Another fact which shows the importance of the notion of properties in magic is that one of the major preoccupations of magic has been to determine the use and the specific, generic or universal powers of beings, things, even ideas. The magician is a person who, through his gifts, his experience or through revelation, understands nature and natures; his practice depends on this knowledge. It is here that magic most approximates science. From this point of view, magic can be very knowledgeable even if it is not truly scientific. A good deal of the knowledge we have mentioned here has been acquired and verified through experiment. Sorcerers were the first poisoners, the first surgeons-we are aware that primitive surgery can be highly developed. Magicans made real discoveries in the field of metallurgy. However, unlike those theorists who have compared magic to science, because of the abstract representations of sympathy sometimes found in the former, it is because of the magician's speculations and observations on the concrete properties of things that we are willing to accord him the title of scientist. The laws of magic discussed above are really a kind of magical philosophy. They were a series of empty, hollow forms bringing in laws of causality which were always poorly formulated. Now, thanks to the notion of property, we have come across rudiments of scientific
laws, involving necessary and positive relationships thought to exist between certain objects. Owing to the fact that magicians came to concern themselves with contagion, harmonies, oppositions, they stumbled across the idea of causality, which is no longer mystical even when it involves properties which are in no way experimental. From this line of thinking they ended up deriving, in authentic fashion, special properties from words and symbols.
We affirm that each magical system has necessarily set up categories of plants, minerals, animals, parts of the body, dividing them into groups which do or do not have special or experimental properties. On the other hand, each system has set about codifying the properties of abstract things-geometrical figures, numbers, moral qualities, death, life, luck, etc. And the two sets of categories have been made concordant.
Here we come up against an objection: we are told that the laws of sympathy determine the nature of these properties. The properties of such and such a plant, for example, derive from the fact that the object or being on which it is supposed to act has the same-or different-colour. In this case, we must reply that far from there being any association of ideas between the two objects due to their colour, we are dealing, on the contrary, with a formal convention, almost a law, whereby, out of a whole series of possible characteristics, colour is chosen to establish a relationship between two things. Moreover, only one, or very few, of the objects having that colour are chosen to share this relationship. This is how it works among the Cherokee Indians when they choose their 'yellow root' to cure jaundice. This kind of reasoning, applied here to colour, can also be used with regard to form, contrariety and all other possible properties.
Furthermore, while it is clear that objects are vested with particular powers, by virtue of their names (reseda morbos reseda), we claim that things act more as incantations than as objects with properties, since they are really kinds of materialized
words. In these cases, the conventions we mentioned above come into greater evidence, since we are dealing with that most perfect of all conventions-a word, whose meaning, sound and everything about it, by definition, are all produced through tribal or national consensus. With a little difficulty we might also include in this argument the notion of magical keys, whereby the properties of things are defined through their relations with certain gods or certain things, which in fact represent their power (for example, the hair of Venus, Jupiter's finger, Ammon's beard, a virgin's urine, Shiva's liquid, an initiate's brain, the substance of Pedu). Here the convention which establishes the sympathetic relationship is a double one. First of all we have the convention determining the choice of name for the first sign (urine = Shiva's liquid) and then there is the other which determines the relationship between the named object, the second sign and the effect (Shiva's liquid = cure for fever since Shiva is the god of fever).
The sympathetic relationship is perhaps more apparent in the case of those parallel series of plants, perfumes and minerals which are said to correspond to planets. However, without commenting on the conventional nature of the attribution of the substances to each planet, we should at least take into consideration the convention which determines the virtues of the planets, virtues which are on the whole moral ones (Mars = war, etc). In summary, far from the idea of sympathy being the presiding principle in the formation of ideas concerning properties, it is the notion of property and the social conventions behind the objects which allow the collective spirit to link together the sympathetic bonds concerned.
