Literally right

Veronica Perez Guarnieri's advice to “stay as literal as possible” when interpreting at depositions seems to have ruffled a few feathers, to judge by the feedback received. Here are some quick takes on literalness in technical interpretation.

Word-for-word instinctive parroting usually needs to be trained away; interpreting students are rightly enjoined to “go for the meaning” and “move away from the source” because, once ingrained, the habit is very hard to break, and usually makes for wildly inaccurate translations or speech clunky enough to be unintelligible.

As a result we develop knee-jerk reflexes to avoid cognates and tend to transform sentence structure to make it sound more natural as a matter of course: we hear “ordinary” in English and eschew “ordinario” because it also means “vulgar”; we hear “passengers have been shocked by this discovery” – quite ordinary-sounding in English – and turn it into “este descubrimiento ha dejado atónitos/estupefactos a los pasajeros”. In simultaneous mode it usually keeps us out of grammatical or semantic trouble, and in consecutive our memory process focuses on the meaning rather than the form. Colleagues who regularly pull off highly creative and totally accurate renditions earn our admiration.

Yet out in the real world of interpreting, specific words used by speakers do matter. As Marc Orlando points out in this issue, literary speakers may choose their words and tenses and modes quite deliberately. Julian Barnes eloquently discusses writer's choices and how they are dealt with in a recent review of Madame Bovary translations (see Language in the News).

Lawyers, too, are famous for their wordcraft – their obfuscation perceived as particularly careful. Scientists, not generally reputed to indulge in verbal flourishes, often choose one quasi-synonym very deliberately over another: a chimpanzee is not a monkey, though the distinction might seem picayune to the unenlightened interpreter. Geeks of any color often combine fanciful naming choices (think “spam”, “charm quarks” or the Babel Fish service) with grammatical fastidiousness – after all, making sure all open parentheses get closed is a basic skill for scientists and IT experts, as it is for interpreters unless the speaker’s too fast.

Experience and subject expertise teach us when to hang onto the original words or structure for best results. No one is saying it's ever a good idea to translate “bell curve” as “curva de campana”, a practically meaningless phrase in Iberian Spanish vaguely suggesting a motor accident because “vuelta de campana” denotes a car rolling over. The phrase might just barely get the message across to an audience of statisticians, who use the word “curva” readily where a more standard word is “gráfica”, and who might make the link between the word “campana” (bell) and the normal distribution curve (“gráfica de distribución normal”), usually termed “campana de Gauss”. But “curva de campana” will just baffle a roomful of Spanish orthopaedic surgeons hearing an aside about evidence-based medicine. Any recognised term or phrase should be chunked as a single unit.

Another good strategy, cognate avoidance, may be rash where ignorance prevails: the above-mentioned “normal distribution curve” has been variously rendered as the equivalent of “ordinary distribution curve” or “common distribution curve” within my hearing. (Perhaps interpreter training should include basic statistical concepts: they do keep cropping up.)

“Deverbalization” is excellent advice for students and for anyone overly attached to the original sentence structure, but does not accurately describe the best interpreters I have heard, who know when to shift between freely worded, well-styled styled renditions and source-hugging, cognate-embracing versions, largely because, aware of their limitations, they study hard and keep learning from mistakes – their own, the speakers' and their colleagues'.

Recommended citation format:
Mary FONS I FLEMING. "Literally right". December 10, 2010. Accessed August 17, 2019. <>.