succeeded in explaining. Thus, in any demoniacal rite the idea of spirits is necessarily accompanied by an impersonal notion of efficacious power.
We may ask whether the idea of power does not itself derive from the idea of spirits. This is a hypothesis which nobody has so far maintained. Nevertheless, it is a logical possibility in a strictly animist theory. A first objection would be that a spirit, in magic, is not, of necessity, an active being. All exorcizing ritual, curative spells and those charms we call origin charms have no other function than to put to flight a spirit whose name, history and activities are pointed out to them. The spirit here is in no way the cornerstone of the rite; it represents simply the object of the rite.
Finally, we should take care not to exaggerate the importance of the idea of persons, even within this class of demoniacal representations. We have said that there are demons who amounted to nothing outside those properties and rites which they so imperfectly personify. In describing them, little else is involved other than the idea of an influence and the passing on of effects. They are , euvia. Even the names of Hindu demons show to what a limited extent they ever attained any individuality: siddhas (those who have obtained power) vidyâdhâras (bearers of learning), the names of 'Prince Siddhi, Prince Shakti' (power) have persisted in Moslem Malay magic. Algonquin manitous are also quite impersonal. This fact also comes out in the frequent vagueness as to the number and the names of the demons involved. They usually form a body of troops, a host of anonymous beings (mobs, ganas), often called by all kinds of collective names. We query even whether these classes of demons ever involved real people at all-apart from the souls of the dead, who are themselves rarely identified, and the gods.
We not only hold that the notion of spirit power does not derive from the notion of magical spirit, but we have reason to
believe that the latter derives from the former. The idea of spirit power, in fact, leads us to the idea of spirit. We find that the Assyrian mâmit, the Algonquin manitou and the Iroquois orenda may all be called 'spiritual', without losing any of their characteristics of general power. On the other hand, is it not a reasonable supposition to imagine that the idea of the magical spirit is the sum of two notions: that of the spirit and that of magical power, the latter not necessarily to be considered as an attribute of the first? Proof of this may be found in the fact that, among the dense crowd of spirits with which society peoples its universe, there are only a limited few that are recognized-through experience, so to speak-as powerful beings and hence involved in magic. This may explain the tendency to bring gods into the system, particularly foreign gods or rejected ones, gods who are, by definition, powerful beings.
Although we were first inclined to favour the animist explanation of magical beliefs over all other theories, we have now noticeably departed from the common animist hypothesis, in that we consider the idea of spiritual force to have preceded the idea of the soul, at least as far as magic is concerned.
To sum up, the various explanations which can be brought forward as motives for beliefs in magical acts leave a residue, which we must now try to describe, in the same way as we described the various elements of magic. And we have reason to believe that it will be here that we shall find the real basis of these beliefs.
We have thus come nearer to determining this further element which magic superimposes on its impersonal notions and its ideas of spirits. At this stage, we hold that it is an element which is superior to these two orders of ideas, and one from which-if it is presented-the others are merely derivative.
It is a complex notion, involving first of all the idea of power, or as it has been rather better described, 'magical potential'. It is the idea of a force of which the force of the magician, of the
ritual and of the spirit are merely different expressions, in accordance with the elements of magic. The fact is that none of these elements acts as such, but does so precisely inasmuch as it is endowed, either by convention or by special rites, with this character of being a force, and not a mechanical force, but a magical one. The idea of magical force is moreover, from this point of view, quite comparable to our notion of mechanical force. In the same way as we call force the cause of apparent movements so magical force is properly the cause of magical effects: illness and death, happiness and health, etc.
This idea also includes the notion of a milieu, where the powers in question exist. In this mysterious milieu, things no longer happen in the way they do in our world of the senses. Distance does not prevent contact. Desires and images can be immediately realized. It is the spiritual world and the world of the spirits at the same time. Since everything is spiritual, anything may become a spirit. Yet although this power is illimitable and the world transcendental, things happen according to laws, those inevitable relations existing between things, relations between signs and words and the represented objects, laws of sympathy in general, laws of properties which are susceptible to being codified into a system of classifications of the same type as those which have been studied in Année Sociologique. The ideas of force and milieu are inseparable, coinciding in an absolute sense. They are expressed at the same time and through the same means. In fact ritual forms, those dispositions aimed at creating magical forces, are also the same as those which create the milieu and circumscribe it before, during or after the ceremony. If our analysis is exact, therefore, we shall find-at the basis of magic-a representation which is singularly ambiguous and quite outside our adult European understanding.
