language. It is for this reason that there is no such thing as a wordless ritual; an apparent silence does not mean that inaudible incantations expressing the magician's will are not being made. From this point of view, the mechanical rite is but a translation of the unspoken incantation: a gesture is a sign, and also a language. Words and actions become absolutely equivalent and that is why we find descriptions of the non-verbal rites presented to us as spells. Without any formal physical movement a magician can create, annihilate, direct, hunt, do anything he wishes with the aid of his voice, his breath, or merely through his will.
The fact that all spells are formulas and that virtually all nonverbal rites also have their formulas shows at once to what degree all magic is formalistic. As for the spells themselves, there has never been any doubt that they are rites, since they are traditional, formal and clothed in an effectiveness which is sui generis. It has never been suggested that words can physically produce the desired effects. For non-verbal rites this fact is less obvious since there may be a very close parallel, often a logical one, between the rite and the desired effect. Obviously steam-baths and magical anointing have been able to relieve aicted persons. Nevertheless, the two types of rites have the same characteristics and lend themselves to similar observations. Both take place in an abnormal world.
Spells are composed in special languages, the language of the gods and spirits or the language of magic. Two striking examples of this kind of rite are the Malaysian use of bhàsahantu (spirit language) and the Angekok language of the Eskimoes. In Greece, Jamblique informs us that is the language of the gods. Magicians used Sanskrit in the India of the Prakrits, Egyptian and Hebrew in the Greek world, Greek in Latin-speaking countries and Latin with us. All over the world people value archaisms and strange and incomprehensible terms. From the very beginnings, practitioners of magic (and perhaps the
earliest are to be found in Australia) have mumbled their abracadabras.
The peculiar and weird nature of non-verbal ritual is paralleled by the enigmatic mutterings of spoken rites. Far from being the simple expression of individual emotions, magic takes every opportunity to coerce actions and locutions. Everything is fixed and becomes precisely determined. Rules and patterns are imposed. Magical formulas are muttered or sung on one note to special rhythms. Both in the Catapatha brâhmana and in Origen we find that intonation is sometimes more important than the actual words. Gestures are regulated with an equally fine precision. The magician does everything in a rhythmical fashion as in dancing: and ritual rules tell him which hand or finger he should use, which foot he should step forward with. When he sits, stands up, lies down, jumps, shouts, walks in any direction, it is because it is all prescribed. Even when he is alone he is not freer than the priest at his altar. Apart from this there are some general canons which are common to both spoken and mechanical rites: these involve numbers and directions. Movements and words must be repeated a number of times. Not any number of times, but according to sacred and magical numbers, such as 3, 4, 5, 7, 9, 11, 13, 20, etc. Moreover, words are pronounced or actions are performed facing a certain direction, the most common rule being that the magician should face the direction of the person at whom the rite is aimed. On the whole, magical ritual is extraordinarily formal and tends to become more and more so, to the point of extreme mystical preciosity-the tendency does not lie in the simplicity of everyday actions.
The simplest magical rites have the same form as those which are the object of the greatest number of definitions. Up to now, we have spoken of magic as if it consisted solely of positive actions. There are, however, many negative rites and these are precisely the simple ones which we now mention. We have already noted them in our list of the preparations performed
prior to a magical rite-the abstinence required of the magician and the other people involved is a case in point. They also include that huge mass of facts we call superstitions. They consist mainly of avoiding certain actions in order to prevent a magical effect. Here the rites are formal, and formal to a superlative degree, because the imperative character of the rite is almost complete. The kind of obligations associated with these rites show that they are the result of social forces even more than other rites which we have shown to be so because of their traditional, abnormal or formalist characteristics. However, on the important question of sympathetic taboos-negative magic as we prefer to call it-we have so far found so little enlightenment in the work of our predecessors, and also in our own researches, that we are hardly in a position to do more than call attention to the subject as offering an important field for study. For the moment we shall confine our attention to noting that these facts provide yet another proof that ritual-as an element in magic-is predetermined by collective forces.
As for positive rites, we have already pointed out how universally limited they are in each magical system. Their composition-and here we mean the combined effects of spells, negative ritual, sacrifices, recipes-is no less limited. There is a tendency for a limited series of stable complexes to grow up-we call them types of ceremonies-which are quite comparable either to technical patterns or what are known as art styles. There is a choice between available forms in each magical system; but once formulated we find the same clearly marked complexes over and over again, used for all kinds of purposes, no matter the logic of their composition. An example would be the variation on the theme of conjuring up the witch through the objects upon which she has cast a spell. When it is a matter of milk failing to produce butter, the milk in the vat is stabbed with a dagger, a custom which is also carried out to ward off many other kinds of evil. Here we have one type of magical rite and not
the only one which furnishes an example on this theme. We also have the medication of two or three dolls, which can be justified only by a similar proliferation. These actions through their persistence and their formalism are comparable to religious ceremonies.
