree that separates the three African apes (human, chimp,
> gorilla) from the other two (orangutan, gibbon). I'm interested
> therefore only in sites that either confirm or deny that hypothesis.

I was under the impression that a site which showed a relationship
between gorillas and humans, which was not shared by chimpanzees, would
deny the hypothesis, because it would show that chimpanzees have a
closer relationship to orangs and gibbons than to humans and gorillas.

> Sites that suggest relationships within the African apes are not
> relevant, because they could be that way whether or not the African apes
> formed a real group, that is whether the branch I'm asking about exists
> or not. Again, perhaps graphics would be helpful, but I would like to
> avoid more graphics if I can.
>
> If I did, though, here's what I would show:
>
> Go C O Go O C
> \ | / \ | /
> \ | / \ | /
> \___|___/ \___|___/
> / x y \ / x \
> / \ / \
> / \ / \
> H Gi H Gi
>
> On both trees, the branch supported by site 1 exists, noted by the x.
> The branch I'm interested in, y, appears on the left tree but not on the
> right tree, but this makes no difference to branch x or site 1. Site 1
> is irrelevant to the question I'm asking.
>
> Is that clear?

Not quite yet. Don't hate me! I think the crux of my misunderstanding
here is that it seems to me that site 1 shows a chimp/gibbon/orang vs.
human/gorilla relationship, which ought to contradict the hypothesis
being questioned. You and many others have told me that this
relationship does not contradict it, however I can't grasp why. I think
that if I were to cure this one fallacy in my reasoning, everything
else in your original post will fall into place and make perfect sense.

.

Relevant Pages

  • Re: Yet again, human evolution: huh?
    ... >>>I'm still not entirely grasping the concept of the nested hierarchy. ... >>provide clues to the structure of that tree. ... > example, all humans are primates, but not all primates are humans; ... would it be so far fetched for an octopus to evolve a skull? ...
    (talk.origins)
  • Re: Avatar observations.
    ... After the humans departed, Jake solved this problem by transferring ... the tree roots. ... As far as we can tell, the bio-net is also sentient. ... commanding the Pandoran animals to attack the RDA forces. ...
    (rec.arts.movies.current-films)
  • Re: Co-optation Today
    ... a tree is nothing but nodes and arcs. ... I don't think a nested hierarchy can exist even in the abstract with no values. ... Sounds reasonable, if by "value" you mean only some kind of label, and if by "leaf" you refer to a branch with one end free (i.e. a node with only one branch connected to it, a terminal node). ... When a message starts a new thread, its reference string is null. ...
    (talk.origins)
  • Re: More proof you boons dont know what your on about.
    ... Or the amount of genes a life form has. ... Screamed spintwitty at the face in the mirror. ... A nested hierarchy is 'groups within groups'; your inexplained 'tree' ...
    (talk.origins)
  • Re: "Logic" trumps science?
    ... scales and fractals exhibit self-similarity across scales, ... in the way that the tree of life is. ... nor is it what makes the tree a nested hierarchy. ...
    (talk.origins)