The Balisage 2014 program is up, and I’m on it, Thursday Aug 7 at 9am. See http://www.balisage.net/2014/Program.html.
The LMNL project I started over a decade ago with Jeni Tennison (or the “rump version” thereof, since everyone else has moved onto other things) is still offering me intellectual rewards. It turns out a range model like LMNL’s is very useful for addressing the question of what one is describing when one marks up a document, because it forces no predisposition to one hierarchical rendition or another (or any) before the document description is mature. So the “ontology” of the markup can remain much looser, to develop iteratively.
This means that LMNL is useful not only for models of texts that have honest MCH (multiple concurrent hierarchies), such as models of poetry that present both verse and grammatical (sentence/phrase) hierarchies together, but also for examining texts that show structural anomalies — the kind of thing that makes us wonder whether and where the hierarchies and “containment” are even to be found.
Since Mary Shelley’s Frankenstein shows such a structural anomaly, it makes an interesting case study. Formally, it’s a hybrid between an epistolary novel (in its framing narrative) and a more conventional first-person narrative (with a long embedded narrative in the center). But overlap between these two structures (at least in almost every printed edition) gives pause: it frustrates a clean and simple representation in a hierarchical model such as a conventionally encoded XML version, and raises questions about the coherence of its representation. Is this a feature (of a gothic horror romance), or a bug?
Interestingly, LMNL offers a way to demarcate the parts of the book without organizing them into a single hierarchy or any hierarchy at all. This turns out to be useful for asking questions about this work and hence about the idea of the OHCO in general, as applied to texts that were not already committed to hierarchical forms in their composition.
Or were they? LMNL helps us ask.