Re: Modern-Human-Neandertal Genetic Admixture


"Dar Habel" <Dar_83001@xxxxxxxxx> wrote in message
news:1129065015.781536.221730@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>
> Jois wrote:
> > "Dar Habel" <Dar_83001@xxxxxxxxx> wrote in message
> > news:1129014770.601438.221290@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> > >
> > [snip]
> [more snip]
[and ditto]

> Nowhere in this paper are the parameters and method of the calculations
> shown. Instead, we are told: "For details of our methods, see the
> electronic edition of this issue on the journal's web site."

This is not unusual, I see the same thing in Science and Nature, even for
charts and other materials that editors (I guess) would rather not expend
another page. Maybe someone in Mikey's or Anne's group could cut and paste
these details into the files.
> > The data that Dale wants from what Mikey calls " . . . a finer
resolution of
> > the archaeological and palaeoanthropological records combined." may
never
> > be available. Does that mean that efforts towards further understanding
> > should halt and all further discussions should be set aside for a later
> > date, or that no one should attempt a "working hypothesis" or maybe a
> > "conditional consensus" in the meantime? Nope - and I don't think
anyone
> > could stop these efforts anyway.
>
> I agree, the efforts are unstoppable. Everyone wants to publish, even
> if what they publish is rubbish. Not saying this is, but.....

LOL but you didn't like it anyway!

I think the question of whether or not Neanders were our ancesters was
resolved years ago.

When I first started reading here the Multi Regional vs OOA (and OOA2) was
nearly a bloody argument, both sides polarized to the extremes and I think
that has softened to a great extent. Both sides seem to have given in a
bit.

The question of the Neanders being a separate species is probably impossible
to resolve. mt DNA can probably function across species so no quantity of
mtDNA will ever do more than "same" or "different". (Also were very hot
discussions here about the molecular clock.)

Nuclear DNA is going to take us no where because we won't ever have anything
from the Neanders that can be counted really useful (too segmented and
short?) and 100% reliable.

What is there besides speculating? (See Palanth - lagar velho -
hybridisation, February 08, 2003, Colin did a nice job but still had to say
he had problems with the underlying assumptions over the numbers. I liked
that one and it seemed logical, too.)

> The problem with this paper is that it sets up two simplified
> alternative scenarios with two options each (actually four scenarios)
> in order to see what comes out. Then it skips to the conclusion.
[snip]

What other alternatives would you like to have seen? Time spans? ( You do
answer this question a few paragraphs after this.)


> Perhaps they did. Without the supplementary information, I don't know
> how they arrived at their conclusions. They've written this paper and
> have concluded that this is the way it was because 'we say so'.

Ah, that technique is used here a lot!

Okay, this is your list of concerns and I'm going to set them in my favorite
format:

> I just think:
1. potential interactions bet. Neanders & AMH are particularly difficult to
prove,
2. let alone try to determine localities,
3. population size of interacting local groups,
4. whether "rapid-disappearance" of Neandertals occurred on a localized
scale when interaction occurred, and
5. dates and locations of these potential interactions.
6. And numerous other demographic difficulties.

Okay, I know I usually don't think of these things until I read some Neander
fossils were found in the mountain caves of some remote place in Italy or
Spain or Hrvatska (LOL).

> The whole framework of their calculations (whatever parameters and methods
of
> calculation) seems overly simplified, as it must, when there are no
> real answers to the above questions (and no doubt other questions could
> be asked). Maybe they've modeled every possibility in order to arrive
> at their conclusion, but somehow I doubt that and, in any case, without
> the supplemental information not included in the primary paper, we just
> don't know how they arrived at their conclusion.
>
> > Maybe when I read Dale's comments a few more times I'll understand them
> > better and have more questions.
>
> Good luck. Dale can be as vague as anyone else.

LOL. I certainly hope so!

Thanks, Dar, come slumming here more often, you add a lot. And you are
going to love the cookie recipes!

Jois


> > http://johnhawks.net/weblog/topics/
>
> > > I'm extremely
> > > leary of interaction probabilities based on calculations made with
> > > unknown parameters. Liars, damn liars, and statisticians!!!!
> > >
> > > Dar


.

Relevant Pages

  • Re: [OT] Hints for an algorithm - avoiding O(n^2)
    ... The obvious way is to compute the force interactions ... > results will not be similar in time (I'm planning to manage about ... You are going to make a simulation, ... to do the calculations between objects much less frequently, ...
    (comp.lang.ada)
  • Re: Modern-Human-Neandertal Genetic Admixture
    ... >> Jois wrote: ... Of course, there is the "supplementary information", and I ... > The question of the Neanders being a separate species is probably impossible ... but the results are from mtDNA calculations. ...
    (sci.anthropology.paleo)
  • Re: Spontaneous particle-antiparticle pair production
    ... When doing calculations in S-matrix theory using the Wick time ordering ... this limits interactions to only ... they interact through bosons, and only through bosons -- and bosons can ...
    (sci.physics.research)
  • Re: Choose: O(n!) on single core vs O(n^2) on multi-core
    ... For example if you have 1000 bodies, divide them in, say, 40 groups of 25 elements. ... An individual UOW should then handle all the interactions between two groups of elements. ... You only need to lock the access once for every group, so you'll be able to compute 25*25 interactions with two sync instruction. ... For a total of 36 calculations, but 6 happening in parallel, so effectivly it's 30 calculations; ...
    (borland.public.delphi.language.basm)