Editing a Difficult Text with the TEI

The Case of The Cantos

Abstract

We will herein identify and describe several kinds of difficulties encountered while creating an XML-TEI encoding of The Cantos of Ezra Pound. This long poem, notoriously difficult for the lay reader, presents equally complex puzzles for its editor to solve.

One is linked with the openness of its form. An exemplary Modernist work, The Cantos displays virtuosic tendencies to question the received boundaries of the poetic form: its integration of peritextual elements, of prose quotations of various lengths, its unsystematic treatment of these “non-poetic” fragments don’t leave untouched the traditional ontologies that underlie the TEI’s standard representation of a poem. This questioning is pushed further, in the late sections of the poem, with the proliferation of non-Western writing systems (especially Chinese logographs), which not only contribute to dismantling the traditional order of reading but create (at the line, stanza and page levels) a new visual organization that must be accounted for with sui generis specifications of the TEI vocabulary. A third series of difficulties emerges with the delimitation of the numerous explicit quotations contained in the poem, whose extremely irregular typographical marking prevents any hope for automated annotation.

In all these cases, we argue, the encoding involves an interpretive dimension, which must be made explicit in order to solve various doubts and hesitations and justify case-by-case decisions. Ultimately, we claim that it is only through an exhaustive, philologically-informed genetic enquiry into the dossier of The Cantos that a reasonably accurate digital edition will be produced.