Abstract
No annotation guidelines concerning substandard Latin are presently available. This paper describes an annotation style of substandard Latin that supplements the method designed for standard Latin by the Perseus Latin Dependency Treebank and the Index Thomisticus Treebank. Each word of the corpus can be assigned only one morphological analysis. In our system, the analysis can be either functional or formal. Functional analysis is applied when a form is language-evolutionarily deducible from the corresponding standard Latin form used in the same (semantico )syntactic function (e.g. solidus pro solidos ‘gold coins’ as a direct object: analysis “accusative”). Formal analysis applies when no connection to the functionally required classical form exists (e.g. heredibus pro heredes ‘heirs’ as a subject: analysis “ablative” or “dative”). When running queries on the corpus, the formally analysed forms can be isolated, and percentages of standard and substandard forms can be counted. In addition, further principles concerning syntax and specific morphological issues are introduced.
Original language | English |
---|---|
Journal | Journal for Language Technology and Computational Linguistics |
Volume | 26 |
Issue number | 2 |
Pages (from-to) | 103-114 |
Number of pages | 12 |
Publication status | Published - Feb 2012 |
MoE publication type | A1 Journal article-refereed |
Fields of Science
- 6121 Languages
- 113 Computer and information sciences