It has been through the discursive processes of similar individual judgments that the science of religion has so far attempted to explain magic. In fact, the theory of sympathetic
magic depends on analogical reasoning or-which amounts to much the same-the association of ideas. Demonological theory refers to individual experiences of consciousness and of dreams. The representation of properties is usually conceived as resulting from experience, from analogical reasoning or from scientific error. This composite idea of force and milieu, on the other hand, avoids these rigid and abstract categories, which our language and reasoning impose. From the point of view of an individual's intellectualist psychology, it would be an absurdity. Let us see whether a non-intellectualist psychology of man as a community may not admit and explain the existence of this idea.
A similar notion exists, in fact, in a certain number of societies. By a logical reversal the fact that it exists, is named and is already relatively differentiated in two of our ethnic groups-which we shall use as special examples-provides confirmation of our analysis.
The idea is that found in Melanesia under the name of mana. Nowhere else is it so clearly evident and it has fortunately been admirably observed and described by Codrington (The Melanesians, p. 119 et seq., p. 191 et seq., etc.). The word mana is common to all Melanesian languages proper and also to the majority of Polynesian languages. Mana is not simply a force, a being, it is also an action, a quality, a state. In other terms the word is a noun, an adjective and a verb. One says of an object that it is mana, in order to refer to this quality; in this case the word acts as a kind of adjective (it cannot be said of a man). People say that a being, a spirit, a man, a stone or a rite has mana, 'the mana to do such and such a thing'. The word mana is employed in many different conjugations-it can be used to mean 'to have mana', 'to give mana', etc. On the whole, the word covers a host of ideas which we would designate by phrases such as a sorcerer's power,
the magical quality of an object, a magical object, to be magical, to possess magical powers, to be under a spell, to act magically. The single word embraces a whole series of notions which, as we have seen, are inter-related, but which we have always represented as separate concepts. It reveals to us what has seemed to be a fundamental feature of magic-the confusion between actor, rite and object.
The idea of mana is one of those troublesome notions which we had thought to have discarded; we therefore experience difficulty in grasping it. It is obscure and vague, yet the use to which it is put is curiously definite. It is abstract and general, yet quite concrete. Its primitive nature-that is, its complexity and confusion-resists any attempt at a logical analysis, and we must remain content to describe the phenomenon. According to Codrington, it invades all magical and religious rites, all magical and religious spirits, the totality of persons and things involved in the totality of ritual. It is really mana which gives things and people value, not only magical religious values, but social value as well. An individual's social status depends directly on the strength of his mana, and this applies particularly to roles in secret societies. The importance and inviolability of property taboos depend on the mana of the individual who imposes them. Wealth is believed to be the result of mana. On some islands mana is the word for money.
The idea of mana consists of a series of fluid notions which merge into each other. At different times it may be a quality, a substance or an activity. First mana is a quality. It is something which possesses the thing called mana, not the thing itself. It is described as being 'powerful' or 'heavy'. At Saa it is 'hot', at Tanna it is something strange, indelible, resistant, extraordinary. Secondly mana is a thing, a substance, an essence that can be handled yet also independent. That is why it may only be handled by individuals who possess mana during a mana action, that is, by qualified individuals during the course of a rite. By its
nature it is transmissible, contagious: mana may be communicated from a harvest stone to other stones through contact. It is represented as a material body. It may be heard and seen, leaving objects where it has dwelt. Mana makes a noise in the leaves, flies away like a cloud or flame. It can be specialized: there is mana to make people wealthy and mana used to kill. Generic forms of mana may be defined even more narrowly. In the Banks Islands there is a special kind of mana, the talamatai, for certain methods of making incantations, and another for casting spells over the traces of an individual. Thirdly mana is a force, more especially the force of spirit beings, that is to say, the souls of ancestors and nature spirits. It is mana which creates magical objects. However, it is not indiscriminately inherent in all spirits. Nature spirits are essentially endowed with mana, but all the souls of the dead are not. Tindalos are active spirits-the souls of dead chiefs, for the most part family heads, and more particularly those in whom mana has manifested itself either during their lifetime or through the performance of miracles after their death. Only these merit the name of powerful spirits, the others being lost among a multitude of impotent shades.
Once again we have an example of the fact that while all demons are spirits, not all spirits are demons. The idea of mana, then, is not to be confused with the idea of spirit. They are closely linked, yet remain profoundly separate. Consequently, there is no possibility of explaining (at least in Melanesia) demonology, and hence magic, through animism alone. Take the following as an example. In Florida, when a man is ill the sickness is explained by the fact that mana has him in its grasp. This mana belongs to a tindalo who is himself associated with a magician (manekisu-endowed with mana) who has the same mana or the mana to act on it, which amounts to the same thing. On the other hand, the tindalo is also associated with a plant. There are certain plant species attached to different kinds of tindalos which, through their mana, cause certain illnesses. The tindalo concerned
is discovered by the following means. The leaves of different species of plants are collected and rubbed between the fingers one by one, the one which contains the mana of the illness aicting the sick person is recognized by a special rustling sound. Now they can confidently call in the tindalo, or the mane kisu who possesses the mana of the tindalo, that is, the individual who is related to the spirit and who is alone empowered to remove the mana from the patient and bring about his cure. Here, in fact, the mana is separable from the tindalo since it is found not only in the tindalo itself, but also in the sick person, the leaves and the magician too. Mana, therefore, exists and functions independently. It remains an impersonal force, alongside the personal spirit. The tindalo contains mana but is not mana itself. Note, in passing, that this mana circulates within a classificatory category and that the things which act upon one another are encompassed within this category.
