THE SEMANTIC BACKGROUND 9
to stand for some referent or thing-meant, but as an attribute of something to be explained by the situation or another word. Thus, this and all similar words have a peculiarly flexible and variable nature, and the senses of such words are even less pinned to specific referents than are those of other sorts of words. In short: variation in the meanings of esthetic and critical terms is the rule, not the exception. If we recognize this fact sufficiently, we should not be surprised by a remark like that of Richards that "until recently Oxford-trained men and Cambridge-trained men hardly ever understood one another on any abstract topicbecause their key explanatory words turned in different locks/' u The opening words of Richards' sentence"until recently"are reassuring, yet it is highly doubtful whether the improvement they indicate can likewise be found in the bulk of supposedly educated discussions upon art and beauty in America.
The difficult semantic situation I have sketched should lead no one to adopt, however, the desperate extreme of Mr. Stuart Chase who, in his popular book The Tyranny of Words, contends that since abstract and general terms have no fixed and exact meanings and are therefore treacherous, one may as well substitute "blab-blab" for each of them. No; it is not necessary, perhaps not even desirable, to use these "fictional" or "wandering" words less, but rather to adopt toward them a different attitude: to treat them not as entities but as metaphors which need frequently to be translated "into concrete terms which hold fewer possibilities of difference of opinion." 12 They should be studied, not eliminated.
Closely allied with and partly dependent upon the false notion that a one-to-one or direct relation exists between words and referents is the equally false notionand this is perhaps the chief misconception leading to semantic befud-
11. I. A. Richards, Interpretation in Teaching (New York, 1938), p. 348. is. Hugh Walpole, Semantics (New York, 1941), p. 171.