Larry W wrote:

....


I admit that I made that statement from memory of Richard Rhodes book ...


I've skimmed it; don't own it. It's massive enough it may take some
digging to find the anecdote... Don't recall if it's indexed well
enough to help or not.

The only case even remotely close at Y-12 I can think of I just don't
think had any connection to NM having pointed out the operation as
troublesome. May have to be careful about how much I say on some of
these; some amazingly enough have details that are (or at least were
when I left OR around 2000 and DOE and the Q behind) still classified
owing to specifics of some of the materials involved (in the separation
process end, not the nuclear material itself) and at least last review
we did for DOE on operations that looked at some of these there had been
no partial declassification on any of the reports--if there was anything
in the report that was classified, the whole document was classified and
a derivative classifier would have to judge what could be released;
the user of the information isn't allowed to decide that just because
they know something else isn't classified per se if it's in the
classified report it can't be released simply on that knowledge. For
one thing, there are combinations of information that are both
unclassified yet the two in proximity to each other can create a
derivative work that may be classified because w/o that direct
connection that the two pieces of information are somehow useful
together isn't obvious. The rules are arcane and full of such
stuf....well, I digress (nothing unusual... )

You've definitely whetted my curiosity, though...

--