W.S. Merwin on the progressive nature of translation from his forward to Selected Translations 1968-1978.
But if we take a single word of any language and try to find an exact equivalent in another, even if the second language is closely akin to the first, we have to admit that it cannot be done. A single primary denotation may be shared; but the constellation of secondary meanings, the moving rings of associations, the etymological echoes, the sound and its own levels of association, do not have an equivalent because they cannot. If we put two words of a language together and repeat the attempt, the failure is still more obvious. Yet if we continue, we reach a point where some sequence of the first language conveys a dynamic unit, a rudiment of form. Some energy of the first language begins to be manifest, not only in single words but in the charge of their relationship. The surprising thing is that at this point the hope of translation does not fade altogether, but begins to emerge.
The poet after sketching out the limit where energy is manifest and where hope emerges continues in workman-like assertion.
Not that these rudiments of form in the original language can be matched — any more than individual words could be — with exact equivalents in another. But the imaginative force which they embody, and which single words embody in context, may suggest convocations of words in another language that will have a comparable thrust and sense.
Take a single word. Trace from there the rudiments of form. Find the comparable convocations.
And so for day 1101