Mana, however, need not be the power possessed by a spirit. It may be the force of a non-spiritual object, such as a stone for making taros grow or for rendering sows fertile, or a plant which brings rain. But it is a spiritual force in so far as it does not work mechanically and can produce its effects from a distance. Mana is the magician's force. The names of those specialists who perform magic are almost everywhere composed from the word: peimana, gismana, mane kisu, etc. Mana is the power of a rite. The word mana is even applied to magical formulas. However, the rite is not only endowed with mana, it may be mana itself. It is because the magician and rite possess mana that they are able to act upon spirits with mana, evoke them, give them orders, possess them. Therefore, when a magician has a personal tindalo, the mana which he uses to act upon his tindalo is not really different from the mana which makes the tindalo function. While there is an infinity of tindalos, we have come to believe that the different manas are but one and the same power, not fixed in any way but simply shared out among beings, men or spirits, objects, events, etc.
We could extend still further the meaning of this word and maintain that mana is power, par excellence, the genuine effectiveness of things which corroborates their practical actions without annihilating them. This is what causes the net to bring in a good catch, makes the house solid and keeps the canoe sailing smoothly. In the farms it is fertility; in medicine it is either health or death. On an arrow it is the substance which kills and, in this case, it is represented by a piece of bone from a dead man which is incorporated in the arrow shaft. And it is a fact that European experts have shown the Melanesian poisoned arrow to be simply a magically medicated arrow-the arrow with mana. However, they are believed to be poisoned, but it is clear that it is the mana and not the arrow point to which they attribute the actual effectiveness of the arrow. It is the same in the case of demons-again mana appears to be distinct from the tindalo, working like a quality attached to an object, without prejudicing its other qualities, in other words, like something superimposed on another. This extraneous substance is invisible, marvellous, spiritual-in fact, it is the spirit which contains all efficacy and all life. It cannot be experienced, since it truly absorbs all experience. The rite adds it to things, and it is of the same nature as the rite. Codrington thought he could call it the supernatural, but then he more correctly says that it is only supernatural 'in a way', that is to say, that mana is both supernatural and natural, since it is spread throughout the tangible world where it is both heterogeneous and ever immanent.
This heterogeneity is always apparent and sometimes manifested in action. Mana is separate from the common world of mortals. It is the object of a reverence which may amount to a taboo. We might add that all taboo objects must contain mana and that many mana objects are taboo. As we have mentioned, these include the mana of a property owner or the tindalo which endows the property taboo with power. There is even reason to believe as well that the place where spells are made, the stones where
tindalos dwell-places and objects with mana-are taboo. The mana of a spirit-dwelling stone will affect any person who walks over the stone or whose shadow crosses it.
Mana is, therefore, seen to be something both mysterious and separate. In sum, mana is first of all an action of a certain kind, that is, a spiritual action that works at a distance and between sympathetic beings. It is also a kind of ether, imponderable, communicable, which spreads of its own accord. Mana is also a milieu, or more exactly functions as a milieu, which in itself is mana. It is a kind of internal, special world where everything happens as if mana alone were involved. It is the mana of the magician which works through the mana of the rite on the mana of the tindalo, and which sets other manas in motion and so forth and so on. In its actions and reactions there are no other forces involved apart from mana. It is produced in a closed circuit, in which everything is mana and which is itself mana, if we may so express it.
The same idea crops up in places outside Melanesia. We find certain indications of it in a number of societies where further research would not fail to uncover it completely. First and foremost, it is widespread among speakers of other Malayan-Polynesian languages. Among the Straits Malays, it is known by a term of Arabic origin with a Semitic root, which has a somewhat more restricted sense-kramât (W. W. Skeat's transliteration) from hrm which means sacred. Things, places, moments, animals, spirits, men, sorcerers are kramât or have kramât; and it is the forces of kramât which are active. To the north, in French Indo-China, the Ba-hnars express a similar idea to mana, when they say that the witch is a deng person, who has deng, who can deng things. They apparently speculate endlessly on the notion of deng. At the other extreme of the dispersal of Malayan-Polynesian languages, in Madagascar, we have the term hasina-of unknown etymology-which refers at one and the same time to the quality of certain things, an attribute of some beings-animals, men and, in particular, the queen-as well as the ritual controlling
these qualities. The queen was masina, she had masina and the tribute presented to her, together with the oath sworn in her name, was hasina. We are convinced that a closer analysis of New Zealand magic where mana plays a role-and even of the Dayaks, where the medicine man is called manang-would provide similar conclusions to the studies carried out in Melanesia.
