SECT. II.—THE COMMON EFFECTS OF POETRY，NOT BY RAISING IDEAS OF THINGS
THE common notion of the power of poetry and eloquence, as well as that of words in ordinary conversation, is that they affect the mind by raising in it ideas of those things for which custom has appointed them to stand. To examine the truth of this notion, it may be requisite to observe, that words may be divided into three sorts. The first are such as represent many simple ideas united by nature to form some one determinate composition, as man, horse, tree, castle, &c. These I call aggregate words. The second are they that stand for one simple idea of such compositions, and no more; as red, blue, round, square, and the like. These I call simple abstract words. The third are those which are formed by an union, an arbitrary union, of both the others, and of the various relations between them in greater or less degrees of complexity; as virtue, honour, persuasion, magistrate, and the like. These I call compound abstract words. Words, I am sensible, are capable of being classed into more curious distinctions; but these seem to be natural, and enough for our purpose; and they are disposed in that order in which they are commonly taught, and in which the mind gets the ideas they are substituted for. I shall begin with the third sort of words; compound abstracts, such as virtue, honour, persuasion, docility. Of these I am convinced, that whatever power they may have on the passions, they do not derive it from any representation raised in the mind of the things for which they stand. As compositions, they are not real essences, and hardly cause, I think, any real ideas. Nobody, I believe, immediately on hearing the sounds, virtue, liberty, or honour, conceives any precise notions of the particular modes of action and thinking together with the mixt and simple ideas and the several relations of them for which these words are substituted; neither has he any general idea, compounded of them; for if he had, then some of those particular ones, though indistinct perhaps, and confused, might come soon to be perceived. But this, I take it, is hardly ever the case. For, put yourself upon analyzing one of these words, and you must reduce it from one set of general words to another, and then into the simple abstracts and aggregates, in a much longer series than may be at first imagined, before any real idea emerges to light, before you come to discover anything like the first principles of such compositions; and when you have made such a discovery of the original ideas, the effect of the composition is utterly lost. A train of thinking of this sort is much too long to be pursued in the ordinary ways of conversation; nor is it at all necessary that it should. Such words are in reality but mere sounds; but they are sounds which being used on particular occasions, wherein we receive some good, or suffer some evil, or see others affected with good or evil; or which we hear applied to other interesting things or events; and being applied in such a variety of cases, that we know readily by habit to what things they belong, they produce in the mind, whenever they are afterwards mentioned, effects similar to those of their occasions. The sounds being often used without reference to any particular occasion, and carrying still their first impressions, they at last utterly lose their connexion with the particular occasions that gave rise to them; yet the sound, without any annexed notion, continues to operate as before.