Diane I. Hillmann wrote:
No Mac, the vocabularies don't assume anything of the kind. If you
check out some of the work we've done with the Deutsche
Nationalbibliothek (in the Content Type and Media Type vocabularies at:
http://metadataregistry.org/vocabulary/show/id/45.html and
http://metadataregistry.org/vocabulary/show/id/37.html). The URIs tie
the language versions together and allow the information to be displayed
in either language.
MARC (or any other such format's) tags can do the same.
For everyday use, URIs are much too cumbersome. If you look at these
two examples, they differ only by their numbers, 45 and 37. Within the
proper context, these numbers as such can mean the exact same thing
as the bulky URIs. Which, however, boil down to the same idea, only
just in an impracticable disguise: Incidentally, MARC 337 stands for the
same thing as that long ...37.html URI.
It is by now a time-tested fact that numbers and codes as shorthand are
the most useful "language" in cataloging. It is proved that humans can
communicate in numbers, but in URIs? For one thing, numbers are
truly international. Dewey was and is successful for the same reason.
But for another thing, equally important, language changes, terms become
obsolete or politically incorrect, numbers don't.
We should therefore never base our encoding on verbal tags (or URIs as
tags).
This was also the biggest mistake the DC people made. Their presumably
simple vocabulary was a set of words. Starting from there, implementers
thought they understood what the words meant, without looking at
the definitions. Inevitably, the metadata they produced was metagarbage,
as Roy Tennant had found out when he wrote "Metadata's bitter harvest":
http://www.libraryjournal.com/article/CA434443.html
This cannot happen with numbers and codes, for to understand them you
got to understand the definitions, and these are much longer than one
word per tag.
MARC is, however, more than a vocabulary, it is also a syntax. Elements
are tied together as subfields in longer, structured element compounds.
The sequence of elements in a record carries meaning as well. Some
of this, but certainly not all of this, probably goes too far and
imposes structure where it isn't helpful (e.g., where it was based on
obsolete card design).
B.Eversberg