In overcoming our self-imposed objection in this way, we do not wish to imply that the properties of an object are not a part of a system of sympathetic relationships. Quite the contrary: we hold that the facts we have just mentioned are of the utmost importance. They have been called 'signatures', that is, symbolic
correspondences. They provide, in our opinion, examples of a classification similar in many ways to those studied in Année Sociologique [1948-9?]. Things which are grouped together under this or that astral sign belong to the same class, or rather to the same family, as the astral body, its region, its mansions, etc. Those which have the same colour, the same shape and so on are believed to be related because of their colour, their shape, their sex. The grouping of things by opposites is also a method of classification. It is really a way of thinking which is basic to all magical systems, that is, the division of everything into at least two groups: good and evil, alive and dead. In this way, the system of sympathetic and antipathetic magic can be reduced to one of classifying collective representations. Things affect each other only because they belong to the same class or are opposed in the same genus. It is because they are members of one and the same family that things, movements, beings, numbers, events, qualities gain a reputation for being similar. It is also because they are members of the same class that one can act on another, it being held that a similar nature is common to a whole class, in the same way as the same blood is held to diffuse throughout an entire clan. As a result, they are involved in relations of similarity and continuity. Furthermore, from class to class we find oppositions. Magic becomes possible only because we are dealing with classified species. Species and classifications are themselves collective phenomena. And it is this which reveals both their arbitrary character and the reason why they are limited to such a small number of selected objects. In fact, when we are dealing with representations of magical properties we find ourselves in the presence of phenomena which are comparable to those of language. Just as no object has an infinite number of names, so, with regard to things, the number of signs is restricted. And just as words have only a distant relationship, or none at all, with the things they describe, between a magical sign and the object signified we have very close but very unreal relations-of number,
sex or image, qualities which in general are quite imaginary, but imagined by society as a whole.
In magic there are other representations-both impersonal and concrete-besides those of properties. They include representations of the power of the ritual, and its methods of action. We mentioned this earlier in our discussion of the general effects of magic, pointing out the concrete forms of such notions as mâmit, mana, effluvia, chains, lines, jets of water, etc. We also find representations of the magician's power and his methods of action, also mentioned earlier when discussing the subject of the magician himself: the power in his look, his strength, his presence, his invisibility, his insubmersibility, his power of transportation, his ability to act directly from a distance, etc.
These concrete representations, along with other abstract representations, provide us in themselves with a conception of the magical rite. There are, in fact, numerous rites which consist of no more definite representations than these. The fact that they are sufficient in themselves perhaps provides justification for persons who see magic as the direct working of ritual and who relegate to a subsidiary role those demonological representations which to us, at least, are necessarily found in all magical systems.
3 Personal representations. Demonology There is no real discontinuity between those ideas involving spirits and the concrete and abstract ideas we have just been discussing. Between the idea of the spirituality of magical action and the idea of the spirit there is only a small gap to breach. The idea of a personal agent, from this point of view, could be considered as the product of the effort made by the magical efficacity of rites and their qualities, in order for their expression to be represented concretely. In fact, it has happened that demonology has been considered a means of expressing magical phenomena: the miasmas are devils, . The idea of the demon, from this point of view, is
not inconsistent with other notions. It is in a way a supplementary idea useful in explaining the play of laws and properties. Here we have a simple substitution of the person as a causal factor for the idea of magical causality.
All magical representations may have personal representations. The magician's double and his animal auxiliary are personal expressions of his power and the way his actions work. Some Ojibway pictograms show it as a manitou of the Jossakîd. In the same way the miraculous sparrow-hawk which carries out Nectanebo's orders represents his magical powers. In both these cases the attendant demon or animal is the personal and effective agent of the magician. Through it he acts from afar. In the same way the power of the rite may be personalized. In Assyria, the mâmit is like a demon. In Greece the that is, the magical wheel, has conjured up demons, and so do certain magical formulas, such as the Ephesia grammata. The idea of properties works the same way. Plants with special virtues are linked to demons who cure as well as bring diseases. We find demons of vegetation of this kind in Melanesia and among the Cherokee, as well as in Europe (the Balkans, Finland, etc.). The bathing devils on Greek vases derive from the practice of using objects from baths in spells. We see from this example that personification can be associated with a minor aspect of the rite. It may equally well be applied to the most general aspects of magical power. In India, Shakti, power, is deified. Obtaining such powers, siddhi, is also deified. Siddhi is invoked, in the same way as the Siddha, those who have obtained it.
Personification is not limited to these examples. Even the subject of the rite may be personified by its ordinary name. This applies first of all to illnesses: fever, fatigue, death, destruction, anything in fact which is exorcized. An interesting story could be told of that doubtful divinity of Atharvanic ritual who is known as the goddess Diarrhoea. Naturally, we find the growth of this phenomenon in the system of incantations, particular
evocations, rather than in purely non-verbal rites, although they may exist here and remain unobserved. In incantations, the illness which the magician is trying to drive away is addressed and in this way treated as a person. For this reason almost all Malaysian formulas are conceived of as invocations addressed to princes and princesses who are no less than the objects and phenomena involved in the rite. Elsewhere, in the Atharvaveda for example, everything magically medicated becomes really personified: arrows, drums, urine. Here we are concerned with something more than a language form. These people are more than simple vocatives. They existed both before and after the incantations were made. They include the Greek , the genies of illness in Balkan folklore, Laksmî (fortune) and Nirrti (destruction) in India. The latter even have their own mythology like other illnesses personified in the majority of magical systems.
