Application of digital humanities in the analysis of Chines poetic discourse model
Résumé
The recitation has been a skill to tokenize verses to craft poetic language. In my Ph.D. dissertation defended last year, I focus on The Book of Odes (Shijing 詩經). I defended a digital method as a simulation of metalinguistic activities for historical readings and chanting of the literary masterwork. The question was how do we handle literary rhythms which could be engendered during corpus’ reading and build some poetic discourse models. In order to trace back rhythmic phenomenons in the text, I proposed to pay attention to repetitious elements:
1. string of n-gram,
2. combination of n-gram,
3. variation between function word and content word, 4. variation between monosyllable and bi-syllable.
A complete automatic survey allows to observe dynamic patterns in a targeted string by binary contrast of frequency for:
1. continuous characters: rare/frequent,
2. combinatorics type of characters: rare/frequent,
3. variation type of word categories: empty (function word)/plain (content word), 4. variation type of phonetic sounds: short (monosyllable)/long(bi-syllable).
Relying on these principles of analysis, the research proposes an experimental method of reading which sets forth two aspects: the frequent pattern that creates the dominant rhythm and the rare pattern that constitutes the mean ideas. On the basis of such inquiry, we can also evaluate the poetic language distance between authors, anthologies, dynasties, regions...etc.