As regards the symbolic interpretation of ritual, this is usually held not to be primitive ; and it is doubtless true that an unreflec tive age is hardly aware of the difference between "outward sign" and "inward meaning," and thinks as it were by means of its eyes. Nevertheless, it is easier to define fetishism (q.v.) (a fetish "differing from an idol in that it is worshipped in its own char acter, not as the symbol, image or occasional residence of a deity," New Oxford Dictionary, Oxford, 190I) than it is to bring such a fetishism home to any savage people, the West African negroes not excluded (cf. A. B. Ellis, The Tshi-speaking Peoples of the Gold Coast of W. Africa, 192). It is the intrinsic mana, virtue or grace residing in, and proceeding from, the material object—a power the communicability of which constitutes the whole work ing hypothesis of the magico-religious performance—that is valued in those cases where native opinion can be tested. More over, it must be remembered that in the act of magic a symbolic method is consciously pursued, as witness the very formulas em ployed: "As I burn this image, so may the man be consumed," or the even more explicit, "It is not wax I am scorching; it is the liver, heart and spleen of So-and-so that I scorch" (W. W. Skeat, Malay Magic, 57o), where appearance and reality are distin guished in order to be mystically reunited.
Now it is important to observe that from the symbol as em bodying an imperative to the symbol as expressing an optative is a transition of meaning that involves no change of form what ever; and, much as theorists love to contrast the suggestional and the petitionary attitudes, it is doubtful if the savage does not move quite indifferently to and fro across the supposed frontier-line between magic and religion, interspersing "bluff" with blandish ment, spell with genuine prayer. Meanwhile the particular mean ings of the detailed acts composing a complicated piece of ritual soon tend to lose themselves in a general sense of the efficacy of the rite, as a whole, to bring blessing and avert evil. Nay, unin telligibility is so far from invalidating a sacred practice that it positively supports it by deepening the characteristic atmosphere of mystery. Even the higher religions show a lingering predi lection for cabalistic formulas.
scious evolution is perhaps the more active factor in primitive times ; imitation is never exact, and small variations amount in time to considerable changes.
On the other hand, there is also deliberate innovation. In Aus tralia councils of the older men are held day by day during the performance of their ceremonies, at which traditions are repeated and procedure determined, the effect being mainly to preserve custom but undoubtedly in part also to alter it. Moreover, the individual religious genius exercises no small influence. A man of a more original turn of mind than his fellows will claim to have had a new ceremony imparted to him in a vision, and such a cere mony will even be adopted by another tribe which has no notion of its meaning (Spencer and Gillen, ibid., 272, 278, 281). Mean while, since little is dropped whilst so much is being added, the result is an endless complication and elaboration of ritual. Side by side with elaboration goes systematization, more especially when local cults come to be merged in a wider unity. Thereupon assimilation is likely to take place to one or another leading type of rite—for instance, sacrifice or prayer. At these higher stages there is more need than ever for the expert in the shape of the priest, in whose hands ritual procedure becospes more and more of a conscious and studied discipline, the naïve popular elements being steadily eliminated, or rather transformed. Not but what the transference of ritualistic duties to a professional class is often the signal for slack and mechanical performance, with con sequent decay of ceremonial. The trouble and worry of having to comply with the endless rules of a too complex system is apt to operate more widely—namely, in the religious society at large —and to produce an endless crop of evasions.
Good examples of these on the part alike of priests and people are afforded by Toda religion, the degenerate condition of which is expressly attributed by Dr. W. H. R. Rivers to "the over development of the ritual aspect of religion" (The Todas, 454-5).
It is interesting to observe that a religion thus atrophied tends to revert to purely magical practices, the use of the word of power, and so on (ibid., ch. x.). It is to be noted, however, that what are known as ritual substitutions, though they lend them selves to purposes of evasion (as in the case of the Chinese use of paper money at funerals), rest ultimately on a principle that is absolutely fundamental in magico-religious theory; namely, that what suggests a thing because it is like it or a part of it becomes that thing when the mystic power is there to carry the suggestion through.