Given souls, acquired by thinking on the lines already described, Mr. Tylor develops Gods out of them. But he is not one of the writers who is certain about every detail. He 'scarcely attempts to clear away the haze that covers great parts of the subject.'[5]
The human soul, he says, has been the model on which man 'framed his ideas of spiritual beings in general, from the tiniest elf that sports in the grass up to the heavenly creator and ruler of the world, the Great Spirit.' Here it is taken for granted that the Heavenly Ruler was from the first envisaged as a 'spiritual being'—which is just the difficulty. Was He?[6]
The process of framing these ideas is rather obscure. The savage 'lives in terror of the souls of the dead as harmful spirits.' This might yield a Devil; it would not yield a God who 'makes for righteousness.' Happily, 'deified ancestors are regarded, on the whole, as kindly spirits.' The dead ancestor is 'now passed into a deity.'[7] Examples of ancestor-worship follow. But we are no nearer home. For among the Zulus many Amatongo (ancestral spirits) are sacred. 'Yet their father [i.e. the father of each actual family] is far before all others when they worship the Amatongo…. They do not know the ancients who are dead, nor their laud-giving names, nor their names.'[8] Thus, each new generation of Zulus must have a new first worshipful object—its own father's Itongo. This father, and his very name, are, in a generation or two, forgotten. The name of such a man, therefore, cannot survive as that of the God or Supreme Being from age to age; and, obviously, such a real dead man, while known at all, is much too well known to be taken for the creator and ruler of the world, despite some African flattering titles and superstitions about kings who control the weather. The Zulus, about as 'godless' a people as possible, have a mythical first ancestor, Unkulunkulu, but he is 'beyond the reach of rites,' and is a centre of myths rather than of worship or of moral ideas.[9]
After other examples of ancestor-worship, Mr. Tylor branches off into a long discussion of the theory of 'possession' or inspiration,[10] which does not assist the argument at the present point. Thence he passes to fetishism (already discussed by us), and the transitions from the fetish—(1) to the idol; (2) to the guardian angel ('subliminal self'); (3) to tree and river spirits, and local spirits which cause volcanoes; and (4) to polytheism. A fetish may inhabit a tree; trees being generalised, the fetish of one oak becomes the god of the forest. Or, again, fetishes rise into 'species gods;' the gods of all bees, owls, or rabbits are thus evolved.
Next,[11]
'As chiefs and kings are among men, so are the great gods among the lesser spirits…. With little exception, wherever a savage or barbaric system of religion is thoroughly described, great gods make their appearance in the spiritual world as distinctly as chiefs in the human tribe.'
Very good; but whence comes the great God among tribes which have neither chief nor king and probably never had, as among the Fuegians, Bushmen, and Australians? The maker and ruler of the world known to these races cannot be the shadow of king or chief, reflected and magnified on the mist of thought; for chief or king these peoples have none. This theory (Hume's) will not work where people have a great God but no king or chief; nor where they have a king but no Zeus or other supreme King-god, as (I conceive) among the Aztecs.
We now reach, in Mr. Tylor's theory, great fetish deities, such as Heaven and Earth, Sun and Moon, and 'departmental deities,' gods of Agriculture, War, and so forth, unknown to low savages.
Next Mr. Tylor introduces an important personage. 'The theory of family Manes, carried back to tribal Gods, leads to the recognition of superior deities of the nature of Divine Ancestor, or First Man,' who sometimes ranks as Lord of the Dead. As an instance, Mr. Tylor gives the Maori Maui, who, like the Indian Yama, trod first of men the path of death. But whether Maui and Yama are the Sun, or not, both Maori and Sanskrit religion regard these heroes as much later than the Original Gods. In Kamschatka the First Man is the 'son' of the Creator, and it is about the origin of the idea of the Creator, not of the First Man, that we are inquiring. Adam is called 'the son of God' in a Biblical genealogy, but, of course, Adam was made, not begotten. The case of the Zulu belief will be analysed later. On the whole, we cannot explain away the conception of the Creator as a form of the conception of an idealised divine First Ancestor, because the conception of a Creator occurs where ancestor-worship does not occur; and again, because, supposing that the idea of a Creator came first, and that ancestor-worship later grew more popular, the popular idea of Ancestor might be transferred to the waning idea of Creator. The Creator might be recognised as the First Ancestor, après coup.
Mr. Tylor next approaches Dualism, the idea of hostile Good and Bad Beings. We must, as he says, be careful to discount European teaching, still, he admits, the savage has this dualistic belief in a 'primitive' form. But the savage conception is not merely that of 'good = friendly to me,' 'bad = hostile to me.' Ethics, as we shall show, already come into play in his theology.