[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]
Re: Query on so-called "correlated" characters, reply
I think about another mistake, worst than the orthographic ones.
I said I though of the absent fenestra as an end of a continuum in
sizes of the antorbital fenestra, and thus to include it in a
continuous character defined on areas. Now I think that we can also
want to define a character related to the lenght of the fenestra, and
in such a case we would need to define lenght as 0 cm in the case of
the absent fenestra, based on what I said in my previous discussion of
the quadratojugal character (and contradicted when accepting inclusion
of absence in a character based on area in the fenestra character). Of
course, 0 in area and 0 in lenght would be correlated for taxa without
fenestrae. This suggests we should put absence/ presence in other,
binary, character, and the sizes measured in different senses in
different characters where we codify the feature as impossible to
apply, as indicated by David (M).
However, there is something I think may be wrong with the case of a
continuous character, specifically the metrics: we would assign to
presence-absence a weight (1) disproportionate from continuous changes
in size of the fenestrae. I would rather think the change in area from
0 cm x cm to 1 cm x cm is not greater than from 1 cm x cm to 2 cm x
cm, at least arythmetically, as continuous characters are ussually
taken (this is something we can also put into doubt, where the 0
metric of absence I suppose will present a similar problem). However,
I cannot think of a better solution yet, because scaling of continuous
characters is until our days a problem, according to Goloboff e al.
(2007, in Cladistics, on continuous characters).
2009/5/23 Augusto Haro <email@example.com>:
> sorry the foramens (sic) and other "sics" around the text
> 2009/5/23 Augusto Haro <firstname.lastname@example.org>:
>> 2009/5/22 David Marjanovic <email@example.com>:
>>> ----- Original Message -----
>>> From: "David Peters" <firstname.lastname@example.org>
>>> Sent: Saturday, May 23, 2009 12:08 AM
>>>>>> 2. Quadratojugal process of jugal present or absent.
>>>>>> Quadratojugal present or absent.
>>>>> The first must be scored as unknown (which, for PAUP*, is the same as
>>>>> "inapplicable") in all taxa that lack a quadratojugal.
>> This seems correct. I have heard of the argument of making the
>> character additive, with states quadratojugal absent (0),
>> quadratojugal present, without process (1), quadratojugal present,
>> with process, (2), as if you were progressively adding things. I think
>> this is likely not certain in this case because the presence of the
>> process is likely a by-product of the formation of the infratemporal
>> fenestra (by delay in ossification, I think to remember). Being more
>> general, if there is any other process in which we observe variation,
>> which cannot be hierarchically correlated with the process already
>> assessed, we should codify with the same logic both processes, and
>> would not consider both additive in the same sense cited above because
>> we would repeat the change from presence to absence. So, I think the
>> best option is doing what David (Marjanovic) indicated.
>>>> And there's the opposite case:
>>>> Antorbital fenestra with or without a maxillary fossa. 'Without a fossa'
>>>> appears on chroniosuchids, pterosaurs and Proterosuchus, which are not
>>>> related. The fossa is not necessarily correlated to the presence of a
>>>> fenestra. It either is present, or it is not.
>>> I'd say they're simply correlated the other way around: the fossa is only
>>> present if the fenestra is present, never otherwise. One ordered multistate
>>> character: both absent (0), fenestra present and fossa absent (1), both
>>> present (2).
>> Well, in this case I think that we can alternatively consider both as
>> not correlated (except hierarchically), for the multistate character
>> you proposed is operationally the same of two binary characters being
>> one presence/absence of the fossa and presence/absence of the
>> fenestra. I think the currently known distributions implies the the
>> character of David (M) is not wrong, as do the fact that according to
>> Witmer (1997, MSVP) the fossa forms by a sac necessarily transversing
>> the fenestra to reach the fossa from the nasal cavity. But I think
>> making two independent characters may better accomodate, without harm,
>> the possibility of a specimen without antorbital fenestra, but which
>> presents a fossa similar in location and depth to the antorbital
>>>>>> 4. Naris larger/smaller than antorbital fenestra. Antorbital
>>>>>> fenestra present or absent.
>>>>> The first must be scored as unknown in all taxa that lack an antorbital
>>>> Yet, if no antorbital fenestra is present, isn't the naris, by default,
>>>> larger than the antorbital fenestra? Why not score it that way? How does
>>>> antorbital fenestra phylogenetically begin? I've seen foramina that might
>>>> be the start of an antorbital fenestra. Do you ignore those?
>>> That needs to be demonstrated first. It needs to be shown that the
>>> antorbital fenestra starts small and then becomes larger. As long as that
>>> isn't established, it's safer to score the relation character as unknown
>>> when the fenestra isn't there.
>>> Why should foramina and fenestrae be homologous? Foramina contain nerves
>>> and/or blood vessels; fenestrae can, but only by accident.
>> Well, here I think we can see foramina and fenestrae as part of a
>> continuum. There are foramens which can enlarge to make fenestrae
>> within which a nerve is located, as with the obturator fenestra of
>> mammals. The suborbital fenestra of diapsids seems to present a great
>> variation in area. Their difference is in area, and there are nerve
>> foramina of different areas, and the same applies for fenestrae. I
>> think we should not deal at this point with which is the origin of the
>> fenestra; as far as we can we should only observe the similarities and
>> dissimilarities (which is not always easy).
>> In my view, an argument for seeing the foramen closer to the absent
>> condition than the fenestra, and so including the absent into a
>> continuum, is that if we define the size as area, we should say that 1
>> cm x cm (suppose this means foramen, the character should be defined
>> better in a ratio with other feature) is closer than 0 cm x cm than
>> (absent condition) than a fenestra of say, 30 cm x cm.