Re: [asa] Testing in the historical sciences (was: What causes students...)

From: Brent Foster <>
Date: Mon Sep 11 2006 - 17:39:01 EDT

The so-called experimental sciences are no less "historic" than geology or astronomy. Time is opaque to us. We cannot directly observe the past any more than we can observe the future. The only way we can know about any past events is by observing (in the present) records of the past. we only think we can observe the recent past because we have an integral record of our past experiences called a memory. But our memories are only records of the past, and not even good ones. I would much rather have a photograph of a recent event than have to rely on someone's memory. Our certainty of knowing a past event depends on the quality of the record, not on when it happened. We can no more observe what happened 50 years ago than we can what happened 50 million years ago. If anyone thinks they can then tell me what you had for dinner 8 weeks ago Tuesday. Now of course our records of the recent past, including our memories are very numerous, and those of the distant past are scarse. But!
  the confirmation of theories in the physics or chemistry lab is no less historic simply because the events that confirm the theory happened in the more recent past. The record of a Trilobite buried in Cambrian strata is no less valid than the written documentation of experimental results.

The examples that Keith and David give of testing a hypothesis in Paleontology involves repeatable observation of records of the past. Infact these data are more repeatable and more reliable than experimental results. Repeatability here occurrs when specific outcrops can be revisited, and fauna from a given stratigraphic interval are repeatedly found in that interval. Experimental results can and have been fabricated. Any record of the past can be fraudulent or corrupted. But I have to think the one written in God's own hand (creation itself) is the most reliable.


---- David Campbell <> wrote:

> The important concept is that common descent provides testable
> expectations. A cladogram (or phylogeny) can be understood as an
> evolutionary hypothesis that makes predictions about the anatomy of
> currently unknown or incompletely known organisms, as well as about the
> relative time of appearance of specific anatomical traits. In
> combination with a knowledge of past paleobiogeography and
> paleogeography, such models of biological relationships can also
> provide expectations of where organisms with such anatomical characters
> might be found in the fossil record.

Another aspect is that, although evolution gives testable predictions, there
are also many details that depend on historical evidence, and other precise
details are unknowable.
An example of the latter is proving that a particular taxon is really the
ancestor as opposed to the uncle or third cousin twice removed, etc. of
another. I don't know of any good reason to doubt that the sequence of
species here in the eastern U.S. of Chesapecten jeffersonius-C. madisonius
is an evolutionary sequence, with the oldest C. madisonius looking more
like jeffersonius than do standard madisonius. However, it's always
possible that the actual ancestral population was one that we have not yet
found in the fossil record, although there's little doubt that they looked
much like the known ones.

Of the former, the exact sequence of appearance of unrelated characters will
be a matter of observation rather than prediction. For example, many
bivalves show a tendency to change from pearly to other shell structure.
Some bivalves also go from free-living to attached living. Thorny oysters
(actually close kin of scallops) had lost the pearly shell before becoming
attached, whereas true oysters were still pearly when the first attached
ones appeared, based on the fossil record. Some bivalves still have pearly
shells; some replaced pearl with other shell structures in the Ordovician.
Some attached bivalves have very recent ancestors that were free-living
(such as rock scallops); some belong to lineages that were attaching since
the Paleozoic.

Some opportunity for prediction can be found in such settings. E.g.,
knowing both the relatively early loss of pearly structure in scallop
relatives and the fact that molecular data put the thorny oysters very close
to the true scallops, closer than some things that look more like scallops
at a glance, strongly suggests that we should expect the oldest thorny
oysters to have non-pearly shells. Thus, one line of evidence about the
sequence of such events can be used to predict results from other lines of
evidence, even though they cannot be a priori deduced from evolutionary
theory. Similarly, astrophysics and cosmology can't a priori predict that
our solar system ought to have 8 planets (by latest official definition),
but it can predict things that we ought to observe about the solar system in
light of other observations.

> Dr. David Campbell
> 425 Scientific Collections
> University of Alabama
> "I think of my happy condition, surrounded by acres of clams"
To unsubscribe, send a message to with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Mon Sep 11 17:39:33 2006

This archive was generated by hypermail 2.1.8 : Mon Sep 11 2006 - 17:39:33 EDT