"For there is in fact no direct communication with God, the Creator God. One cannot speak to Him, that is pray to Him. 'The pot doesn't know the woman who made it. The field of millet doesn't know the sower. The cloth forgets its weaver'." Jack Goody, The Interface Between Written and the Oral (CUP: Cambridge, 1987), 153. (Goody is apparently quoting Alfred Erbs, Approche de la religion des Birifior (Paris, 1975), 9.)

This remark struck me with great precision. Here, well in advance of Richards or Barthes, an African tribe announced the death of the author. But in truth, if we were to research this matter carefully there can be few cultures in which this point would not be found to be a truism; but one that cohabits with the strictest intentionalism, since in regard to speech we must presume a living, thinking, person like ourselves (it is only courteous to do so), otherwise one slides into solipsism. With writing, of course, it is different, and there is a choice to be made:

a) This is a stand-in for the face to face speech, so it is essential, again out of politeness, to invoke the intention of the writer. Of this kind are most letters, bills, and such similar documents; but it should benoted that with time such documents may, as far as a reader is concerned, be allowed to slip into the second category,

b) Writings which are cast in the written form, and are meant to be read. Just as one might write someone a poem, although they live in the room above you. Here writer has implicitly accepted the possible drift.– The poet who writes for posterity must be resigned to the irrelevance of his intention.

All the fuss about intent is due to an oddity of reader response. When we go through a poem, or any text, we are very likely to think of it as a substitute for speech; it is a form of arrogance, as if Shakespeare were speaking to me, hard something particular to tell you. But when the audience is unpredictable (in name, or merely in mood) no intention can be broad enough to circumscribe the flux, even when we are able, via historical research to discover something of the wishes of the author, and in such a case it is important to to see that evidence of that kind is only adequate to expand the reference, not constrict it. I can show that author A's text a was alluding to author B's b to make the point Ac, but I cannot with the same material rule out the possibility that a also makes the point 'Ac (not Ac). – Ambiguity, then, can be seen as a feature of writing, especially, though present in speech, but controlled there by tone, pitch, gesture, facial expression, each of which exerts a truly enormous semantic pressure; so great in fact that it is not even necessary to know the language of an angry or an unhappy man. Writing is, by comparison, quite undetermined, except in those cases when, out of politeness, as I have already noted, we pretend that many of the contextual clues are present. The truth is that poems are valuable, if they are, because of their elusive profundity, suggesting but never delivering extreme significance. It is only the timid who try to turn them into business letters.

There is a sense in which the writer has something to say, but by putting it into written form, with so indefinite an application, he has committed it to the logic of written language. A potter may design an orange glaze, but there is a tint of green as a result of an irregularity in the kiln, or, even more clearly, what comes out as he planned, ages unexpectedly and uncontrollably to an even more pleasing colour. (To proceed with their work, what faith the cleaners of the Sistine Chapel must have had.) So the writer, like the sower of millett, resigns control. – But it is essential to add that for every reader there are limits: it is almost as if the intention of the author must be replaced by that of the poem or the drift becomes intolerable. When we read we must be honest and say that it produced in us, unambiguously (though containing many ambiguous precessions), an affective response, and not some other response.

Then how is it that some poets more reliably turn out work that we value? – Some play into the hands of the ambiguity. They start their poems off in a state of rolling amnesia, they sense productive avenues of development in the words just written and push their poems in that direction. This produces texts that seem strangely impersonal. It is not the only means of production. Time also works. It may take a thousand years to clean a poem that is not scoured before birth, and by then there may be nobody to read it. Many of the most beautiful things in English are fragments of popular and even nursery verse that are now sinking into and on the verge of oblivion. The pot, the millett, the cloth are themselves sometimes completely forgotten.