I'm a cataloger who has been following this discussion with interest,
but not necessarily understanding all of it. I'll try to add what I
can regarding the rules for constructing LCSH headings.

> My understanding is that Education--England--Finance *is* authorized,
> because Education--Finance is and England is a free-floating
> geographic subdivision.  Because it's also an authorized heading,
> "Education--England--Finance" is, in fact, an authority.  The problem
> is that free-floating subdivisions cause an almost infinite number of
> permutations, so there aren't LCCNs issued for them.

Ross is essentially correct. Education is an authorized subject term
that can be subdivided geographically. Finance is a free-floating
subdivision that is authorized for use under subject terms that
conform to parameters given in the scope notes in its authority record
(680 fields), but it cannot be subdivided geographically. England is
an authorized geographic subject term that can be added to any heading
that can be subdivided geographically. Thus, Education -- England --
Finance is a valid LCSH heading, whereas Education -- Finance --
England would not be. This is wonky, and it's stuff like this that
makes LCSH so unwieldy and difficult to validate, even for humans who
actually have the capacity to learn and adjust to all of the various
inconsistencies.

I don't know how relevant it is to this particular discussion, but
going forward I'm not sure how important it is to validate LCSH
headings. I really appreciate developers who seek to preserve the
semantic relationships present in the headings as much as possible; I
believe many of them have value. But aren't there ways to
preserve/extract that value without getting too bogged down in the
inconsistent left-to-right structure of the existing headings?

I hope this helps, at least a little bit. I'd be happy to answer
additional questions.

Shirley

Shirley Lincicum
Frustrated Cataloger