Arts and crafts have styles which might be called tribal or, more precisely, national. In the same way it could also be maintained that each magical system has its own recognizable style which is characterized by the predominance of certain types of ritual: the use of dead men's bones in Australian sorcery, fumigation with tobacco in America, the benedictions and credos of the Muslims and Jews and other magical systems influenced by the religions of these last two. The Malays seem to have been the only people to know the curious ritual theme involving assemblies.
While it is true that different societies exhibit specific styles of magic, within each system-or, from another perspective, within each of the larger ritual groups we have described separately-there are important variations. The choice of the type of magic to be used is partly a matter for specialist magicians who use one rite or a limited number of rites in all the cases for which they are qualified. Each magician is bound by his materials, his tools and his medicine bag, which he inevitably uses every time he works. More often than not a magician is set apart from his colleagues by the type of rites he performs, rather than by the powers he possesses. We might add that the people we have called amateur magicians have an even more limited knowledge of rites and tend to go on repeating the same ones endlessly. This is the reason why certain formulas which are used over and over again, without rhyme or reason, end up by becoming completely unintelligible. Once again we find form taking precedence over content.
Nevertheless, what we have just said on the formation of styles in magical ritual does not mean that they are in fact classifiable.
Apart from the question of the existence of a vast amount offloating ritual, the fact remains that the development of special styles is quite a matter of accident and does not correspond to any real diversity of function, in magic there is nothing properly comparable to religious institutions.
Magical practices are not entirely without sense. They correspond to representations which are often very rich and which constitute the third element of magic. As we have seen, all ritual is a kind of language; it therefore translates ideas.
The minimum of display which each magical act involves is display of its effect. But this display, however rudimentary it may conceivably be, is already highly complex. It has several components, several levels. We can indicate at least some of these, so that subsequent analysis will not be purely theoretical; for there are magical systems which are perfectly conscious of their diversity and refer to it with special words and metaphors.
In the first place, we assume that magicians and their followers will not represent the special effects of the rite concerned without taking into account, at least unconsciously, the general effects of magic. Each magical rite seems to have arisen from some kind of syllogistic reasoning with the major term perfectly clearly expressed sometimes even in the spell: Venenum veneno vincituri natura naturam vincit. 'We know where you come from…. How can you kill like this?' (Atharva Veda, vii, 76, 5, vidma vai te… janam…Kaltham ha tatra tvám hano…). No matter how different the results of each rite, when they are working, they are thought to have common characteristics. In every case, in fact, we have either the imposing or suppressing of a characteristic or a circumstance: from being bewitched to being delivered, from possession to exorcization. In a simple word there has been a change of state. We are prepared to claim that all magical acts are
represented as producing one of two effects: either the objects or beings involved are placed in a state so that certain movements, accidents or phenomena will inevitably occur, or they are brought out of a dangerous state. The actions vary according to the initial state of the individual, the circumstances determining the significance of the change, and the special ends assigned to them. Nevertheless, they share one feature in that their immediate and essential effect is to modify a given state. The magician, therefore, knows full well that his magic is consequently always the same. He is always conscious that magic is the art of changing-the mâyâ as the Hindus call it.
However, apart from this quite formal conception, there are other more concrete elements behind a magical rite. Things come and depart: the soul comes back to the body, fever is driven away. An attempt is made to make sense of an accumulation of images. The bewitched person is ill, lame, imprisoned. Somebody has broken his bones, dried up his marrow, peeled off his skin. The favourite image is of something holding him, and it is tied or untied: 'an evil thread which has been maliciously tied', 'a chaining up which has been etched on the earth', etc. Among the Greeks the spell is, a . The same idea is expressed in Latin, although more abstractly, in the word religio, which has the same meaning. In a spell directed against a type of throat disease, we read, after an enumeration of technical and descriptive terms: Hanc religionem evoco, educo, excanto de istis membris, medullis (Marcellus, 15, 11); religio is here treated as a kind of vague being, with a diffuse personality that can be chased and caught. Elsewhere, the effects of the rite are expressed in ethical images: peace, love, seduction, fear, justice, ownership. This kind of representation, whose features are now and then vaguely glimpsed, may also be condensed into distinct notions and given special terms. The Assyrians expressed the idea through the word mâmit; in Melanesia, the equivalent of mâmit is mana, which is seen as a product
of the rite; among the Iroquois (Huron) the substance thrown by the magician is called orenda; in ancient India it was the brahman (neuter) which worked. We call it a charm, enchantment or a spell and the words we use show how unscientific the idea is. The idea is represented as a concrete, material object; a spell or a rune is thrown down, a charm is washed, drowned or burnt.
