5. The Skaldic Database 6. Database structure 1. Concepts
This is not currently part of the peer-reviewed material of the project. Do not cite as a research publication.
The structure of the database is generalised as far as possible to include skaldic, Old Norse and other material from other databases that share the same broad database structure but is distinguished at different levels.
The basic structure of the text in the database is a tree of one-to-may relationships. These and other phenomena are linked using a relational data model:
- Author ('skalds'): an identifiable individual responsible for composing one or more textual outputs. Poets (skalds) are distinguished by the 'skaldic' (boolean) column. May also be used for convenient groupings of anonymous texts (sagas, eddic poems, etc.). Anonymous groupings are distinguished by the 'person' (boolean) column.
- Text ('text'): an identifiable separate textual work, such as a saga, poem, compilation or group of stanzas, with the same author(s) throughout. The 'type' (prose/poetry/other) column distinguishes poetic texts from prose works. Compilations are distinguished by the 'compln' (y/n) column. Texts within compilations are identified by a link to an entry in the same table ('text_id') to the work that the particular text forms part of.
- Segment ('verses'): a first-level division of a text: a chapter, stanza, text segment or other division. There is no nesting within this table; if a larger division is used (e.g. chapter) then this should be used to divide the whole text without mixing with smaller-level divisions (stanzas or segments). If smaller segments are used, these should be used throughout with headings occupying a single segment. The type of division is indicated by the 'inclusion' column (full [stanza], refonly [stanza], xref [stanza], eddic [stanza], runic, prose [teaching text segment], chapter, heading, segment, picture, other). This table can include a full text and translation of the segment but these are analysed in more detail below.
- Line ('line'): a second-level division of a text. The type is defined by the segment to which it is linked. For stanzas, this is a line in the stanza but may also be accompanying prose before or after the stanza (defined in those cases by the 'type' column); for runic inscriptions it is a physical side of the inscription; for other segment types this is generally redundant but included to complete the structure; future applications may be for second-level divisions of texts such as paragraphs.
- Word ('word'): a linguistic word or word element. Compound words are links of words in this table that are stored in a separate table ('compound'). The word table links different text versions that could also be stored in a separate table occupying a sixth level in the tree. Three versions of each word are stored in this table: the main text, a reordered text if necessary and a translation into English. For each there are columns for the word, preceeding and following punctuation and ordering.
Other tables in the structure include
- Collections, manuscripts, references, images
- Orderings, versions
- Notes, variants
- Bibliography (works, journals, authors)
- Lexicon, kennings, compounds
- Interface information tables