Good Translations from Poor Originals?
It is accepted by the translator community as self-evident that the translation should accurately reflect the original, possibly in both form and content. So what happens if the translator receives a poorly written original or one containing ambiguities and even outright factual errors (as opposed to typos)? Is the translator supposed to produce good English out of poor German, Russian, or Chinese? Are the ambiguities to be left in the target text (when this is possible) on the assumption that they may have been intentional? Is the translator supposed to correct the errors of the original? How far should the translator go in his or her research to ensure that the original is factually correct?
If the translator is in a position to contact the author of the original, they can discuss certain aspects of the text, but this is not always possible. The text may have reached the translator via a middleman or may have been extracted from a publication whose author is no longer available. A translator's note addressed to the client does not always solve the problem, because the client may be more clueless about the text than the translator. On the other hand, an error left in the text may come back to haunt the translator later.
How should the problem of poor ST be handled?