[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]
Re: New AVES definition refined (more testable?)
> Paraphrasing (I think), you're saying that it is inappropriate
> to make a list of characters and conclude that a species which
> has X number of those characters must be related to another species
> or group of species.
I think he says it's inappropriate to use the possession of _any_ out of a
number of characters as a synapomorphy. As a fictional example, to define a
group by "the possession of either D-sectioned or unserrated pmx teeth, or
both" is certainly not a good idea.
> Your reason is that:
> <As far as I know, some genetic agent that groups a certain number
> of characters together, then expresses a random sampling of them
> in every generation, is not known.>
> [...] isn't it possible to say that (just inventing) someone's
> descendants may be identified by either or both of two uniquely
> shaped birthmarks? It is theoretically possible to have alternative
> unique connections, no?
The occurrence of such a situation gets difficult to test a posteriori. I
think such a phenomenon would never be used as an apomorphy, it would be
considered individual variation... which is what it is, actually, no?
> Also, and more significantly, as a lineage evolves through time,
> some of the characters which linked a species to its ancestors
> may disappear. Part of the complexity of cladistic logic is
> that of linking one species to another, which in turn has characters
> which help to identify certain characters as derived/shared from
> the mrca. Thus, a descendant species need not have many of the
> identifying marks of an ancestral species to be included.
> The list of characters linking two species may not be the same
> as the list of characters linking those species to their mrca,
> In that sense, if you drew up a list of characters used in establishing
> a whole hypothesis of relationships, some of those animals would
> have some items on the list and others not.
But you won't get a random pattern. You might get something like:
In this (erm, _very_ simple) case, you'll be able to produce a most
parsimonious cladogram by hand, despite the reversals:
A: acquisition of ch. 2
C: loss of 1
D: 4, loss of 2
E: loss of 3
G: loss of 4
6: reversal of 1
8: reversal of 2
9: reversal of 1
But actually... I just see that 6 might come out in a trichotomy with 1 and
A. Also, it takes 2 steps to reach 8 from (the metataxon) 7, but also from
(the metataxon) 2. It's not possible without some calculating to find the
MPT here :-] Take-home message for me: I've used too few characters compared
to the number of taxa. Gotta look for more.
> If I'm reading you correctly, your observation is about the question
> 'Are these two species related' as opposed to the question, 'Do
> these two species share a common ancestor?'
Certainly not -- as a cladist, he equates these 2 questions :-)
> Cladistic logic, as I tromp along in its path, has an amazing
> number of conditional assumptions and complex associations.
Which might be why cladistics is a scientific and not a philosophical method
Though in this case I think you should just read better literature on
cladistics, such as the archives :-)