Hi Amanda, and all,
Amanda Gailey wrote:
> (...) Have you considered taking this one step further and even
> removing the editorial markup to its own file? In the case of a
> Spenser edition we're working on, two competing editorial interests
> in the texts (literary/logical and physical bibliographic) seem to
> demand that a proper standoff system would involve a clean
> transcription as the base file to avoid competing hierarchies. Not
> all editions are interested in noting both of these views of a text,
> and so could, of course, include the editorial markup in the base
> file. But that would from the outset require that future standoff
> files not hierarchically conflict with whatever the base file's
> annotations say about the text. Is that a problem for you, or do you
> view that as an edition's prerogative?
I believe basic editorial markup should be stored with and within the
edited text, for a number of reasons:
(1) There is no such thing as a 'clean' text or a text without markup.
You can't really conceive of text without line breaks, whitelines,
headings, titles, notes and note numbers, page numbers, signatures,
dedications, etc. Even a so-called clean text needs pseudo-markup to
indicate some form of textual structure and some indication of
(2) If there were such a thing as clean text, you couldn't do anything
with it. A researcher who wants to study a book of poems that I edited
should be given a division into poems and stanzas, even if only to be
able to say I did that wrongly. I must provide metadata - I could decide
not to give a tei header, but then I must use some other means to say
what my source was and who the author of these poems is. If I feel I
can't reliably make decisions about poems and stanzas I should describe
a physical structure, perhaps in terms of folio numbers or even
locations of fragments on pages. But a 'clean text', if it is anything,
is just string soup.
(3) In most cases, it is fairly easy to create editorial markup that
represents both the logical and the physical structure. I don't think
there's anything wrong with the usual representation of pages by
milestone elements. The occasional line split between clauses in a play,
or a quotation that runs over multiple paragraphs, is certainly
inelegant but usually not a major problem.
(4) If for some reason you feel two competing hierarchies must be given
equal weight, and don't want to favour one over the other, you still
have two options to keep markup with the text:
= use milestone elements only, or
= switch to another markup language, such as TexMECS.
True, both are ways of handicapping yourself and others, but so is a
text that doesn't include editorial markup.
(5) There is no reason why an editorially defined hierarchy should make
it impossible to annotate phenomena that belong to another hierarchy. If
the annotation tool can create annotations by pointing at arbitrary
start and end locations (as our EDITOR tool can), annotations can
overlap structural boundaries as much as you like. Display of these
annotations can then be integrated into the display of the edition as
based on the editorial markup. (I have a paper on this forthcoming in
the Jahrbuch für Computerphilologie).
(6) Text without markup cannot be preserved. This is really a special
case of (2), highlighted because of its importance. A text without
markup needs functioning software in order to be understood, probably in
a database system somewhere, that no-one has a reason to believe will
run twenty years from now. At the very least, one will need a document
that says where markup should be inserted in the text. And probably yet
another document, to say which text should be combined with which markup
So, to finally answer your question: yes, I considered this, though I
didn't know how strongly I felt about it before writing this reply. I'm
not a lawgiver, but I'd say it is not so much a prerogative of the
edition to propose a base structure for the edited text, as it is a duty.