2.5.1. Editing the corpus
Tarrin Wills 2012, ‘Editing the corpus’ in Diana Whaley (ed.), Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035. Skaldic Poetry of the Scandinavian Middle Ages 1. Turnhout: Brepols, p. xxxvii.
Having created a resource to manage data related to skaldic poetry, and providing a mechanism for altering that data at a distance, it has been relatively straightforward to extend that resource to managing the content of the edition itself. A significant issue in the editing process has been that of maintaining a central, authoritative version of the edition while different parts are being produced, entered, corrected and updated by different people involved in the project. Some of the information which was to be included in the print edition was already contained in the database, including classification information (both for the new edition and the Skj arrangement), manuscripts and locations, and some edition information, including the Skj page references. Other information forms a natural extension to the database. Variant readings, for example, are linked electronically to the manuscripts, eliminating errors in reproducing sigla and ordering manuscripts. The prose order is constructed from a reordering of the verse order; errors that may result from the duplication of this information are therefore reduced.
Other mechanisms for linking data provide extensions for the electronic edition: notes linked to words can appear with those words, and lemmata for the variant apparatus are extracted from the base text, so that any changes to the base text are automatically reflected in the textual apparatus. From early on editors were asked to mark up their editions with brackets to indicate the structure of the kennings, in both the prose word order and translation. For example, the editor of Gizsv Lv1/7I would mark up the following kenning thus in the prose order: í {éli Ála} and present the translation as ‘in {the blizzard of Áli <legendary king>} [battle]’. These structural brackets do not appear in the printed text. When entered into the database, kennings can then be indexed automatically by their referents, so that, for example, all kennings for ‘battle’ can be viewed. Likewise, words and lines have also been indexed, along with other features that can be extracted from the edition, such as heiti.