In cases such as these, we are not dealing with simple matters of fraud. In general, the magician's simulations are of the same nature as those observed in nervous conditions. As a result, it is both voluntary and involuntary at the same time. Even when it starts off as a self-imposed state, the simulation recedes into the background and we end up with perfect hallucinatory states. The magician then becomes his own dupe, in the same way as an actor when he forgets that he is playing a role. Nevertheless, we must ask why he pretends like this. Here we must be careful not to confuse true magicians with those charlatans who turn up at fairs, or Brahman jugglers who brag to us about spirits. The magician pretends because pretence is demanded of him, because people seek him out and beseech him to act. He is not a free agent. He is forced to play either a role demanded by tradition or one which comes up to his client's expectations. It may appear that the magician vaunts his prowess of his own free will, but in most cases he is irresistibly tempted by public credulity. Spencer and Gillen found a host of people among the Arunta who declared they had taken part in magical excursions, known as kurdaitchas, where the liver fat of an enemy is 'removed'. A good third of the warriors have, as a result, had their toes disjointed, since this is a condition of the accomplishment of the rite. And the whole tribe declared they had seen, really seen with their own eyes, the kurdaitchas roaming their camps. In fact, most
of them were loath to remain outside all this atmosphere of 'fanfaronade' and adventure. The wish to 'encourage belief' was mutual and general throughout the social group, since credulity was universal. In cases of this kind, the magician cannot be branded as an individual working on his own for his own benefit. He is a kind of official, vested by society with authority, and it is incumbent upon the society to believe in him. We have already pointed out that the magician is appointed by society and initiated by a restricted group of magicians to whom society has delegated its power to create magicians. Quite naturally he assumes the spirit of his function, the gravity of a magistrate. He is serious about it because he is taken seriously, and he is taken seriously because people have need of him.
Thus, what a magician believes and what the public believes are two sides of the same coin. The former is a reflection of the latter, since the pretences of the magician would not be possible without public credulity. It is this belief which the magician shares with the rest, which means that neither his sleights of hand nor his failures will raise any doubts as to the genuineness of magic itself. And he himself must possess that minimal degree of faith-a belief in the magic of others, when he is a spectator or patient. Generally speaking, while he does not see the causes at work, he does see the effects they produce. Indeed, his faith is sincere in so far as it corresponds to the faith of the whole group. Magic is believed and not perceived. It is a condition of the collective soul, a condition which is confirmed and verified by its results. Yet it remains mysterious even for the magician. Magic as a whole is, therefore, an object a priori of belief, a belief which is unanimous and collective. It is the nature of this belief that permits magicians to cross the gulf which separates facts from their conclusions.
'Belief' implies the adherence of all men to an idea, and consequently to a state of feeling, an act of will, and at the same time a phenomenon of ideation. We are, therefore, correct in
assuming that this collective belief in magic brings us face to face with a unanimous sentiment and a unanimous will found in the community or, in other words, precisely those collective representations which we have been looking for. Some people, no doubt, will query the theory of belief we are putting forward and object that a single scientific error, naturally of an intellectual order, through its diffusion, may give birth to beliefs which in time become unanimous, beliefs which we can find no reason for not calling collective, yet which do not derive from collective forces, examples of such beliefs might include canonical beliefs in geocentrism and the four elements. We must now turn our attention to finding out whether magic depends entirely on ideas of this kind, ideas which cease to be doubted simply because they have become universal.
AN ANALYSIS OF IDEOLOGICAL EXPLANATIONS CONCERNING THE EFFECTIVENESS OF RITUAL
In examining magical representation, we have already considered those ideas by which magicians and theorists explain the efficacy of magical beliefs. These are: 1, sympathetic formulas; 2, the notion of property; 3, the notion of demons. We have already seen that these ideas are far from simple and continually overlap with each other. We shall now show how none of these ideas, by themselves, has ever been sufficient to justify a magician's belief. If we analyse magical ritual in order to reveal the practical application of these different notions, we shall always find that there is something left over, a residue, which the magician himself is also aware of.
Of course, no magician, and no anthropologist either, has ever attempted to reduce the whole of magic to one or other of these notions. This should put us on our guard against any theory which attempts to explain magical beliefs in these terms. We should also point out that, while magical facts do form a unique
category of facts, they usually depend on a single principle, which is alone capable of justifying these beliefs, of which they are the object. While it is true that each of these representations corresponds to a certain type of rite, the whole ritual ensemble must correspond to another representation which is quite general in nature. In order to determine what this may be, let us find out to what extent each of the notions enumerated above fail to explain fully the rites with which it is especially associated.
