To express ideas dearly and effectively, you must first have them clear in your own mind. It follows that if you wish to re-express someone else‘s ideas without having the possibility of repeating them word for word - which is the ease for the interpreter - then you must make a clear, structured analysis of them. To make that analysis, you have to understand the individual ideas that are the basic building blocks of a speaker‘s line of reasoning. In order, then, the three basic stages of a consecutive interpreter‟s work are understanding, analyzing, and re-expressing. These three notions have to be looked at in turn. Understanding The ‗understanding‘ here is not of words but of ideas, for it is ideas that have to be interpreted. Obviously, you cannot understand ideas if you do not know the words the speaker is using to express them, or if you are not acquainted sufficiently with the grammar and syntax of the speaker‘s language to follow the ideas. In connection with the notion of ‗not knowing words‘, it is best to address here what is probably one of the two commonest questions put to conference interpreters by non-interpreters: ‗What do you do if you do not know a word or an expression that you hear in a speech?‘ The answer to this has already been partially given when we say that the interpreter has to understand ideas, not words. It may well be perfectly possible to understand a speaker‘s meaning without actually understanding every single word and expression they use, and without having to reproduce all of those terms in the interpretation. For example, imagine a delegate says: 15 I don‟t think that the advisory committee is the appropriate forum for discussion of this point. What is important is that the groundwork be done in the technical working parties, in order to prepare the basis for a decision in the executive committee. Let‘s assume the unlikely, namely that the interpreter understands neither forum nor groundwork. Yet this does not prevent them from understanding that (1) the advisory committee is not the right place to discuss the matter, and (2) the question has to be properly prepared for the executive committee by the technical working parties. The interpretation is possible without all the words and without changing the meaning. There are other occasions, however, where a word is too important to be skated round in this way. Let us say that Norway is being discussed and the motorway network is referred to: Given the topography of the country, the construction of motorways has been very expensive. The Norwegians have found the solution to their financing problems by imposing tolls. And these tolls arc pretty expensive. The roads are wonderfully built, a pleasure to drive upon, with beautiful scenery, but when the poor driver gets to the end of their journey and has to pay the toll, they certainly feel that their wallet is much lighter. The key word here is toll, and if the interpreter does not know it they can hardly avoid it. But the interpreter can also benefit from working in consecutive. By the time they start interpreting they will have heard the whole speech, and should have been able to deduce the meaning of toll from context, given the number of clues they have. Thus, again, it is possible for the interpreter to work satisfactorily, indeed in this case totally accurately, without their having known in advance all the vocabulary used by the speaker. 16 Two further points should be made here. First, interpreters must accept that there are times when they do not know a word or an expression, can neither avoid it nor deduce its meaning from context, and are consequently stuck. In those circumstances, the interpreter, particularly in consecutive where there is a straightforward human contact with the delegates, must admit their ignorance, and, if necessary, clarify the question with the delegates. The interpreter cannot be expected to be a walking multilingual dictionary-cum-encyclopedia and has a perfect right not to know certain things. On the other hand, the interpreter does not have the right to ‗betray‘ the delegates by missing things out or guessing at meanings in order to hide their ignorance. Second, in order to understand meaning without knowing all the lexical items, and still more in order to deduce meaning from context, the interpreter must in any case have a thorough knowledge of their passive language(s). Arguing that interpreting is possible without knowing all the words should not be distorted into the argument that an interpreter does not need to know their passive languages properly. To return to the question of „understanding‟, we must stress that pure linguistic understanding (what we might call ‗comprehension‘), although necessary, is not a sufficient condition for the interpreter to be able to re-express ideas efficiently in another language. Interpreters must be able to seize meaning in a split second, and must therefore listen constantly in an active, attentive way, always asking themselves ‗What does the speaker mean?