News:Previewing the previews
Previewing the previews (*selgā March 2013)
Much of the work that has been lurking behind the scenes has been an exercise in preparation for improvements to the software (SMW) that have not yet been implemented in official versions of the product. One such improvement relates to the use of so-called ‘subobjects’: rather than having the page dictate the base unit for semantic annotation, so that semantic differentiation is primarily determined by the use of different properties, it will allow us to associate smaller chunks of information together. These smaller datasets then remain tied to the entry for which they are written but are otherwise treated as independent units in data querying. While this feature is already in place, the essential point is that the new software changes will make it a better fit for use in templates.(1)n. 1 If you are familiar with SMW and wondering about this, I’m referring here to better support for value arrays using delimiters.
Now do not worry if this pithy description manages to do little more than to raise your eyebrows, or bring them together in a confused frown. What this means in practice and why this is useful can be easily demonstrated. As you can see on the page for the composite Middle Irish text of Cath Maige Mucrama (‘The battle of Mag Mucrama’) or that of Aided Diarmata meic Cerbaill I, a separate tab is reserved for a summary of the text, subdivided into a number of episodes or at least, what appear to be relatively free-standing narrative units. In addition to a descriptive subheading and a synopsis, each of these can carry with it a variety of metadata, such as poems cited, ‘people’ (by which is actually meant any personified being), placenames, ‘events’ and other, miscellaneous keywords. A general section for comments is intended for more metadescriptive aspects, such as possible relationships with other texts and scholarly coverage.
So why is this worth having, apart from the convenience of layout? Whenever the corpus of available information on texts is queried, whether in a preformatted query or through a custom search facility where you (the user) can select or enter keywords, the use of subobjects will enable us to give you much more pertinent details than is possible using a generic page-based query. To get the context, you should always consult the main entry, i.e. the source page itself rather than a derivative item in a query. In the initial stages, however, when you have been given a large number of potentially relevant links, you do not want to have to do so for each one of them. In other words, it should be possible to retrieve better previews of what you might, and might not, be looking for.
Because this change in data structure comes with new requirements for querying, a new search facility is being developed. You can already try it here and use a keyword such as Áed to get a basic idea of what this means for searching the catalogue.
Another reason for adopting this section-based approach is that it allows us to refine our information about manuscript witnesses and coverage in secondary literature in a format which is easy to understand and use; especially when the manuscript texts are incomplete or far from uniform and when a published scholarly discussion centres only on a discrete part of a text. See, for instance, the work in progress for the legal tract Bretha nemed dédenach, where partial translations are noted where relevant.(2)n. 2 Like Bretha nemed toísech, this text is still awaiting a full translation. Finally, there are also advantages to be gained for our workflow. Because this catalogue is a progressive effort, having at least a framework to fall back on will allow editors to pitch in and home in on small parts that they would like to focus their attention and energy on.
There is much more that can be said about this, but for now one final remark should suffice. There are cases when it is appropriate to extend the hierarchy by dedicating a separate entry/page to an individual text section. This may be true, for instance, when poems or other parts of a narrative are also independently attested in the manuscripts; when it could be argued that subnarratives in a larger textual framework merit separate attention (e.g. some of the stories in Bretha nemed dédenach); or simply when multiple pages are required to cover an extensive text or compilation (e.g. Táin bó Cúailnge). We have usually indicated this with a link to the page in question (such as » entry or a link to a poem included).
To be continued...
