As with many issues of language, I believe that multitudinous difficulties of translation categorize well as lexical and syntactical. Lexical difficulties stem from the fact that there is not a one-to-one correspondence between words in sundry languages. The most obvious case is translating between two languages, one of an extensive vocabulary and the other a scarce, but with a functional correspondence of words in the former to those of the latter.1 Translating from the former to the latter, the translator typically faces little interpretive challenge, as only one word in the final language matches a word in the source language. However, if he translates word-for-word in this manner, he fails to convey the original precision. Thus, a conceptually accurate translation must add words, and one stylistically accurate must add ambiguity. The reverse, translating from the poorer to the richer, encounters the opposite problem. A given word in the source language will, in different contexts, translate to a different word in the final. Thus, to maintain the original ambiguity of meaning, he must add words to indicate the ambiguity; to maintain the style, he must act as interpreter of meaning, adding his own opinion of the precise meaning of the original. And this is in an optimal situation, where the correspondence of words is functional in one direction. In real translation, meanings will overlap, so that the translator risks adding ambiguity in some respects and removing it in others. For example, in English words may be vague in exact meaning, and specific in context; in Greek, the converse is true. L&S defines ŒªœåŒ≥ŒøœÇ as “Computation, reckoning, relation, correspondence, proportion, explanation, debate, continuous statement, narrative, verbal expression or utterance, a particular utterance, saying, thing spoken of, subject-matter, speech”. To an English speaker, this seems to be a very broad range of meanings. At the same time, it conveys a fairly specific general meaning, applied in a variety of contexts. In English, we use a different word for each context, but frequently use a word for a variety of meanings within one context for which Greek uses multiple words. English better specifies context; Greek meaning. Thus, word-for-word translation from Greek to English specifies the context more strongly and the meaning less; the converse, the inverse. The only means of circumventing these difficulties is to attempt to force a correspondence, as the scholastics did in creating a technical terminology within Latin to parallel the nuances of Aristotle’s Greek.
Moreover, lexical difficulties may pale in comparison to syntactical. Syntactical difficulties further resolve into conceptual, accentual, and stylistic. Conceptual difficulties are very similar to lexical; certain forms in one language are more precise than comparable forms in another, e.g. the careful distinction in Greek between natural and artificial result, difficult to express except by convention in English. Accentual difficulties arise particularly when translating from a highly inflected language into a weakly; highly inflected languages tend to use word-order to express accent, while weakly syntactical role. Thus, maintaining the original word order introduces odd word orders where the original was natural and even grammatical errors, while transposing words changes emphasis. Going the other way, maintenance of the original word order implies an emphasis not in the original; transposition an interpretation of the intended emphasis. Yet more challenging are stylistic difficulties. Style, more so than any other aspect of a language, is relative to context; what may seem informal and idiomatic, when literally translated, may appear formal and stilted. Thus, preserving the original style may require unidiomatic translation.
1. A function associating each member of the range to one member of the domain, but not necessarily the converse.