background image
Here the ref-attribute indicates which parts of the emblem belong together. An encoder
recording the sequence of emblems would ignore the actual location where the motto or pictura
are found and do the following:
<emblem id="no1">
<motto/>
<pictura/>
</emblem>
<emblem id="no2">
<motto/>
<pictura/>
</emblem>
...
Although encoding the page structure would have some advantages, the emblem scholar is
accustomed to focusing on the emblem structure rather than the page structure.The encoding of
the emblem structure facilitates collecting the various emblem parts together, inscriptio, picturae
and subscriptio, without being forced to encode the relations between them when they occur in
different sections of a book. It suits the need of libraries, however, to digitize and encode the entire
emblem book according to the page sequence, because a sequence of emblems cannot be includ-
ed in the library catalogue. Formally speaking, a library catalogue contains books, not individual
emblems. Above and beyond this essential point, the encoding of the book structure allows for
approaches that may be of little or no interest for the emblem scholar, but is useful for other
researchers, considerations which a library must satisfy as well.This is why the Herzog August
Bibliothek,Wolfenbüttel, scans entire emblem books rather than a choice of pages containing
emblems.This permits the cataloging of books as electronic editions in the main library catalogue.
Additionally, these digitized books are enriched by emblem related metadata.
On the one hand, an emblem exchange format does not necessarily require the encoding of
the entire book and, in fact, few emblem digitizing projects do so.And it is also not necessary for
the metadata exchange format to comply with the individual format an individual institution or
project may employ. Illinois,
5
for instance, uses an extended Dublin Core format, Utrecht
6
and
Wolfenbüttel
7
TEI,
8
though in a slightly different way. Wolfenbüttel encodes the entire book,
whereas Utrecht focuses on the emblems. All of the projects have good reasons to employ their
chosen method of encoding, insofar as their encoding serves different, project-related goals. On
the other hand, it is fairly easy to extract emblem specific elements, for instance from book-ori-
ented documents, to set up entirely new databases, provided that some provisions are made and
all of the relevant information is encoded in the source document.Thus, it probably does not
make much sense to try to enforce a specific format for all possible projects in which emblems
may play a role, but it will suffice to agree on a set of categories, or schema structure, that allows
for data exchange and establishing sophisticated emblem databases, a schema into which source
data of any origin can be transferred.
The question of mandatory and optional data fields should also be raised here. Not all of the
emblem projects will encode the same data. Some may have the means to transcribe subscriptiones
in full, some may limit themselves to the motto and the description of the pictura.While Wolfen-
büttel encoded the subscriptio for 10 emblem books, Illinois did not encode subscriptions at all.
To produce consistent data at least some core elements must be agreed upon.The necessity of
transcribing the motto would seem to be a minimal requirement.The description of the pictura
is welcome, but one can debate whether it should be mandatory. Subscriptions must be option-
al since only a few of the projects will have the means to transcribe them in full.
5
http://images.library.uiuc.edu/
projects/emblems/.
6
http://emblems.let.uu.nl/
emblems/html/index.html.
7
http://www.hab.de/forschung/
projekte/emblematica.htm.
8
http://www.tei-c.org.
93
DC_Emblemsbook_180204 19.02.2004 11:26 Uhr Seite 93