The writing of a poem is like a child throwing stones into a mineshaft. You compose first, then you listen for the reverberation.

I've not been a prolific poet, and it always seemed to me to be a bad idea to feel that you had to produce in order to get... credits. Production of a collection of poems every three years or every five years, or whatever, looks good, on paper. But it might not be good; it might be writing on a kind of automatic pilot.

In rap, as in most popular lyrics, a very low standard is set for rhyme; but this was not always the case with popular music.

An aria in an opera - Handel's 'Ombra mai fu,' for example - gets along with an incredibly small number of words and ideas and a large amount of variation and repetition. That's the beauty of it. It's not taxing to the listener's intelligence because if you haven't heard it the first time round, it'll come around again.

Those who actually set out to see the fall of a city or those who choose to go to a front line, are obviously asking themselves to what extent they are cowards. But the tests they set themselves - there is a dead body, can you bear to look at it? - are nothing in comparison with the tests that are sprung on them. It is not the obvious tests that matter (do you go to pieces in a mortar attack?) but the unexpected ones (here is a man on the run, seeking your help - can you face him honestly?).

Considering the wealth of poetic drama that has come down to us from the Elizabethan and Jacobean periods, it is surprising that so little of any value has been added since.

For poets today or in any age, the choice is not between freedom on the one hand and abstruse French forms on the other. The choice is between the nullity and vanity of our first efforts, and the developing of a sense of idiom, form, structure, metre, rhythm, line - all the fundamental characteristics of this verbal art.