How do simultaneous interpreters translate while someone speaks?

Simultaneous interpreters do three things at once: they listen to and understand the speaker, structure the message, and reproduce it in another language. They start speaking a short moment after the speaker begins, without waiting for pauses, and keep going as the speech continues. The European Commission's Knowledge Centre notes that these are the same actions used in consecutive interpreting, but all happening at the same time. [1]

A central concept is the ear-voice span, the gap between when the speaker starts a thought and when the interpreter starts to render it. According to the American Translators Association, the right gap varies by language pair. If it is too long, the interpreter forgets what was said. If it is too short, the interpreter risks copying the source grammar or falling for false cognates, such as the Spanish compromiso, which usually means commitment rather than compromise. [2]

Anticipation is another tool. The Knowledge Centre says a good interpreter must predict what the speaker might say, especially when the speaker's language has a very different sentence structure from the target language. A 2010 article in the Journal of Language Teaching and Research likewise describes prediction, based on language ability, knowledge and experience, as a way to save time and keep pace with the speaker. [1] [3]

The same article describes working along the original sentence structure: the interpreter divides what they hear into parts and recombines them to express the full meaning, which shortens the delay. It also describes reformulation as the overall strategy, because word-for-word rendering is impossible between languages as different as English and Chinese. Its point is that the interpreter delivers meaning, not wording. The ATA similarly says interpreters focus on concepts rather than individual words, and reflect the speaker's tone. [3] [2]

Because the interpreter is speaking while listening, interference from the language being heard is more likely. The Knowledge Centre says this can mean moving away from the original syntax, splitting long sentences into shorter ones and avoiding false friends. [1]

In the traditional setup, the speaker talks into a microphone, and the audio goes to the interpreter in a sound-proof booth, who listens through headphones. The interpreter speaks into a microphone, and the interpretation is transmitted wirelessly to headsets, where attendees choose their language channel. Interpreters may also use a portable system without a booth, called a bidule, or whisper the interpretation directly into a listener's ear, which is known as chuchotage. [4] [5] [1]

The 2010 article calls the task unusually demanding, requiring intense concentration that tires the interpreter fairly quickly and can then affect performance. It adds that only a few interpreters handle it well. The ATA says interpreters generally swap with a partner about every 30 minutes, and keep a notebook for figures, dates and names. Trainees usually learn consecutive interpreting first, since simultaneous uses the same skills with added complexity. [3] [2] [1]