Warring States Project
Stylistics

People may use rare words for an archaic effect, or long sentences for a sonorous one. These devices are part of style. And the way the language of a piece of writing holds together, its articulation, is also an aspect of style: the aspect most independent of content. It will be obvious that rarity of words can be measured, and that length of sentences can be measured. It is a little less obvious whether the degree of articulation of a piece of writing can be measured, and if so, how.

It turns out that articulation can be measured. We first define style in articulation terms, as the way a piece of prose hangs together. We then create a tool capable of registering the degree of stylistic similarity between two samples of text. And what happens when we have made that measurement? We must interpret the result. This requires some subtlety and some literary acquaintance. It turns out that dissimilarity in style need not imply difference of authorship; writers may vary from their own norm if they are agitated, or depressed, or if they are temporarily influenced by the style of some other text. But along with any other information that may be available, stylistic difference data are analytically useful. Like every other observational result, style analysis results will in the end be tested against the investigator's literary sensibility. But it is an advantage if the stylistic analysis itself had first been reached by an objective method that others can independently duplicate, and is not simply the shadow of the investigator's literary sensibility.

The tool we have devised for this purpose is universal in application, but it has so far been implemented only for English and Chinese; a Greek extension is currently underway. Work in this direction is currently going forward under the Project's direction, with a focus chiefly on certain classical Chinese texts. The results of those investigations will be made public as appropriate. For a past result, shared with the public at Leiden in 2003, see the lecture entitled Lord Shang Revisited.

However sophisticated its tools, history cannot be done in one corner:

16 Sept 2009 / Contact the Project Director / To Project Home Page