THE SEMANTIC BACKGROUND 15
all semantic ailments, I shall indicate two significant limitations in the application and usefulness of them.
(a) The necessity for and the nature of volitional definitions by no means imply that any one definition is as satisfactory as any other; for while a word has neither a "true," "single correct," nor "essential" meaning, words are not wholly conventional; they do "apply to groups of similar situations, which might be called areas of meaning" 22 For this reason, we remarked earlier that, "up to a point" we may give meanings to words. Thus, while definitions may be arbitrary in the sense of being deliberately chosen, they must never be unnecessarily or excessively arbitrary; for even if they cannot reasonably be considered true or false, they may certainly be considered sensible or foolish. The meanings of words, in short, should be "waringly and sparingly" altered since, to quote Locke, "He that applies his names to ideas different from their common use, wants propriety in his language, and speaks gibberish." 23 Since this is so, I suggest that there are certain general criteria of good volitional definitions which may be summarized in the words of J. R. Reid: the definitions should be "(1) clear and intelligible, (2) as convenient or useful as possible in dealing with a given subject-matter, and
(3) should as nearly conform to established usage, if any, as is compatible with clarity and usefulness in the context." 24 We shall have occasion to refer frequently to these criteria.
(b) Most of the time we necessarily rely upon "established usage" in order to express our meanings; for while usage does not permanently fix the meanings of words, it does largely regulate the verbal customs which enable us to communicate with one another.25 And in the last few centuries, dictionaries
22. S. I. Hayakawa, Language in Action (New York, 1941), p. 71.
23. "Of Words," An Essay Concerning Human Understanding (Oxford, 1894), II, 144. A. C. Fraser, ed.
24. A Theory of Value, p. 25.
25. In Interpretation in Teaching, chaps, xv and xvi; in The Philosophy of Rhetoric, lecture iii; and in How To Read a Page, chap, iii, Richards