Re: mesosaurus

[David Marjanovic <david.marjanovic@gmx.at>]

> BTW, how does PAUP differentiate massive convergence versus homology? Or
> does it take human intervention (weighin characters, etc>) to note the key
> differences?

Erm... huh? PAUP* does cladistics. This means that it will give you those 
distributions of character states across trees ( = it will give you those 
trees) that can be explained with the fewest assumptions of convergence ( = 
the most parsimonious ones). It tries to distribute all characters so that 
each state change happens once, and balances them against one another so 
that the fewest additional changes must be assumed. If... I hope this 
sentence isn't too long... :-/ if there are enough characters the most 
parsimonious distribution of which contradicts the assumption that the 
character change in question happened only once, the character change in 
question will be shown as having happened several times convergently.
       If you don't trust the result because the character is related to 
(say) burrowing, then add more characters (see "enough" above) and/or more 
taxa to test your distrust. More taxa, especially more basal taxa, help 
because if the basal taxa of two clades retain plesiomorphies that the 
derived ones lack, the corresponding apomorphies can no longer come out as 
synapomorphies of the two clades.
       Something tells me I should use PowerPoint to explain this...

> The answer will not only affect the mesosaur question, but another one
> that recently came up about "worm lizaards" and "legless lizards", which
> are all definitely lizards, but are they all related? Or does burrowing
> impose convergence?

Which ones do you mean, amphisbaenians and anguids? They seem to be rather 
distantly related... yes, burrowing does impose a lot of convergence.