Answer: It depends on the context.
The above joke is probably only funny to translators and people who interact with them on a regular basis, and specifically, to anyone who has ever had the frustrating experience of asking a translator how to translate a particular word and has been met with this same answer. But the fact is, when replying to your deceptively simple question as to how to say “echar” in English, the familiar refrain “it depends on the context” is not merely an attempt to be evasive, because it really is important to know first whether your intention is to “echar algo a la basura” (throw something into the garbage), “echarme una llamada”(give me a call) or “echarme la culpa” (put the blame on me), among many other possibilities.
Indeed, the seemingly innocent question of how to translate any single word into another language strikes at the very heart of a problem that is always at the forefront of the translator’s mind: the problem of equivalence. The notion of a supposed equivalence between languages, so blithely assumed by bilingual dictionaries and MT software, is problematic because the sheer complexity of language means that one-to-one equivalence between words, sentences or texts in two different languages is simply impossible. And yet, achieving this equivalence is the translator’s mission; equivalence is, in a sense, our philosopher’s stone.
At the level of individual words, the quest for equivalence is complicated by the fact that all words have multiple layers of meaning. In an attempt to classify these layers, the linguist D.A. Cruse has posited four categories of lexical meaning: propositional, expressive, presupposed and evoked meaning. It is the first of these, the propositional meaning (also commonly referred to as denotative meaning), that people are usually asking for when they ask for a straight translation of a given word: it is the element or aspect of reality that the word refers to, its explicit definition. Even this denotative meaning can be extremely complicated, as illustrated in the above example of “echar”, as the context of a word’s use may in fact change its denotation. But it is further complicated by the word’s expressive (or connotative) meaning, which adds an evaluative element that the purely referential denotative meaning does not contain. For example, while the propositional meaning of the phrases “he did not do it” and “he failed to do it” is essentially the same, the use of the verb “fail” in the second case suggests a critical evaluation on the part of the reader that is absent in the use of the negative adverb. The transferral of both propositional and expressive layers of meaning in the translation process is but one of the daily challenges faced by translators, as is the avoidance of inadvertently introducing expressive layers that were not present in the source text.
The other layers of meaning according to Cruse’s taxonomy pose obstacles that often trip up inexperienced translators. Presupposed meaning relates to words that commonly appear together in a particular language (collocations). Violations of collocational restrictions will often result in language that sounds “strange” to native speakers. The literal translation of “lavarte los dientes” – “wash your teeth” – is a good example of this, as the use of “wash” instead of “brush” may signal to the reader that the author is not a fluent English speaker. On the other hand, evoked meaning is related to the dialect and register appropriate to a given location or situation, which if not duly observed could completely undermine the credibility of the text; imagine, for example, the bewildering effect of hearing the characteristically British adjective “splendid” spoken by an American cowboy, or seeing the specifically legal term “hereinafter” in the advertising copy for a new shampoo. These are obvious examples, but even much subtler digressions from register can have a disconcerting effect on the reader, which is why it is so important for translators to be intimately familiar with the kind of lexical content typical of the genre of the text they are translating. The nature of this lexical content is highly arbitrary and idiosyncratic, and can thus only really be “learned” through extensive reading in the target language in the genre concerned. Thus, even the most experienced literary translator in the world might not be up to the challenge of translating a patent, unless s/he was familiar with the kind of words and phrases that patents habitually use.
And the above are only a few of the dilemmas faced by translators, and only at the level of individual word choices. Beyond this, there are questions of grammatical structures, idioms and fixed expressions, text organization and textual cohesion that are very distinct in different languages, and pose further obstacles in the translator’s impossible quest for equivalence. And because languages are such fluid and constantly changing phenomena, and their layers of meaning (in spite of the best efforts of language academies and dictionaries) are determined by implicit convention rather than prescriptive norms, the search for equivalence between languages will always be subjective, always contestable, and always, of course, highly dependent on the context.