The introduction of the idea of spirits does not necessarily modify the magical rite. In the main, spirits in magic are not a free force. They must simply obey the rite, which indicates how they should go about their work. It is, therefore, a possibility that nothing betrays their presence and that they need not even be mentioned in spells. All the same, it often happens that the spiritual auxiliary does play its part, and sometimes a big part, in magical ceremonies. In some, the image of the auxiliary animal or genie is conjured up. We find in ritual and prayers hints of offerings and sacrifices which have no other object than to evoke and satisfy the demands of personal spirits. If truth be told, these rites are frequently supererogatory in relation to the central rite, the schema of which always remains symbolic or sympathetic in its principal lines. Yet they are sometimes so important that they entirely swamp the ceremony. Thus, an exorcism rite may be encapsulated in a sacrifice or a prayer addressed to a demon who is to be driven out, or to a god who drives him away.
In dealing with these kinds of rites, it would be true to say that
the notion of the spirits is the pivot on which they turn. It is obvious, for example, that the idea of a demon will take precedence over all other thoughts in the mind of the officiant, when, for example, he addresses a god in Greco-Egyptian magic in order to beg him to send a demon to work for him. In cases like this, the idea of the rite, along with everything involving automatic inevitability, fades into the background. The spirit is an independent servant and assumes, in magical practices, the role of chance. The magician ends up admitting that his science is not an infallible one and that his will is not necessarily accomplished. He is dealing with a power. In the same way the spirit is both a subject and free form, merged with the ritual and separate from the ritual. Here we seem to be in the presence of one of those antinomian confusions which abound in the history of both magic and religion. The understanding of these apparent contradictions belongs to the theory of the relationship of magic and religion. However, we can state here that the most common magical facts include those where the ritual seems to constrain. But we do not wish to deny the existence of other facts, which will be explained elsewhere.
What are magical spirits? We shall attempt a very summary explanation, a very rapid enumeration, in order to show how magic has recruited its bands of spirits. We shall at once find that these spirits have other than magical qualities and that we also come across them in religion.
Our first category of magical spirits embraces the souls of the dead. Magical systems even exist which-either from the beginning or through a process of reduction-have only this kind of spirit. In western Melanesia, people have recourse to spirits known as tindalos-all of whom are spirits of the dead-in both their magic and their religion. Any dead person may become a tindalo, if he proves his power through the performance of a miracle, a maleficent action, etc. However, in principle, only people who possessed religious or magical powers when alive
may become tindalos. The dead, in this case, provide the spirits. They do so in Australia and in America as well, among the Cherokee and Ojibway. In ancient and modern India the dead, deified ancestors are invoked during magical ceremonies. In spells, however, the spirits are invoked of dead persons for whom funeral rites have not yet been appropriately performed (preta), of those who have not been buried, of those who died a violent death, of women who died in childbirth, of the spirits of stillborn children (bhûta, churels, etc.). In ancient Greece, the , that is, magical spirits, are given names which indicate that they are souls of the dead. We even have the occasional mention of or , although they are most frequently those of demons who died a violent death () or who were not buried , etc. In Greek areas another type of dead person provides magical auxiliaries: these are the heroes, that is, the dead who form the object of a public cult. It is not clear, however, whether all the magical heroes were official ones. On this point the Melanesian tindalo may be reasonably compared with the Greek hero, since although he may be a person who was not deified after death, he is necessarily conceived in this way. In Christianity the dead all have properties which may be useful, qualities deriving from their death. However, magic makes use only of the souls of non-baptized children, those who died violent deaths and dead criminals. Even such a short exposition shows that the dead are magical spirits either by virtue of a general belief in their divine powers, or because of their special qualities which, in the phantom world, give them a special role in relation to religious beings.