1. We hold that sympathetic formulas (like produces like; the part represents the whole; the opposite acts on its opposite) will not be sufficient to represent the totality of a rite of sympathetic magic. The remaining elements are not negligible. We shall consider only those sympathetic rites for which we have a complete description. The following ritual related by R. H. Codrington (The Melanesians, Oxford, 1891, pp. 200, 201) gives a fairly exact idea of their working:
In Florida the manengghe vigona, when a calm was wanted, tied together the leaves appropriate to his vigona and hid them in the hollow of a tree where water was, calling upon the vigona spirit with the proper charm. This process would bring down rain to make the calm. If sunshine was required he tied the appropriate leaves and creeper-vines to the end of a bamboo, and held them over a fire. He fanned the fire with a song to give mana to the fire, and the fire give mana to the leaves. Then he climbed a tree and fastened the bamboo to the topmost branch; as the wind blew about the flexible bamboo the mana was cast abroad and the sun shone out.
We have used this only as an example of a concrete illustration, since sympathetic rites are generally bound up in a complex contextual situation. For this reason, we must conclude that the symbols themselves are not sufficient to constitute a magical
rite. In fact, while a magician, such as an alchemist, sincerely imagines that his sympathetic practices are intelligible, he still expresses astonishment at the extent of the accretions accumulating around what was abstractly conceived to be the schema of the ritual. 'Why is it?', writes one alchemist, called the Christian, 'that there are so many books and evocations to demons? Why all this fabrication of furnaces and machinery when everything is so simple and so easy to understand?' Yet all the paraphernalia which surprised our Christian is not without its use. It is an expression of the fact that, along with the idea of sympathy, we also have the idea of the unleashing of power on the one hand and the magical milieu on the other.
There are quite a few indications of this notion of a power present during rites. First of all, there are sacrifices, which appear to have no other purpose than the creation of usable forces. We have already seen that this was one of the attributes of religious sacrifice. The same applies to prayers, invocations, evocations, etc., and also to negative rites, taboos, fasting, etc., which are a burden on the sorcerer or his client or sometimes on both, or indeed on their families, rites and ritual precautions which mark at the same time the presence and the fleeting nature of these forces. We should also take into account the powers belonging to the magician himself, powers which he carries with him, and the invocation of which is always at least possible. As for the sympathetic rite itself, we have already shown that the mere fact of its being ritualistic implies that it will necessarily produce, in turn, its own special forces. Magicians have always been conscious of this, in fact. In the Melanesian rite quoted above we saw how mana came out of the leaves and rose up to the sky. In Assyrian ritual, also mentioned above, mâmit is produced. Let us now consider a sympathetic rite in one of our so-called primitive societies, one which lacks mystical doctrines and where society still exists in the magical state. According to Frazer, the law of sympathy functions regularly and
on its own in these societies. We are immediately aware not only of the presence of forces but of their movements. Among the Arunta a sympathetic rite performed on an adulterous woman is thought to work the following way. An evil power, known as arungquiltha is, in effect, created. A stone-soul is charged with it (the image is used to fool the person's soul and persuade it to come to the rite as if it were still in its own natural body). The evil power is further activated by gestures which simulate the killing of the woman. It is this power which is finally thrown in the direction of the camp where the woman has been abducted. This rite provides an example of a situation where the sympathetic image is not even causal, since it is not the image which is thrown but the charm which the magician has just fabricated.
This is not the whole story. In the same case we find that, apart from the making of the image-where the soul is said to reside for a temporary period only-the rite also involves a collection of additional images which had previously been medicated-spirit stones, needles-and given their power well before the rite. Moreover, the ritual is performed at a secret spot, a spot validated by myth. We shall be bold enough to generalize from this example and conclude that sympathetic rites never occur in the same way as any ordinary act. They must take place in a special milieu, a milieu constructed by all the requisite magical conditions and practices. The milieu may be closed off by boundaries of taboos, and there are both entry and exit rites. Everything which enters this milieu belongs to the same nature as the sympathetic rite, or is endowed with it. The general tenor of all gestures and words becomes affected by it. The explanation of certain magical rites by reference to the laws of sympathy leaves us, therefore, with a twofold residue.
