I have asked this question in various forms in
various places before. I have not yet been able
find a satisfactory answer. Where is there a
textbook for the sort of "foundations" course
in geology that one would expect to see at a
senior undergraduate or beginning graduate
level? One would expect that at this level, now
that a student is familiar with the essentials of
the subject and is presumably a serious student,
that there would be a course in which the basic
logical foundations of the subject are thoroughly
examined. I have never been able to find such
a book.
It seems to me that one of the most essential
notions in the study of geology is the claim that
one can make reliable rock/formation correlation's
over great enough distances and with great enough
frequency to be able to make sense out of the
geologic pile over large distances. Now I realize
that one need only step outside to be convinced that
one can make correlations over small distances.
What I have never seen is a thorough quantitative
examination of the degree to which one can make
reliable correlations over great distances. I find
this incredible, that many geologists would like to
give the impression that theirs is an academic
discipline, yet it seems so difficult to find a good
treatment of this question which one would
think would be at the foundation of their subject.
I can accept that to some extent geology is a craft,
much as pouring cement is a craft. I would not
necessarily expect to find a "cement 401" textbook
examining the foundation principles of pouring
cement. However, many geologists give the
impression that theirs is also an academic discipline.
They seem to give the impression that the rest of
society should be convinced of the truth of their
conclusions. I find many texts giving a "description"
of the principles of geology. I have never found
a text giving a rationally convincing quantitative
"defense" of the principles of geology.
I find this lack amazing. I am a statistician working
in an aerospace engineering environment. In that
world when someone makes a design mistake
something crashes and there is hell to pay. Other
mistakes can waste large amounts of money.
( The US space shuttle program is a good example. )
Often there is the situation where large communities
of engineers are working to get some particular
approach to building something sold to a customer,
when in fact their approach is a bad one. Good
engineers know enough to be careful of this. You
often hear them ask the question "Are we drinking
our own bath water here?" (Other more colorful
phrasings also come to mind.) They are asking
whether in their enthusiasm to promote a project,
they are coming to believe their own falsehoods.
What I keep looking for is a text where geologists
examine whether they are drinking their own
bath water.
I realize that the rock pile we all live on is
enormously complex and a complete description
is impossible in a textbook. What would I expect
to see in such a foundations text? First, I would
at least expect to see some sort of index or quantity
defined for a formation that is a measure of the
distance over which points in the formation can be
reliably correlated using only those properties of
the rock unit that are present everywhere in it,
such as color, chemical composition, texture etc.
I would expect to see example scatter plots each
for a specific location, on which each formation
present at that locality is represented by a point
on the plot - the axes of the plot would be the
formation thickness versus this correlation
distance index. I would expect to be informed
about where I could go to find such scatter plots
for most any locality I was interested in. I would
expect also to see some index developed indicating
the quality of the correlations represented by each
such scatter plot. I would expect to see various
distributions showing the variation pattern of such
a quality index over the earth. I don't claim to have
the best way figured out to do this but that is not
the task of the student. I would expect to see some
scheme showing this information, though. Again, I
am amazed to not be able to find such a text from
an academic community many of whom seem to
think that their conclusions should be believed by
the rest of society.
A second group of notions I would expect to see
developed in such a foundations text would be
notions relating fossil correlation and/or other
correlations based on potentially sparse data points,
to the above types of correlation indices. I would
expect to see indices developed relating fossil
densities to the above homogeneous properties
correlation distance measures. Again, I would
expect to see scatter plots and distributions showing
the global variation in such fossil correlation quality.
Thirdly, I would expect to see a rational case made
for being able to make correlations over great
distances of whatever quality is claimed, using
these measures. Of course, this quality will vary a
great deal. I would expect to see a quantitative
description of this variation.
Is this expecting too much from a large community
of scholars who claim to have an academic
discipline? I have a hard time thinking that it is.
Let me go a little further to put my question in
a less abstract setting. I have no doubt that
rock correlations of various types which are
of great practical value, can be made with
considerable confidence. It seems clear that
this can be done well especially within a basin
or other such region, in the case of some of the
formations. For example, I live near the "hogback"
that runs along the mountains west of Denver CO.
It is very easy to correlate the Dakota formation
or the Fountain formation along that hogback
over great distances. The Dakota forms the ridge
of the hogback and the Fountain gives Red Rocks,
Roxborough and Garden of the Gods Parks their
distinctive formations. However, beneath the Dakota
and above the Fountain are the Morrison and other
formations. I have a good deal of difficulty in visibly
correlating the Morrison versus other formations
beneath it, between roadcuts that are only a couple
of miles apart. I am not claiming that there is no way
of reliably making such correlation by means other
than what is blatantly visible, I am suggesting that
after reading such a foundations text, I should
know how to go about firmly establishing the
degree of confidence in such a correlation.
Naturally I also find myself in question about the
confidence of correlations between my side of
the mountains and the other side of the divide,
such as between say the Morrison formation
near Red Rocks Park near Denver and the Morrison
in Dinosaur National Monument in western Colorado.
Norm Smith
74532.66@compuserve.com
4/17/98