Perhaps we need not even decline this comparison. But just see to what far-reaching differences we have come by penetrating somewhat into the seemingly harmless problems of the psychology of errors, differences which at this stage we do not at all know how to reconcile. I offer you a preliminary compromise on the basis of the analogy of the judge and the defendant. You will grant me that the meaning of an error admits of no doubt when the subject under analysis acknowledges it himself. I in turn will admit that a direct proof for the suspected meaning cannot be obtained if the subject denies us the information; and, of course, that is also the case when the subject is not present to give us the information. We are, then, as in the case of the legal procedure, dependent on circumstances which make a decision at one time seem more, and at another time, less probable to us. At law, one has to declare a defendant guilty on circumstantial evidence for practical reasons. We see no such necessity; but neither are we forced to forego the use of these circumstances. It would be a mistake to believe that a science consists of nothing but conclusively proved theorems, and any such demand would be unjust. Only a person with a mania for authority, a person who must replace his religious catechism with some other, even though it be scientific, would make such a demand. Science has but few apodeictic precepts in its catechism; it consists chiefly of assertions which it has developed to certain degrees of probability. It is actually a symptom of scientific thinking if one is content with these approximations of certainty and is able to carry on constructive work despite the lack of the final confirmation. But where do we get the facts for our interpretations, the circumstances for our proof, when the further remarks of the subject under analysis do not themselves elucidate the meaning of the error? From many sources. First of all, from the analogy with phenomena extraneous to the psychology of errors; as, for example, when we assert that the distortion of a name as a slip of the tongue has the same insulting significance as an intentional name distortion. We get them also from the psychic situation in which the error occurred, from our knowledge of the character of the person who committed the error, from the impressions which that person received before making the error, and to which he may possibly have reacted with this error. As a rule, what happens is that we find the meaning of the error according to general principles. It is then only a conjecture, a suggestion as to what the meaning may be, and we then obtain our proof from examination of the psychic situation. Sometimes, too, it happens that we have to wait for subsequent developments, which have announced themselves, as it were, through the error, in order to find our conjecture verified.