Does this apply in every possible case? We are inclined to believe that this residue is an essential part of magical rites. In fact, once all trace of mystery disappears we enter the realms of science and technology. This is precisely what our Christian
alchemist was trying to say. When he discovers that alchemy refuses to be scientific, he bids it become religious. If prayers are required, it is preferable that they be made to god rather than to the devil. This is to admit that alchemy and, as an extension, magic depend essentially on mystical powers. In cases where sympathetic formulas appear to be functioning on their own, we still find, accompanying the minimal form which every rite has, the presence of a minimal mysterious force-this is a matter of definition. Added to this, there is also the force of active property, without which, as we have already pointed out, there would be no way of properly conceiving a sympathetic rite. Moreover, we are still inclined to believe that so-called simple rites have been incompletely observed or have been incompletely performed, or else they have suffered a contraction which makes them useless as examples. As for the really simple rites, involving laws of sympathy, we shall call them sympathetic taboos. It is precisely these rites which best reveal the presence, the instability and the violence of those hidden spirit forces, the intervention of which, to our way of thinking, makes for the effectiveness of magical rites.
We have just seen that sympathetic formulas are never the complete formula of a magical rite. We could produce facts to show that, even when they are present in the clearest fashion, they are still only accessory elements. This is true of the practices of alchemists. They have always formally stated that their operations are rational deductions based on scientific laws. These laws, as we have seen, involve the notion of sympathy: one is the whole, the whole is in one, nature triumphs over nature. There are also special pairs of sympathies and antipathies, a whole complex system of symbols through which they order their operations-signs which are astrological, cosmological, sacrificial, verbal, etc. All this paraphernalia acts as a kind of fancydress for their techniques; it cannot even be considered as the imaginary principles of a false science. At the beginning of their
books, prefacing each chapter of their manuals, we find an exposition of their doctrines. The rest of the text, however, does not fit the introduction. The philosophical idea is prefixed in the manner of a caption, a heading or allegory, like the man of copper who was changed into gold by sacrifice. This quasi-scientific study can, in fact, be reduced to myths, myths which on occasion provide incantations. The same applies to their experimental precepts. There are algebraical formulas and schemas of actual operations, diagrams of apparatuses which once served a purpose but have since been transformed into unintelligible magical signs, no longer used for performing experiments; they are no more than power-inducing charms. Apart from such principles and formulas, of whose worth we are now perfectly aware, alchemy is an empirical study. It involves such activities as boiling, melting, vaporizing substances whose properties and reactions are understood empirically or traditionally. The term scientific is only a fancy title. The same once applied to medicine. Marcellus of Bordeaux headed a good number of chapters with such phrases as 'Remedia physica et rationabilia diversa de experimentis'. But immediately afterwards we read a sentence like this: 'Ad corcum carmen. In lamella stagnea scribes et ad collum suspendes haec', etc. (Marcellus, xxi, 2).
We may well conclude from the above that the formulas of sympathetic magic are not the laws of magical rites nor even those of sympathetic rites. They are but the abstract expression of very general notions which we have found to be diffused throughout magic. They are nothing more. Sympathy is the route along which magical powers pass: it does not provide magical power itself. In a magical rite the residue after the sympathetic formulas have been abstracted provides us with the essential elements of magic. If we take another example and look at those rites which Sidney Hartland described as sympathetic ritual by contact, the kind of spells by which a sorcerer dries up a woman's milk by kissing the child, we should like to stress
the fact that popular beliefs in spells such as these attach less importance to the idea of contact than to the evil eye, or the magical powers of the sorcerer or evil fairy.
2. We also claim that the idea of magical properties in themselves, even in cases where they predominate, cannot explain a belief in magical facts.
In the first place, the idea of properties is not the only element involved. The use of objects which have properties is usually prescribed by ritual. There are rules about the way they should be collected. Conditions of time, place, means, intention and so on have to be fulfilled wherever possible. A plant must be picked from the side of a river, by a crossroad, at the full moon, using two special fingers, with the left hand, approaching from the right, after going first here and then there, without thinking this or that, etc. And there are similar prescriptions for the collection of metals, animal products, etc. Finally, there are regulations regarding their use, the time, place, quantities involved, without going into the sometimes vast array of accessory rites which accompany them and which allow the utilization of their properties and the application of their sympathetic mechanisms. There are systems of magic-in India, for example-where every element involved in a magical rite, either as a secondary charm or an active substance, must be medicated or sacrificed beforehand.
In the second place, magical attributes are not conceived as being naturally, absolutely and specifically contained in the object to which they are attached; they are always relatively extrinsic and conferred. Sometimes this is achieved through ritual: sacrifice, blessing, bringing into contact with holy or polluted objects or other general sympathetic procedures. In other cases, the existence of the said property may be validated by myth, but even then it is considered accidental or acquired: such and such a plant grew in the footprints of Christ or Medea;
aconite flourished in the teeth of Echidna; Donnar's broom and the plant of the celestial eagle are magical objects whose qualities are not naturally inherent in the nut tree of the Hindu plant.