A third moment in our total representation comes when it is believed that there is a certain relationship between the persons and things involved in the ritual. This relationship may be conceived as a sexual one. An Assyro-Babylonian spell creates a kind of mystical marriage between demons and the images meant to represent them: 'People, evil and wicked ones who have taken off N., son of N., and will not let him go, if you are male let this be your wife, if you are female let this be your husband.' (C. Fossey, La Magie assyrienne, Paris, 1902, p. 133.) There are a thousand other ways of conceiving this relationship. It may be represented as something which is shared between the magicians and their victims. The magicians can be reached through their victim who thus has a hold over them. In the same way a spell can be removed by bewitching the magician who, for his part, naturally has control over his own spell. It is also said that it is the magician, or his soul, or his demon which has possessed the victim; in this way he gains control of his victim. Demoniacal possession is the strongest expression, simple spell-binding the weakest, with regard to the relations established between the magician and the subject of his rite. There is a distinct idea that there is a kind of continuity between the agents, the patients, the materials, the spirits and the end-object of a magical rite. Taking everything into consideration, we find the same idea in magic which we found in sacrifice. Magic involves a terrific confusion of images, without which, to our way of thinking, the rite itself would be inconceivable. In the same way that the central person in the sacrifice, the animal victim, god and the sacrifice itself become merged into one, the magician, the magical rite and its
effects give rise to a motley of indissociable images. Moreover, this very confusion may be the object of the representation. However separate the different moments in the representation of a magical rite may be, they also form part of a total representation whereby cause and effect become confused. Here we have the basic idea behind magical actions, an idea involving immediate and limitless effects, the idea of direct creation. It is the absolute illusion, the mâyâ as the Hindus so aptly named it. Between a wish and its fulfilment there is, in magic, no gap. This is one of its distinctive traits as we see in fairy tales. All those representations which we have been describing are only different forms, different moments, if you like, in this idea of magic. However, there do exist more straightforward representations of magic and these we shall attempt to describe now.
We shall classify these representations as impersonal and personal according to whether the idea of individual beings is involved or not. The first group may be further divided into abstract and concrete classes; the second group, of course, will be only concrete.
I Abstract, impersonal representations. The laws of magic Impersonal representations of magic involve laws which have been proposed, either implicitly or explicitly, at least in the case of alchemists and doctors. Over the past few years a great deal of importance has been given to these types of representation. It was believed that magic was dominated by them and the quite natural conclusion was reached that magic was a kind of science; when laws are involved we have science. In fact, magic gives every outward appearance of being a gigantic variation on the theme of the principle of causality. But this teaches us nothing, since it would be quite remarkable if it were otherwise, because magic's exclusive aim, apparently, is to produce results. On this subject we can only concede that if its formulas were simplified, it would be impossible not to consider magic as a scientific discip
line, a primitive science-and this is exactly what Frazer and Jevons have done. Magic can also function as a science and take the place of sciences not yet developed. The scientific character of magic has been observed throughout the world and has been consciously cultivated by magicians. This tendency towards a scientific orientation is naturally more obvious in the superior forms of magic, those which presuppose a body of acquired knowledge or refined techniques and which are performed in cultures where the idea of positive science is already present.
From the tangle of changing images it is possible to extract three principal laws. They could in fact all come under the heading of laws of sympathy, if antipathy is also covered by the notion of sympathy. These are the laws of contiguity, similarity and opposition: things in contact are and remain the same-like produces like-opposites work on opposites. E. B. Tylor and others after him have noticed that these laws are none other than the association of ideas (among adults, we would add), with one difference, that here the subjective association of ideas leads to the conclusion that there is an objective association of facts, or in other words that the fortuitous connexion between thoughts is equivalent to the causal connexion between things. If the three principles were to be combined into one, we could state that contiguity, similarity and contrariety equal simultaneity, identity and opposition, in both thought and deed. But we should be left wondering whether these formulas exactly reflect the way in which these so-called laws have been conceived.
