140 MYTHOLOGY rise to the myth of Cronus, and how do you explain that god s mutilation of his father, and his habit of swallowing down alive each of his own children, whom he afterwards disgorged? Grant that Dionysus only means the vine and its influence, and how do you explain his unspeakable conduct as recorded in the mysteries at Halimus in Attica 1 How, in short, as Arnobius asked the heathen, how, if the myths represent pure natural facts and phenomena, do they come to be crowded with the obscene details which disgusted philosophers six hundred years before Christ? In what stage of society did this " impure way of stating pure facts " win favour 1 Mr Miiller must fix the period at which such details were invented, some time between his Mythopoeic age and the age of Xenophanes and Theagenes. In the meantime his system does not explain and scarcely touches on the very facts that most call for explanation. Why did the Greek poets relate divine myths of which we find the parallels "among the lowest tribes of Africa and America " 1 (2.) Mr Muller s system is a result of the philological discoveries that establish the linguistic unity of the Indo- European peoples, and is founded on an analysis of their language. But myths precisely similar in irrational and repulsive character to those of the Aryan races exist among Australians, South Sea Islanders, Eskimo, Bushmen in Africa, among Solomon Islanders, Iroquois, and so forth. The facts being identical, an identical explanation should be sought, and, as the languages in which the myths exist are essentially different, an explanation founded on the Aryan language is likely to prove too narrow. Mr Miiller indeed, has ventured into Finnish philology and mythology, but a wider examination is needed. Once more, even if we discover the original meaning of a god s name, it does not follow that we can explain by aid of the significance of the name the myths about the god. For nothing is more common than the attraction of a more ancient story into the legend of a later god or hero. Myths of unknown antiquity, for example, have been attracted into the legend of Charlemagne, just as the bons mots of old wits are transferred to living humorists. Therefore, though we may ascertain that Zeus means " sky " and Agni " fire," we cannot assert, with Mr Miiller, that all the myths about Agni and Zeus were originally told of fire and sky. When these gods became popular they would inevitably inherit any current exploits of earlier heroes or gods. These exploits would therefore be explained erroneously if regarded as originally myths of sky or fire. We cannot convert Mr Miiller s proposition " there was nothing told of the sky that could not in some form or other be ascribed to Zeus" into "there was no thing ascribed to Zeus that had not at some time or other been told of the sky." This is also, perhaps, the proper place to observe that names derived from natural phenomena sky, clouds, dawn, and sun are habitually assigned by Brazilians, Ojibways, Australians, and other savages to living men and women. Thus the story origin ally told of a man or woman bearing the name "sun," "dawn," "cloud," may be mixed up later with myths about the real celestial dawn, cloud, or sun. For all these reasons the information obtained from philological analysis of names is to be distrusted. We must also bear in mind that early men when they conceived, and savage men when they conceive, of the sun, moon, wind, earth, sky, and so forth, have no such ideas in their minds as we attach to these names. They think of sun, moon, wind, earth, and sky as of living human beings with bodily parts and passions. Thus, even when we discover an elemental meaning in a god s name, that meaning may be all unlike what the word suggests to civilized men. A final objection is that philologists differ widely as to the true analysis and real meaning of the divine names. Mr Miiller, for ex ample, connects /cpo^os with xpovos, "time"; Preller with Kpcuvw, "I fulfil," and so forth. (3.) The objection to Mr Miiller s doctrine of the Rhematic and Dialectic periods of language as bearing on mythology is that he either supposes man to have had no myths in these periods, or takes no account of the myths they may have had. Yet it is certain, and admitted by himself, that we do find myths in languages of all known sorts. If man on his way to being Aryan or Semitic, if man in the Rhematic and Dialectic stages, possessed no myths, he must have differed from all men of whom we know anything. If he did possess myths these cannot have been produced by the conditions of Mr Miiller s Mythopoeic age, because that age had not yet been reached. And if man in the Rhematic and Dialectic periods had myths, and if they survived into the Mytho- pceic and later periods, they cannot be explained by Mr Miiller s Mythopoeic theory. Especially if these earlier myths crystallized round a god or hero of later date will the effort to explain the earlier stories by analysis of the later names be fruitless. As to the Mythopoeic period itself, as described by Mr Miiller, it was rather an age when the materials for myths were accumulated than when myths themselves were developed. (4.) Mr Miiller attempts to show how the conversation of men in the Mythoposic age became the source of myths. Here he is endeavouring, really, to account for that universal attribution of life, sex, action, and thought to all phenomena which is, indubitably, the essential con dition of mythology. He explains this " animism " as an erroneous condition of thought into which men essentially civilized were driven by the nature of language. In lan guage all words denoted gender, hence (he thinks) men were led to suppose that gender and sex, with all that follows, were possessed by all objects. But it is scarcely necessary to reply (the truth being so obvious) that the gender-termina tions of words reflected, and must have arisen from, a state of the human intellect in which all things were regarded as persons. The civilized men of the Mythoposic age were not compelled, as Mr Miiller thinks, to believe that all phenomena were persons, because the words which denoted the phenomena had gender-terminations. On the other hand, the gender-terminations were survivals from an early stage of thought in which personal characteristics, including sex, had been attributed to all phenomena. This condition of thought is demonstrated to be, and to have been, universal among savages, and it may notoriously be observed among children. Mr Miiller explains it as the result of reflexions on gender- terminations, but how does he explain these terminations themselves 1 His theory is that, somehow, gender -terminations arose in language. Then, when they had become a tradition of language, they " reacted on the mind with irresistible power," and men, who had previously been as sensible as ourselves, felt themselves obliged to animate and personify all pheno mena. Mr Miiller remarks " there is some truth in this," (that is, in the contention that a belief in the personal character of phenomena reflected itself in gender-termina tions), " but it only serves to confirm the right view of the influence of language on thought, for this tendency, though in its origin intentional, and therefore the result of thought, became soon a mere rule of tradition in language, and then it reacted on the mind with irresistible power " (Sel. Ess., i. 604). Mr Miiller thinks that men first held nature to be animated and personal. This belief reflected itself in language, but it (apparently) produced no myths. In a later and civilized age language brought back the old state of intellect, and myths were produced. This becomes a superfluous hypothesis of degradation, for the original state
