Similarly, we find out when we interpret dreams what is the real meaning of the doubts and uncertainties, so frequently mentioned by dreamers, whether a certain element did actually appear in the dream, whether it was really this and not rather something else. As a rule, there is nothing in the latent thoughts corresponding with these doubts and uncertainties; they originate wholly through the operation of the censorship and are comparable to a not entirely successful attempt at erasure. One of our most surprising discoveries is the manner in which _opposites_ in the latent dream are dealt with by the dream-work. We know already that points of agreement in the latent material are replaced by condensation in the manifest dream. Now contraries are treated in just the same way as similarities, with a marked preference for expression by means of the _same_ manifest element. An element in the manifest dream which admits of an opposite may stand simply for itself, or for its opposite, or for both together; only the sense can decide which translation is to be chosen. It accords with this that there is no representation of a “No” in dreams, or at least none which is not ambiguous. A welcome analogy to this strange behaviour of the dream-work is furnished in the development of language. Many philologists have maintained that in the oldest languages opposites such as: strong—weak, light—dark, large—small, were expressed by the same root word (_antithetical sense of primal words_). Thus, in old Egyptian “_ken_” stood originally for both “strong” and “weak.” In speaking, misunderstanding was guarded against in the use of such ambivalent words by the intonation and accompanying gestures; in writing, by the addition of a so-called “determinative,” that is to say, of a picture which was not meant to be expressed orally. Thus, “_ken_” = “strong” was written in such a way that after the letters there was a picture of a little man standing upright; when “_ken_” meant “weak,” there was added the picture of a man in a slack, crouching attitude. Only at a later period did the two opposite meanings of the same primal word come to be designated in two different ways by slight modifications of the original. Thus, from “_ken_” meaning “strong—weak” were derived two words: “_ken_” = “strong” and “_kan_” = “weak.” Nor is it only the oldest languages, in the last stages of their development, which have retained many survivals of these early words capable of meaning either of two opposites, but the same is true of much younger languages, even those which are to-day still living. I will quote some illustrations of this taken from the work of C. Abel (1884): In Latin, such ambivalent words are: _altus_ = high or deep. _sacer_ = sacred or accursed. As examples of modifications of the original root, I quote: _clamare_ = to shout. _clam_ = quietly, silently, secretly. _siccus_ = dry. _succus_ = juice. and, in German, _Stimme_ = voice. _stumm_ = dumb.