Re: Nature of the fossil record

["Doug Erwin," <MNHPB028@xxxxxxxxxxx>]

Thanks to Dave Lindberg for a good reply.
To Stefan: I largely agree with you.  I was not arguing that we should
attempt to identify the causes of the S-L effect - although in individual
sections it might be possible (for example to pick up pronounced size
bias, or the loss of silicification in the P/Tr section in Abadeh, Iran)
but not generally.  I completely agree that it is largely a statistical
phenomenon, which is sort of the point of thinking about null models.
  To my mind, the S-L effect, like Pete Sadler's work and David Shindel's
all demonstate that there are definate limits on the sorts of questionns
that we can ask as paleontologists AND expect to get meaningful answers to.
I would differ with Stefan in that I see great promise in the Marshall and
Sadler techniqes, even for dealing with the last meter of the Cretaceous,
for they do allow one to potentially place limits on the range of true
patterns which may have produced the observed pattern, thus potentially
excluding some hypotheses.  With the appropriate techniques, properly used,
the absence of data can be useful  and tell us "something real".
  (actually I don't believe anything younger than the Triassic is "real")
Doug Erwin