> AFAIK, humans who do simultaneous interpretation are provided with at least an outline, if not full script They are usually provided with one, but it is by n

> AFAIK, humans who do simultaneous interpretation are provided with at least an outline, if not full script They are usually provided with one, but it is by no means necessary. SI is never truly simultaneous and will have a delay, and the interpretor will also predict based on the context. Which makes certain languages a bit more difficult to work with, e.g. Japanese, sentences of which I believe often have the predicate after the object, rather than the usual subject-predicate-object order, making the "prediction" part harder.