The Malayan-Polynesian world can claim no monopoly of these concepts. In North America we find the same in certain regions. Among the Huron (Iroquois) it is called by the name orenda. Other Iroquois seem to have called it by a term which has the same root. J. N. B. Hewitt, a Huron by birth, and a distinguished ethnographer, has given a valuable description-a description rather than an analysis, since the orenda is no easier to explain than mana. (American Anthropologist, 1902, new series, 4, i, pp. 3246.)
The idea is too general and too vague, too concrete, covering so many things and so many obscure qualities that it is only with difficulty that we can begin to understand it. Orenda is power, mystical power. There is nothing in nature, particularly anything endowed with life, which is without orenda. Gods, spirits, men, animals are all endowed with orenda. Natural phenomena, such as storms, are produced by the orenda of the spirits of these phenomena. The fortunate hunter is one whose orenda has defeated the orenda of his prey. The orenda of animals hard to catch is said to be intelligent and cunning. Everywhere among the Huron there are examples of struggles between different orenda-in the same way as we found struggles between different manas in Melanesia. And the orenda, like mana, is distinct from the objects to which it is attached, to such a degree that it can be exhaled, thrown into the air-the spirit which brings thunderstorms throws up his orenda in the form of clouds. Orenda is also the sound of an object. Animals crying, birds singing, rustling trees, the blowing of the wind-all are expression of orenda. In the same way the voice of the magician is orenda. The orenda of things is like an incantation.
In fact, the name Huron, when uttered aloud, is none other than orenda. In addition, orenda means, in its original sense, prayers or chants. This meaning of the word is confirmed by the terms which correspond to it in other Iroquois dialects. But although incantations are orenda, par excellence, Hewitt expressly informs us that all ritual is orenda, and this aspect again reminds us of mana. Orenda is, above all, the power of the shaman. He is called rarediowa'ne, somebody whose orenda is great and powerful. A prophet or diviner, ratreñ'dats or hatreñdotha is someone who habitually exhales or effuses his orenda, and in this way learns the secrets of the future. It is orenda which is magic's active ingredient. Everyone who practises magic is said to be possessed by orenda, activated by it rather than by virtue of any physical properties. This is what gives power to spells, amulets and fetishes, mascots, lucky charms and, if you like, medicines. It is particularly active in black magic. All magic, therefore, derives from orenda.
We have some hints that lead us to believe that orenda works through a system of symbolic classification. The cricket is called the ripener of the corn, because it sings on hot days, that is its orenda which brings warmth to make the corn grow; 'the rabbit "sings", and by barking the underbrush at a suitable height, indicates the depth to which the snow must fall. Thus his orenda controlled the snow.' The hare is the totem animal of a clan in one of the Huron phratries, and this clan has the power to bring fog and snowfalls. It is, therefore, the orenda which unites the various classificatory terms which include the hare, the totemic clan, fog and snow on the one hand, and on the other, the cricket, heat and corn. In this classification it plays the role of middle term. These texts also give us an idea of the way the Iroquois represent causality. For them, the cause, par excellence, is the voice. To sum up, orenda is not material power, it is not the soul, nor an individual spirit, nor is it strength nor force. Hewitt establishes, in fact, that there are other terms to express these various notions and he correctly defines orenda as
a 'hypothetic potence or potentiality to do or effect results mystically.'
The famous concept of manitou found among the Algonquins and particularly among the Ojibway is basically the same as our Melanesian mana. The manitou, according to Father Thavenet-the author of an excellent French-Algonquin dictionary still in manuscript-refers, in fact, not to a spirit, but to a whole species of spirits, forces and qualities (Tesa, Studi del Thavenet, Pisa, 1881, p. 17).
'It means being, substance, the state of being animate and it is quite clear that to a certain extent all beings with souls are manitous. But it particularly refers to all beings which still have no common name, which are not familiar. A woman who came across a salamander said she was afraid, thinking it to be a manitou. The people laughed at her and told her the name of the animal. Trade beads are manitou's scales, and cloth-wonderful as it is-is said to be the skin of a manitou. A manitou is an individual who performs extraordinary feats-a shaman is a manitou. Plants have manitous. A sorcerer who uses the tooth of a rattlesnake will say that it is a manitou; when it is found to have no power to kill, he says that it no longer has manitou.'