A second category of magical beings embraces demons. The word 'demon', of course, is not used here as a synonym for devil, but for words such as genie and djinn. Demons are spirits: on the one hand, they are distinguishable from the souls of the dead, and on the other, they are those who have not yet attained
the divine nature of gods. In character they are rather tame, yet they sometimes represent something more than a simple personification of a magical rite, a magical property or object. All over Australia, it seems that they were thought of as distinct in form and even when we have sufficient information about them they still appear quite specialized. Among the Arunta we find magical spirits, Orunchas and Iruntarinias, who are really local genies and who are set apart by their relatively complex nature. In eastern Melanesia spirits are invoked which are not souls of the dead, nor are they all gods, properly speaking. They are spirits which play a considerable role, especially in nature rites: vui in the Solomon Islands, vigona in Florida, etc. In India, along with the Devas, the gods, we have Pisâkas, Yâksasas, Râkshasas, etc., and the whole group, from the moment we have a classification, forms the category of Asuras, the main personalities of which include Indra's rival Vrtra and Namuci. We all know that Mazdaism considered, on the contrary, that the Daevâs, servitors of Ahriman, were the adversaries of Ahura Mazda. Now and then, in these two cases, we come across specialized magical beings. They are evil genies, it is true. Nevertheless, their very names betray the fact that there was never a radical distinction between them and the gods-at least in the beginning. Among the Greeks these magical beings were the , which, as we saw, were similar to the souls of the dead. The specialized nature of these spirits was such that magic was defined in Greek by reference to its relations with demons. There are demons of all kinds, of both sexes, of all shapes and sizes-some are localized on earth, others people the atmosphere. Some are given proper names, although they are all magical. The fate of all the Sxiuoves was to become evil genies and go and live among the Cercopes, Empusae, Keres, etc., as a category of evil spirits. Furthermore, Greek magic showed a marked preference for Jewish angels, particularly the archangels and the same applies to Malayan magic. Eventually all these archangels, archontes, demons, eons, etc. formed a genuine
pantheon of hierarchic magic. This was later inherited by medieval magic, in the same way as the whole of the Far East inherited the magical pantheon of the Hindus. Demons, however, were changed into devils and set up alongside Satan-Lucifer, on whom all magic depended. Nevertheless, in the magic of the Middle Ages and also that of our own times, in places where the old traditions have been preserved, we find that our system has other genies, fairies, sprites, goblins, kobolds, etc.
Magic, however, need not address itself to specialized genies. In fact, the different classes of spirits we have just mentioned were not always exclusively magical in nature. And once they have become magical, this does not mean they do not relinquish the religious role: we do not consider Hell as a magical idea. On the other hand, there are countries where the functions of gods and demons are not distinguishable from each other. This is the case all over North America, where Algonquian manitous constantly change from one to the other. The same occurs in eastern Melanesia where the tindalo behave in the same way. In Assyria, there are whole series of demons which may or may not be gods. In the scriptures their names usually have divine affixes. The main ones include the Igigi and the Anunnaki, whose identity is still somewhat mystifying. All in all, demoniacal activities are not at all incompatible with those of the gods. Moreover, the existence of specialized demons does not mean that magic cannot make use of other spirits, endowing them, at least for the time being, with a demoniacal function. And we find gods-and in Christian magic, saints-who crop up as spiritual auxiliaries of the magicians. In India, the gods play their part even in black magic, in spite of the degree of specialization which has developed there, and they play essential roles in all other magical ritual. In countries which have undergone Hindu influence, Malaysia and Câmpa (Cambodia), the entire Brahman pantheon figures in their magic. As for the Greek magical texts, they first of all mention a host of Egyptian gods, either by their Egyptian or
their Greek names, as well as Assyrian and Persian gods, Iahwé (Jehovah) and the whole gamut of Jewish angels and prophets-they are all gods who were outside Greek culture. However, they also used the 'high gods', in their form, referring to them by their Greek names, Zeus, Apollo, Aesculapius, even associating them with their particular localities. In Europe, the Virgin, Christ and the saints are the only spirits which appear in most of the spells and particularly in the mythical charms.
Personal representations in magic have presented a sufficient consistency for myths to have grown up. Mythical charms, of the kind just mentioned, depend on myths pertaining to magic. There are others which explain the origin of magical tradition, of sympathetic relations, ritual, etc. However, while magic may have its myths, these are only rudimentary and very specific in nature, dealing only with things rather than with spiritual beings. Magic has little poetry. We do not find many stories about its demons. Demons are like soldiers in an army, they are troops, ganas, bands of hunters or cavalcades; they lack any real individuality. This applies even more to the gods which have become involved in magic. They are stripped of personality leaving-if we may be allowed to express it this way-their myths on the doorstep. Magic is not interested in them as individuals, but as wielders of properties, powers whether generic or specific in nature. Moreover, they may be transformed to suit a magician's purpose, and are often reduced to mere names. In the same way that spells can invoke demons, so the gods may also end up as nothing more than mere incantations.
The fact that magic has made use of gods shows that it has been able to take advantage of the obligatory beliefs of society. It is because gods were believed in that magic used them for their own ends. But demons, along with gods and the souls of the dead, are objects of collective representations, which are often obligatory and sanctioned, at least in ritual. This is the reason why they became magical forces. In fact, each magical system