When a voice other than the poet's reads a poem why does the new voice fail to ruin the verse's music and meaning? Ezra Pound wrote to Mary Barnard in 1934: 'Thing is to cut a shape in time. Sounds that stop the flow, and durations either of syllable, or implied between them, "forced onto the voice" of the reader by the nature of the "verse."' The fundamental concern here is with cadence and rhythm. Pound asserts that real verse shapes the voice. Are intonational implications governed by the same principle? Eric Griffiths comments on Pound's desire to indicate typographically 'the reader's intonation' and annoyance at the difficulty of indicating the 'exact tone and rhythm'. His important study of Pound's predecessors - Tennyson, Browning, Hopkins and others - is concerned with how poets are at pains to shape a voice which may be heard in the writing and printing of verse. They are also involved in a relationship over which they cannot have entire control: 'Print does not give conclusive evidence of a voice; this raises doubts about what we hear in writing but also gives an essential pleasure of reading, for we meet the demands a text makes on us for our voices, we are engaged in an activity of imagination which is delicately and thoroughly reciprocal.' Pound's masterful '"forced onto the voice'" has become Griffiths' 'we meet the demand a text makes'. His phrase has its own ambiguity: do we merely encounter the demand, or do we ...