In general, magical properties, even an object's specific property, are considered to derive from characteristics which, from all the evidence, can only be regarded as secondary. This applies, for example, to the accidental shape of stones which resemble taros, pig's testicles and pebbles with holes in them. It is the colour of a lizard's head, in India, or a lump of lead, river foam, etc., which explains the connexion they have with evil substances. Other characteristics include an object's toughness, its name, its rarity value, its mysterious presence in a particular spot (a meteorite, prehistoric stone axes) or the circumstances of its discovery. The magical properties of an object derive from a kind of convention, a convention which plays the role of a sort of embryonic myth or rite. Anything which possesses magical properties, by its very nature, is a form of rite.
In the third place, the idea of properties plays such a relatively insignificant role in magic that it is always confused with very general ideas of power and nature. While people's idea of a desired effect is a very precise one, the idea of special qualities and their immediate action is always quite obscure. On the other hand, we do find very clearly in magic the idea of objects possessing infinite powers: salt, blood, saliva, coral, iron, crystals, precious metals, the mountain ash, the birch, the sacred fig, camphor, incense, tobacco, etc., all incorporate general magical forces susceptible of application or specific use. Moreover, the magician's attitude towards these properties is very commonly general and vague in the extreme. In India things have either a good or bad augury. Those with a good augury are the ones containing urjas (power), tejas (brightness), varcas (lustre, vitality), etc. For the Greeks and the moderns we also find holy, sacred and mysterious objects, which bring either good or bad luck. In sum, magic seeks philosophers' stones, cure-all, panaceas, divine waters.
Let us return to the alchemists, who developed a theory of magical powers based on sympathetic operations. These operations were, for them, forms, ???, of a generic nature, of nature, . If we break up the ???? we find the . However, as we have already stated, they were not concerned with an abstract conception of nature, but conceived it as a kind of essence, , or force, with vague spiritual properties, which nonetheless have a corporeal basis. Thus, once we are confronted with the idea of nature we also have the idea of force ,. Nature and force in their most abstract conception, are represented as a kind of impersonal soul, a power distinct from the objects themselves, yet one which is intimately bound up with them, understood though unconsciously. Before leaving the alchemists, we should remember that while the notion of spirit was found to be linked to the idea of properties, the converse is also true. Property and force are two inseparable terms. Property and spirit are often intermingled. The virtues of the pietra buccata come from the follettino rosso lodged there.
The idea of properties is also bound up with the idea of magical milieu. This is defined by negative or positive prescriptions involved in the use of things which we have already discussed. Finally, this representation is perfectly expressed in a certain number of traditions which imply that contact with a certain object immediately transports us into a magical world: magic wands, magic mirrors, eggs laid on Good Friday. Nevertheless, the residue left behind from the idea of properties, when we try to analyse magical ritual as the product and sum of these properties, is much smaller than it was in the case of sympathetic formulas. This is because the idea of property already partly expresses the idea of force and magical causality.
3. Demonological theory seems better able to account for rites in which demons figure. It even seems to provide a total explanation for rites which involve an appeal or a command addressed
to a demon. We could, at a pinch, extend the idea to the whole of magic, although it would be difficult to explain the basic nature of demoniacal rites through the notions of sympathy or magical properties. On the one hand, there are no magical rites which do not betray the presence of personal spirits to some extent, even if they are not necessarily specifically mentioned. On the other hand, the theory implies that magic has to operate within a special milieu, everything taking place in a world peopled by demons, or more precisely under such conditions that the presence of demons would be feasible. Finally, this theory clearly brings out one of the essential characters of magical causality-its involvement with spirits. Nevertheless, the theory has its drawbacks.
Demons cover only one part of the forces involved in a magical action, even in demoniacal rites. The idea of spirit beings is not a sufficient representation of anonymous general forces which are the basis of a magician's power, the strength behind his words and actions, the power of his looks and intentions, spells and death. This idea of a vague power, then, which we have covered as the residue of the other series of representations, is the total representation of a magical rite. It is so essential that magic has never been able to express its totality, in the form of demons, in a demoniacal rite. Something else must always be left over to explain at least the theurgical action of the rite on the demons, who possibly may be independent, but are not free agents. On the other hand, if the idea of spirits explains how a magician is able to act at a distance and how the ritual is multiple, it cannot explain either the existence of the ritual or its special features-sympathetic actions, magical substances, ritual prescriptions, private languages, etc. In fact, although demonological theory may suffice as an analysis of part of the residue remaining from other formulas, it is only explaining a part and it therefore also leaves a residue-consisting of everything which the other theories almost