So I return to Aristotle. Aristotle (I shall in fairness say) does not anticipate Dr Gummere, to contradict or refute him; he may even be held to support him incidentally. But he sticks to business, and this is what he says ("Poetics," C. IV):
Poetry in general seems to have sprung from two causes, and these natural causes. First the instinct to imitate is implanted in man from his childhood, and in this he differs from other animals, being the most imitative of them all. Man gets his first learning through imitation, and all men delight in seeing things imitated. This is clearly shown by experience….
To imitate, then, being instinctive in our nature, so too we have an instinct for harmony and rhythm, metre being manifestly a species of rhythm: and man, being born to these instincts and little by little improving them, out of his early improvisations created Poetry.
Combining these two instincts, with him, we arrive at harmonious imitation. Well and good. But what is it we imitate in poetry?— noble things or mean things? After considering this, putting mean things aside as unworthy, and voting for the nobler—which must at the same time be true, since without truth there can be no real nobility—Aristotle has to ask `In what way true? True to ordinary life, with its observed defeats of the right by the wrong? or true, as again instinct tells good men it should be, universally?' So he arrives at his conclusion that a true thing is not necessarily truth of fact in a world where truth in fact is so often belied or made meaningless—not the record that Alcibiades went somewhere and suffered something—but truth to the Universal, the superior demand of our conscience. In such a way only we know that "The Tempest" or "Paradise Lost" or "The Ancient Mariner" or "Prometheus Unbound" can be truer than any police report. Yet we know that they are truer in essence, and in significance, since they appeal to eternal verities—since they imitate the Universal—whereas the police report chronicles (faithfully, as in duty bound, even usefully in its way) events which may, nay must, be significant somehow but cannot at best be better to us than phenomena, broken ends and shards.