FRBRlogical Investigations (was Re: FRBR and classes)

I should have liked to produce a good post. This has not come about, but the
time is past in which I could improve it.

1.  Discussion of the concept of meaning is essential to the semantic
enterprise; but where problems have proven resistant to solution for several
thousand years, it is wise to avoid issues of metaphysics unnecessary to the
resolution of the issues at hand.

2. Work, Expression, Manifestation, and Item do not form a genus/species
hierarchy of #$Collections.  However, there are other forms of inheritance
that exist between related members of the different #$Collections.

3. In order to understand WEMI, it can be helpful to examine the concepts in
reverse from the particular to the general - iMEW.

4. It can be easier for people from a computing background to understand
iMEW by considering the different concepts as they might be apply to a
replicated version control system.

5.  It can also be helpful to consider iMEW in terms of relative identity;
when are two things the same Manifestation, etc.  This is not unlike Martha
Yee's concept of Near Equivalents - when are two items similar enough that
either may satisfy a specific patrons request?

6. Items have an undeniable physical existence; they consist of specific
particles having a specific extent in space-time.
For example, the copy of *Philosophical Investigations* that is sitting on
my desk to my right;  the copy of umbel_reference_concepts_consolidated.n3 that
is stored magnetically on the hard drive of my laptop.

7. In the official FRBR report,  Manifestations are also considered to
physically exist.  If this is taken literally, one must make an ontological
commitment to treating a Manifestations as a Substance, made up of the
discontinuous regions of  space-time occupied by the Items that make up that
Manifestation. This is completely valid mereology, but some of the
implications are distracting; for example, the that a Manifestation has a
weight which changes every time a new copy is created or destroyed.

8. It may be simpler to instead treat Manifestations as being abstract;
 collections of Items whose characteristics other than those of physical
composition are default identical.  For example, all copies of the pape 3rd
edition paperback G.E.M Anscombe's translation of *Philosophical
Investigations *; all copies of the n3 encoding of version 1.00 of
umbel_reference_concepts_consolidated
.

9. It might be argued that if two items are physically indistinguishable,
they must be of the same manifestation- for example, if two computer files
are made up of  identical bit strings.  However, this criteria, whilst
necessary, is not sufficient.  For example, it is possible for identical bit
sequences to be derived completely  independently- for example, different
scans of different blank pages; calibration output from sensors on different
instruments.   These examples suggest that it is problematic to strictly
identify bitstrings with Manifestations (or Expressions following
 Wickett/Renear).  It is nevertheless important, especially when building
systems for long term digital preservation,  to be able to identify these
bit strings.

10. There are other possible equivalence classes of Items other than
Manifestation. These equivalence classes may facilitate different
applications within the bibliographic universe.  For example, for purposes
of ILL,  one may not care which of several copies of a particular
manifestation owned by a particular institution are used to service the
request.  One might describe these as  "Has-ings".

11. Expressions are unquestionably abstract.  Just as Manifestations can be
treated as having fixed default physical characteristics, Expressions have
fixed default intellectual characteristics - that is, the propositional
content of two Manifestations that are of the same Expression is identical.
For example, all manifestations of the 3rd edition of Anscome's translation
; an XML encoding of  version 1.00 of umbel_reference_concepts_consolidated ;
a gzip compressed version of the same.

12. Compare a  TIFF encoded image with a PNG encoded version of the same
image. Are there two Expressions or one?

13. Compare a TIFF encoded image with a JPEG encoded version of the same
image. Are there two Expressions or one?

14. In a Version Control model, what might correspond the concept of a Work?
Consider the contents of a file named by  a specific path in a version
controlled repository-  consider the entity named by
http://umbel.googlecode.com/svn/trunk/umbel_reference_concepts_consolidated.n3.

15. If a new proposition is added to the entity named by this identifier,
and a new version committed to the repository, a new Expression has been
created.  However, the name has not changed- if not a Work what is it that
these two expressions are versions of?

16. By the same reasoning, all three editions of Anscome's translation of
Philosophical Investigations are expressions of the same Work.

17. But what of Wittgenstein's original, German, text?  In what relationship
does it stand to Anscome's translation?  Is the relationship solely one
between works, or are both subsumed within  what Carlyle termed a
Superwork?

18. As Quine showed, there is no fundamental difference between classes and
properties.  Whether one chooses within ones ontology to treat something as
property or class is a matter of convenience and not of correctness. It is
important to understand this; it is also important to understand that
different communities may traditionally think about the subject matter in
terms of one or the other, and that it is these concerns that are primary.

19.  That the logics used OWL cannot readily model domains requiring
non-monotonic reasoning is a weakness of OWL.

20. That inexperienced ontologists cannot  readily model some domains in OWL
may be a weakness in the ontologist.

21. Anything that requires the use  Restriction Classes may fairly be blamed
OWL.