[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Color revisited (was Re: Origins (was: Re: Sharovipteryx))

Brian McCarthy  <philidor11@snet.net> wrote (first quoting me):

>> Contrast that with another example I'll pull out of thin
>> air.

Oops, I forgot about that part when I said in my most recent message
"the only part that was a joke was..."  The "thin air" was also a
joke.  There were two jokes in that message: somebody should write a
paper, out of thin air, and my ruthless efficiency... three!  There
were three jokes in that message!

My apologies to Monty Python.

> Your argument, persuasive as it sounds to me, is still an assemblage
> of inferences, no?

Yes, and paleontological science is not different from any science in
that respect.  I don't want to venture too far down that road here
because it is yet another philosophical tangent, but I'll pursue it
briefly below.

> Wouldn't you say that there is a difference in the testing of your
> hypothesis that dinosaurs had color vision and the testing of the
> hypothesis that a particular virus in your lab causes a particular
> type of influenza?

In principle, no.

> The contrast here is not between true/false but between
> direct/indirect proof.  And indirect proof is permanently vulnerable
> to more types of refutation.

We could get hopelessly bogged down in semantics -- in my experience
that's the typical outcome of a philosophical debate.  However, I
would submit that your distinction between direct and indirect in this
context is illusory.  Sticking with Chris' example, I suppose you know
that viruses are too small to be seen with light.  Aside from their
inferred effects, we only "know" viruses exist because we can "see"
them with electron microscopes or we can bind a lot of them (or
constituents thereof) into things which will clump together into a
structure large enough to be resolved by light.  There are many many
steps involved in both means of detection, and any one of those steps
could be seen as a level of indirection.  Closer to home, as David
Hume argued quite forcefully, you can't ever "prove" causality; you
can only show that one event is correlated with another which tends to
precede it.  Traditionally this is where people start talking about
pool balls, but I'll stick to viruses.  The way to "prove" that an
infectious agent causes a disease was outlined by Robert Koch in the
1870s.  People still argue about "Koch's postulates" (see for example,
the October 9th, 1998 issue of _Science_).  One of the primary
arguments is that modern technology allows us to perform *less direct*
tests that generate results we should accept as allowing us to assert
causality even though the biology of the organisms (or our state of
knowledge about them) do not allow us to satisfy Koch's postulates
explicitly.  Do you deny that HIV causes AIDS?  I think it's
irrational to think that, but some people do anyways.  There's a
firestorm brewing about that particular subject in South Africa right

But I'm digressing so let me pull us back.

While evilly grinning, Josh Smith <smithjb@sas.upenn.edu> wrote:

] However, if birds see UV frequencies, and assuming that birds are
] truly descended from coelurosaurs, and assuming that UV processing
] didn't arise somewhere else along the line of decent or conversely
] assuming that the vision of avian ancestral taxa hasn't changed
] significantly, you might be able to wave your arms a bit about the
] possibility that perhaps coelurosaurs might have been able to
] perhaps maybe process UV frequencies.

Typed like a man who truly did not read my paper!  The case is much
stronger than Josh allows, though I will use his comments as the
springboard for my reporting that there are nuances I didn't know when
I wrote the paper.  The paper's barely a month old and already
partially out of date!  Phylogenetic trees of visual pigment genes
show that the UV pigments of goldfish, parakeet, anole (etc.) are more
closely related to each other than they are to the other visual
pigments retained by those same animals.  That is, for example, the
goldfish's UV-sensitive (UVS) pigment diverged from the UVS pigment of
parakeet later than the UVS pigment of goldfish diverged from any of
the other visual pigments in the goldfish retina.  That means that the
UVS pigment in goldfish and parakeet was inherited from their most
recent common ancestor.  Consequently, barring secondary losses that
pigment would exist in all Tetrapods.  Ironically, despite our
complete lack of UV sensitivity(1) we retain that particular pigment.
One way to view the data is to deduce that in us that pigment's
absorption spectrum shifted toward longer (read visible) wavelengths.
In my paper I suggested that this shift occurred after mammals
diverged from reptiles, since in reptiles the pigment is generally (so
far as we know, but keep in mind that we've investigated a fairly
small number of animals) still primarily sensitive to UV wavelengths.
Now here's my out of date part...  I saw a couple of relevant posters
at the annual meeting of the Association of Researchers in Vision and
Ophthalmology earlier this month.  The authors of those posters were
characterizing differences in the amino acid compositions of the
visual pigments to see which differences are most important for
setting the particular features of the pigments' absorption spectra.
It turns out that goldfish and birds tune their UVS pigments into the
UV region of the spectrum via different parts of the amino acid
chains.  It could be that these differences resulted from an analog of
"silent substitution" -- that is, perhaps the pigment always retained
its UVS status in both lineages, and each time a mutation affecting
its tuning occurred in one part of the chain there was a compensatory
mutation bringing the pigment's overall sensitivity back into UV
wavelengths.  However, a viable alternative is that UV sensitivity in
birds, lizards and fish is the result of parallel evolution which just
happens to use the orthologous gene.  The same argument could be made
for the mice which have sensitivity to UV.  They use the orthologous
gene to generate their UVS pigment, but is that because the pigment
was always UVS in those rodent lineages, or did mice also shift their
sensitivity to the UV in another parallelism?  UV sensitive mice use
yet another set of amino acids to push their pigment's sensitivity in
the UV, so perhaps this *is* just a massive parallelism contra my
suggestion (now archived for posterity).  We can test that hypothesis
by performing additional comparative analyses.  And so we return...

Hypothesis: dinosaurs were sensitive to UV light.  Current evidence
still suggests that the answer is yes.  Just like current evidence
suggests that HIV causes AIDS.

1) aside from lacking a primary UV pigment, our lenses and corneas
strongly filter out light with wavelengths shorter than 400 nm --
that's the real reason for our lack of sensitivity to UV light; the
visual pigments we have will respond to light at UV wavelengths, and
people who've had their lenses removed do indeed see UV.

Mickey Rowe     (rowe@psych.ucsb.edu)