Imagine 100 FMLers buying a brand new file, the type you might buy at a
hardware store, and use it periodically for the next 10 years.  At the end
of that time period, all 100 FMLers mail their file to a central location
where they are inspected.  So, are all 100 of the files going to look
alike, with identical amounts of wear and rust?
 
You know they won't.  Stronger people will push harder.  Some will use
their file more.  Some will clean and oil their file, others will just
toss it in a drawer.  Some will only use the file as it was intended,
others will use it as a hammer, prybar, paint stirrer, or even a lever.
Because of the complicated individuality (use history) of each file, you
could honestly say each file was unique.  They do not look the same.
 
Imagine you took the 100 files and laid them side to side, so one end was
the least worn and rusted and the other was the most worn and rusted.
Imagine you compared them to a pile of 100 files just 5 years old.  What
would you see?  You would find that wear and rust formation on individual
files would be about the same.  Both sets would have files that were
"hardly worn and rusted," "moderately worn and rusted," and "heavily worn
and rusted." If anything, you might be able to say the older file pile had
more worn and rusted files than the younger file pile, but HOW could you
say WHICH individual file was actually older?  Mix the two file piles
together and pull out just the old ones.  See the problem?  You would know
the old file pile would have more worn files, but you cannot determine
WHICH specific file was worn and rusted because it was older.  In other
words, you couldn't tell the 10 yr olds from the 5 yr olds.
 
What you have discovered is it is USE and CARE which defines the wear
and rust on the files, NOT the age.  You cannot use file condition to
accurately determine the age of a file because so much depends on the
environment.  The brand of file, number of file grooves, exact composition
of the metal, file owner strength, care, even relative humidity ALL
influence final condition.  Sure, a number of the least worn would be
younger, just like a number of the most worn would be older, BUT YOU
CANNOT PREDICT WHICH SPECIFIC FILE WOULD BE OLDER BASED ON CONDITION.
 
Ferret teeth are the same.  You could look at thousands of teeth, do
complicated math and statistical manouvers, but you still end up with the
simple question, "which ferret jaw is older?" Unless you can learn the
exact history of each individual tooth, you can never control for genetic,
use and care variables (many right now are not understood well enough for
that type of control).  You can NEVER accurately predict the age of any
single ferret based on tooth wear or transparency.
 
Oh, sure, you can generate percentage predictions, like, "in 90% of the
cases, the ferret has a 50% chance of being 3, 4 or 5 years old."
Depending on the degree of variability of tooth wear and transparency in
the pile of ferrets, it could even be 2 OR 3 OR 4 OR 5 OR 6 years old.
These techniques are somewhat accurate in demographics, but is it accurate
with individuals?  "Bob, how old is my ferret?"  "Let's see... based on the
degree of tooth wear and transparency, there is a 90% chance that your
ferret has a 50% chance of being either 3, 4 or 5 years old." What the hell
kind of an answer is that?  That is the ONLY kind of an answer that Ed's
experiment could generate, providing his samples were large and random
enough, his method and technique beyond reproach, and his criteria used to
determine wear and transparency are valid.
 
That last requirement is the biggest problem of all.  For Ed to be able to
say "there is a 90% chance your ferret has a 50% chance of being 3, 4, or 5
years old," he MUST FIRST be able to define "wear" and "transparency" so
anyone else can effortlessly duplicate his efforts.  Such definitions are
difficult at best and impossible at worst.  This is such a major problem
that some fish and game wardens carry actual (or cast) jaws of deer and
elk so they can age by direct comparison instead of trying to use photos
or drawings.  Even with the actual jaws, there is STILL difficulty.  For
example, if wear looks part way between 4 and 5 yrs, which one do you
choose?  What if one part of the jaw looks 6 and the other part looks 4?
Would the guy next to you agree with your decision?  How about the other
1000 people?  Now, if we can't all agree EXACTLY on tooth wear and
transparency, how accurate are our observations?  Since age predictions are
dependent on *OPINION*, how accurate are they with 1000 people judging?
 
The problem with Ed's idea isn't because Ed is dumb. These are the types
of questions that concern a handful of scientists and I am fortunate to be
a zooarchaeologist where such questions have vast importance. Ed just
didn't know; quite understandable. The absolute bottom line is Ed's
experiment CANNOT produce the results he wishes, which is a way for people
to look at a picture and say "My ferret's teeth look like that, so it must
be 4 years old." The truth is, educated guesses of age, based on muscluar
condition, fur quality, teeth, behavior, joint stiffness, etc., give a
much better prediction of age than transparency and wear of teeth. And
THAT takes simple experience. It is not something you can learn by looking
at photos (or skulls), but only through handling a lot of ferrets. There
are no shortcuts for newbees. C'est la Vie. Get over it. Das ist alles.
 
And THAT is why I felt it was important to explain exactly why Ed's idea
would not work and why I urge people to not send dead ferret carcasses to
him.  All you and Ed would be doing is generating the equivelent of
scientific smoke, where, because light is blocked, you think substance
exists.  There is a real risk that should "data" be collected, it will
confuse the issue and throw a lot of people off for a long time, requiring
major efforts at reeducation (like the "Out of Egypt" ferret origin theory
which STILL plagues us).  Tooth transparency and wear will NOT give you an
accurate method of ageing.  I urge Ed to drop the "experiment," and I urge
each of you to not send dead ferrets to Ed for this purpose.
 
Bob C and 19 Mo' Who Wonder if Poofing Causes Shadows and Substance
[Posted in FML issue 2889]