Reference:

Oskar Kohonen, Sakari Katajamäki, and Timo Honkela. In search for volta: Statistical analysis of word patterns in Shakespeare's sonnets. In Ann Russell, Timo Honkela, Krista Lagus, and Matti Pöllä, editors, Proceedings of AMKLC'05, International Symposium on Adaptive Models of Knowledge, Language and Cognition, pages 44–47, Espoo, Finland, June 2005. Helsinki University of Technology, Laboratory of Computer and Information Science.

Abstract:

The sonnet is one of the most canonical modes of poetry in Western literature. The English or Shakespearean sonnet falls in to three quatrains in iambic pentameter, with a turn at the end of the line 12 and a concluding couplet often of a summary or epigrammatic character. The turn normally is both semantic and stylistic for the rhyme scheme abab cdcd efef of the first part of the poem changes to a form gg in the closing couplet. We analyze if the semantic turn, or volta, can be found with statistical analysis of word distributions, using the Self-Organizing Map for exploration and visualization. The Self-Organizing Map is a neural network architecture based on unsupervised learning. We conclude that our methods can be useful for finding semantic and stylistic turns that can then be studied in detail using other methods. We propose extensions to our methods for other literary analysis.

Suggested BibTeX entry:

@inproceedings{ShakespeareAMKLC05,
    address = {Espoo, Finland},
    author = {Oskar Kohonen and Sakari Katajam{\"a}ki and Timo Honkela},
    booktitle = {Proceedings of {AMKLC'05}, International Symposium on Adaptive Models of Knowledge, Language and Cognition},
    editor = {Ann Russell and Timo Honkela and Krista Lagus and Matti P\"{o}ll\"{a}},
    month = {June},
    pages = {44-47},
    publisher = {Helsinki University of Technology, Laboratory of Computer and Information Science},
    title = {In search for volta: {S}tatistical analysis of word patterns in {S}hakespeare's sonnets},
    year = {2005},
}