"Ermes Marana dreamed of a literature made
entirely of apocrypha, of false attributions, of
imitations and counterfeits and pastiches"¹.

Gregory Rabassa, in a short essay marked by refined sense of humor, gives some useful indications useful for sweeping away any pretence to equivalence in a light but clear-cut way within the framework of all languages including, he says, artificial languages like mathematic languages. Since we are used to thinking of an expression like 2 = 2 as correct, we do not realize that "the second 2 is obviously a hair younger than the first and therefore not its equal"². But here Rabassa is evidently joking, attaching importance to the age of a mathematic sign, almost decontextualizing the number to insert it into a less scientific, more human world.
  Nowadays when the equal sign is placed between two numbers mathematicians are more careful than some time ago (although, fortunately, accountants and book-keepers continue their business as usual), and prefer the word "approximation" to "equivalence", a stronger reason yet that we should use a grain of salt when expressing certainties about the meanings of words.
  In a sense, allowing that 2 is equal to 2 makes sense from many points of view, and this is not surprising since numbers were created by man within tan isomorphic framework - one consisting of elements that have a form similar or comparable to that of the other elements. Relations between numbers are expressible in different but consistent forms. For example, 3 is the result both of 2 + 1, and of 1 x 3, and of 1+1+1 etc. Numbers exist inasmuch as the relations between them exist, so we are not amazed at the fact that 9 ¸ 3 is 3, since 3 + 3 + 3 is 9.
  Words, on the other hand, were and continue to be created and freely and spontaneously altered, and their semantic content varies over time, in space, in individuals, in cultures. Therefore, any attempt is fruitless to force them into rigid formats of meaning, or to use pseudo-mathematic expedients in an attempt to say that "run = walk + fast". Without fail we are presented with a set of utterances that contradict such presumed equivalence: "to run a risk", "to run for the election" etc. To say nothing of marathon men, who walk fast but never run.
  Unlike numbers, words express not only an object meaning, but also the attitude of the speaker toward such object. "A word is nothing but a metaphor for an object or, in some cases, for another word"³. The implicit comparison implies a peculiar way of expressing the indicated object, not a "neutral" expression of it.
  Very appropriately Rabassa recalls a passage from the Gulliver's Travels in which, at Lagado Academy, the problem of the margin of misunderstanding is resolved in a very original, if not very practical, way:

Swift as usual, pretending to speak about other times and other places points his accusatory finger against society and culture in which he lives in his present, and is caustic in his implied judgment of word decoding problems. Actually, only the simultaneous presence of author and translator, together with the object the text he wants to express, could contribute to the elimination of possible misunderstandings of verbal expression, but with them also some of the polysemic richness of the utterances.
  In the case of Lagado's academics, the interpretive triangle formed by object, interpretant, sign in the prototext would share one vertex with the relative interpretive triangle of the metatext: the object vertex.

  While in our reality the two triangles face one another, and the only vertex they don't share, but have very close, is the sign vertex, through which pass, first the interpretive act and, then, the translation act.

  Rabassa never names the interpretant, but the notion is implicitly present throughout the article. He speaks of the words expressing "dog" in various cultures, about the Muslim culture's contempt toward this animal (that has negative connotations as soon as it is named, for this reason), about peculiar connotations in each different, not only national, but also individual culture: just think of the different interpretants of "dog" in two persons that as children had, one, important affectionate relationships with dogs and, in the other case, has been bitten, for example.
  Borges, in order to stress the inadequacy of words, proposed one of his translators not to translate what he said, but what he meant to say. The invitation is evidently ironic because an author's communicative intention is never transparent or obvious (it would be too easy).
  Since, in Rabassa's opinion, when writing an author does nothing but choose the metaphor that best becomes the sense of what he wants to express, and since, evidently, metaphors are all but scientifically formed ("golden", for example, can be used as a metaphor for color or richness or brilliancy and God knows how many other things, so there is no isomorphism in the network of the possible metaphors), the translator has the hard task of abductively reconstructing the process that induced the author to use given metaphors and then she has to understand his presumed communication intention.
  Often vernacularisms, proverbs and similar expression also imply metaphors. Rabassa quotes the saying «Out of sight, out of mind»⁵. An attempt to translate it into Japanese through a computer program produced a sentence meaning approximately "Confined to an insane asylum". Undoubtedly, "Out of sight, out of mind" could well mean that too, but the metaphor activated by the translator is not the same as that implied by the author, so, produces a serious communication problem.

Bibliographical references

CALVINO I. If on a Winter's Night a Traveller, London, Random House, 1998, ISBN 0-749-39923-6.

RABASSA G. No two snowflakes are alike: translation as metaphor, in The Craft of Translation, edited by John Biguenet and Rainer Schulte, Chicago, London, The University of Chicago Press, 1989, ISBN 0-226-04864-3, p. 1-12.

SWIFT J. Gulliver's Travels, 1726, in The Writings of Jonathan Swift, New York, Norton, 1973, ISBN 0-293-042839, p. 1-260.