Now we return, perhaps with increased faith, to our problem. Apparently it is very probable that the dreamer knows of his dream; the question is, how to make it possible for him to discover this knowledge, and to impart it to us? We do not demand that he give us the meaning of his dream at once, but he will be able to discover its origin, the thought and sphere of interest from which it springs. In the case of the errors, you will remember, the man was asked how he happened to use the wrong word, "_vorschwein_," and his next idea gave us the explanation. Our dream technique is very simple, an imitation of this example. We again ask how the subject happened to have the dream, and his next statement is again to be taken as an explanation. We disregard the distinction whether the dreamer believes or does not believe he knows, and treat both cases in the same way. This technique is very simple indeed, but I am afraid it will arouse your sharpest opposition. You will say, "a new assumption. The third! And the most improbable of all! If I ask the dreamer what he considers the explanation of his dream to be, his very next association is to be the desired explanation? But it may be he thinks of nothing at all, or his next thought may be anything at all. We cannot understand upon what we can base such anticipation. This, really, is putting too much faith in a situation where a slightly more critical attitude would be more suitable. Furthermore, a dream is not an isolated error, but consists of many elements. To which idea should we pin our faith?"