Using corpus linguistics to explore the language of poetry: a stylometric approach to Yeats' poems

This chapter demonstrates the value of combining corpus stylistic techniques with stylometric methods of analysis in order to generate stylistic insights into the language of poetry. In particular, we analyse a specially constructed corpus of William Butler Yeats’ poetry containing 307 of his published poems to determine whether there is textual evidence of a change in Yeats’ style over time. Using cluster analysis, principal components analysis and keyword analysis, we are able to provide support for some literary critical claims as to the distinct periods of Yeats’ writing career, as well as linguistic insight into what made these periods distinctive. The chapter provides an explanation of each method of analysis employed, making it a practical guide for newcomers to multivariate analysis and corpus stylistics.