Let us first look at the law of contiguity. The simplest expression of the notion of sympathetic contiguity is the identification of a part with the whole. The part stands for the complete object. Teeth, saliva, sweat, nails, hair represent a total person, in such a way that through these parts one can act directly on the individual concerned, either to bewitch or enchant him. Separation in no way disturbs the contiguity; a whole person can even be reconstituted or resuscitated with the aid of one of these parts:
totum ex parte. It is not necessary to give examples of beliefs which have become so well known by now. The same law may be expressed in another way: the personality of a being is indivisible, residing as a whole in each one of its parts.
This formula applies not only to people but also to things. In magic the essence of an object is found in a piece of it, as well as in the whole. The law is therefore, very general and concerns a property, which is attributed to both the individual's soul and the spirit essence of objects. This is not all: each object contains, in its entirety, the essential principle of the species of which it forms a part. Every flame contains fire, any bone from a dead body contains death, in just the same way as a single hair is thought to contain a man's life force. These observations lead us to believe that we are not only concerned with concepts involving an individual's soul; thus, the law cannot be explained by properties which are implicitly attributed to souls. Neither is it a complementary aspect of a life theory concerning life tokens. A belief in life tokens, on the contrary, is only a special case of totum ex parte.
This law of contiguity, moreover, comprises other features. Everything which comes into close contact with the person-clothes, footprints, the imprint of the body on grass or in bed, the bed, the chair, everyday objects of use, toys and other things, all are likened to different parts of the body. There is no need for contact to be permanent or frequent, or actually made-as in the case of clothes or objects of everyday use. A road, objects touched by mere accident, bath water, a fruit that has been bitten into, etc.-all can be used magically. Magic performed over the residues of meals-which is practised throughout the world-follows from the idea that there is a continuity and complete identity between the remains, the food consumed and the one who has eaten-the latter being substantially identical to the food partaken by him. A similar relationship of identity exists between a man and his family. It is through his relatives that he
can be harmed most effectively and it is always deemed a useful practice to name them in spells or to write their names on magical objects designed to bring him harm. The same relationship exists between a man and his domestic animals, his house, roof, fields, etc. There is also continuity between a wound and the weapon that caused it: a sympathetic relationship is involved in the curing of the wound through the intermediacy of the weapon. The same tie links a murderer and his victim. The notion of sympathetic continuity leads to a belief that the corpse bleeds when the assassin approaches it. It returns immediately to the state it was in at the time of the crime. The explanation of this phenomenon is a valid one, since we have several clear examples of this kind of continuity. It even spreads further than the guilty one. It was believed, for example, that if a man maltreated a robin-redbreast his cows would give forth red milk (Simmenthal, Switzerland).
As a result, we find that both individuals and objects are theoretically linked to a seemingly limitless number of sympathetic associations. The chain is so perfectly linked and the continuity such that, in order to produce a desired effect, it is really unimportant whether magical rites are performed on any one rather than another of the connexions. Sidney Hartland says that a girl, deserted by her lover, may make him suffer sympathetically by winding some of her hair round the feet of a frog or, alternatively, twining it into a cigar (Lucques). In Melanesia (the New Hebrides and the Solomon Islands, apparently) the friends of a man who has wounded another are placed in a position, as a result of the blow itself, of poisoning the victim's wound by magic.
The idea of magical continuity, realized through the relationship between parts and the whole or through accidental contact, involves the idea of contagion. Personal characteristics, illness, life, luck, every type of magical influx are all conceived as being transmitted along a sympathetic chain. Although contagion is
already one of the best known of all magical and religious notions, we shall spend a little time on the idea. In cases of imaginary contagion, a fusion of images is produced in exactly the same way as we found in sacrifice. The result is relative identification of the things and beings in contact. In a manner of speaking, it is the image of the thing to be displaced that runs along the sympathetic chain. This fact is often a feature of the rite itself. In India the victim must touch the magician at a certain moment in the main rite; in Australia the person being medicated has a thread or chain attached to him and the illness is chased along it. However, magical contagion is not only an ideal which is limited to the invisible world. It may be concrete, material and in every way similar to physical contagion. In order to diagnose maladies, Marcellus of Bordeaux advised patients to be sent to bed for a period of about three days with a puppy which had not yet been taken away from its mother. The patient had to give the dog milk from his own mouth and at frequent intervals (ut aeger ei lac de ore suo frequenter infundat), after which all that remained to be done was to open the dog's belly. Marcellus adds that the death of the dog cured the man. An exactly similar rite is practised among the Baganda of central Africa. In both cases the fusion of images is perfect. More than an illusion is involved-there is also hallucination. You can really see the illness leaving the person and being transmitted elsewhere. Here we have a transfer-rather than an association-of ideas.