Date of this Version
Published in Transactions of the Nebraska Academy of Sciences, Volume 3 (1976).
In natural languages there are various ways of indicating the probable character of sentences: (i) by simply adding operators like "probably," "perhaps," and so on; (ii) by using the subjunctive; (iii) by using the future tense: and (iv) by forming sentences which - intentionally or unintentionally _ do not express adequately the situation under consideration. An example for (iv) is for instance the sentence "The situation looks good" when in reality it is rather not so good. There are, of course, mixtures between (i), (ii), (iii), and (iv) possible. One of the linguistic differences between (i). (ii), and (iii) on the one hand and (iv) on the other hand is, that case (iv) can lead to a change of meaning. (iv) comprises those expressions of natural languages which we call "not literal," "figurative," and the like. It is obvious that at least case (iv) may be explained best by a probabilistic semantics. As a specific example of (iv) we will discuss here euphemism and add one traditional explication of "euphemism:" a euphemism is a pleasant way of referring to something unpleasant; a kakophemism, by the way, is simply the opposite of a euphemism.
For theoretical reasons we will in general exclude euphemistic words and expressions shorter than a sentence, and also religious euphemisms (Leinfellner, 1971). Only when we will explain the change of meaning we will return to the concept of euphemism as a single word.
We will modify here the standard method of allotting probabilities to sentences, where a sentence is probable with respect to another one, P(S1, S2) = r, by using a relation of euphemistic difference, e, which is to be founded pragmatically. The difference between the cases (i)-(iii) and (iv) thus lies also in the pragmatic foundation or nature of e: i.e. in order to explain case (iv) we have to take into consideration the personal attitude of the speaker, the language user in general. In the following model of euphemism we have to assume that there exists always an empirically true sentence B - for practical reasons we exclude here empirically false sentences --- which is empirically true with respect to the same empirical situation to which the euphemistic sentence E refers euphemistically: e(E, B) = r. e has two limit cases: a euphemism ceases to be one when r = 1; and it turns into a total lie when r = 0. In the case that r lies between 0 and 1, 0r, we could compute the absolute euphemistic shift, either of the sentence E, or, as we will see later, of the euphemistic meaning compared to the basic or empirical one. It is also clear that we can insert between 0 and 1 various degrees of euphemistic "efficiency;" they are topological values of variously "shaded" euphemisms.