Skip to main content

Full text of "The elementary forms of the religious life, a study in religious sociology"

See other formats








The Elementary Forms 

of the 

Religious Life 



M. A. 






This book is copyright under the Berne Convention. 
Apart from any fair dealing for the purposes of private 
itudv, research, criticism or review, as permitted under 
the Copyright Act, ig^G, no portion may be reproduced 
hy any process without written permission. Enquiry 
should be made to the publisher. 

(g) George Allen & Unwin Ltd. 1 9 1 5 






CûtO .% 



Subject of our Study : Religious Sociology and the Theory of 


I. — Principal subject of the book : analysis of the simplest religion known 
to determine the elementary forms of the religious life — Why they are 
more easily found and explained in the primitive religions 

II. — Secondary subject of research : the genesis of the fundamental 
notions of thought or the categories — Reasons for believing that 
their origin is religious and consequently social — How a way of 
restating the theory of knowledge is thus seen . . . . . j 



Definition of Religious Phenomena and of Religion 

Usefulness of a preliminary definition of religion ; method to be followed 
in seeking this definition — Why the usual definitions should be 
examined first .......... 23 

I. — Religion defined by the supernatural and mysterious — Criticism : the 

nption of mystery is not primitive ....... 24 

II. — Religion defined in connection with the idea of God or a spiritual being. 
— Religions without gods — Rites in deistic religions which imply no 
idea of divinity .......... 29 

III. — Search for a positive definition — Distinction between beliefs and 

rites — Defiinition of beliefs — First characteristic : division of things 

^.between sacred and profane — Distinctive characteristics of this 

definition — Definition of rites in relation to beliefs — Definition of 

religion . . . . . , . - . . . .36 

IV. — Necessity of another characteristic to distinguish magic from 
religion — The idea of the Church — Do individualistic religions exclude 
the idea of a Church ?......... 42 

Leading Conceptions of the Elementary Religion 
I. — Animism 
Distinction of animism and naturism ....... 48 

I. — The three theses of animism ; Genesis of the idea of the soul ; Forma- 
tion of the idea of spirits ; Transformation of the cult of spirits into 
the cult of nature .......... 49 


vi Contents 


II. — Criticism of the first thesis — Distinction of the idea of the soul from 

that of a double — Dreams do not account for the idea of the soul . 55 

III. — Criticism of the second thesis — Death does not explain the trans- 
formation of a soul into a spirit — The cult of the souls of the dead is 
not primitive .......... 60 

IV. — Criticism of the third thesis — The anthromoporphic instinct — 
Spencer's criticism of it ; reservations on this point — Examination 
of the facts by which this instinct is said to be proved — Difference 
between a soul and the spirits of nature — Religious anthropomorphism 
is not primitive . . . . . . . . . . . 65 

\ V. — Conclusion : animism reduces religion to nothing more than a system 

of hallucinations .......... 68 


Leading Conceptions of the Elementary Religion — {continued) 

II. — Naturism 

History of the theory .......... 71 

I. — Exposition of Max MUller's naturism ...... 73 

II. — If the object of religion is to express natural forces, it is bard to see 
how it has maintained itself, for it expresses them in an erroneous 
manner — Pretended distinction between religion and mythology . 78 
III. — Naturism does not explain the division of things into sacred and 

profane ........... 84 



I. — Brief history of the question of totemism ...... 88 

II. — Reasons of method for which our study will be given specially to the 
totemism of Australia — The place which will be given to facts from 
America ........... 93 



ToTEMic Beliefs 
The Totem as Name and as Emblem 
I. — Definition of the clan — The totem as name of the clan — Nature of the 
things which serve as totems — Ways in which the totem is acquired 
— The totems of phratries ; of matrimonial classes . . . .102 

II. — ^The totem as emblem — Totem ic designs engraved or carved upon 

objects; tatooings or designs upon the body . . . . .113 

III. — Sacred character of the totemic emblem — The churinga — The 

nurtunja — The waninga — Conventional character of totemic emblems 119 


Totemic Beliefs — {continued) 
The Totemic Animal and Man 
I. — Sacred character of the totemic animals — Prohibition to eat them, kill 
them or pick the totemic plants — Different moderations given these 
prohibitions — Prohibition of contact — The sacred character of the 
animal is less marked than that of the emblem . . . .128 

Contents , vii 


II. — The man — His relationship with the totemic animal or plant — 
Different myths explaining this relationship — The sacred character 
of the man is more apparent in certain parts of the organism : the 
blood, hair, etc. — How this character varies with sex and age — 
Totemism is not plant or animal worship . . . . .134 

Totemic Beliefs — {continued) 
The Costnological System of Totemism and the Idea of Class 
I. — The classification of things into clans, phratries and classes . . 141 

II. — Genesis of the notion of class : the first classifications of things take 
their forms from society — Differences between the sentiment of the 
differences of things and the idea of class — Why this is of social origin 144 
III. — Religious significance of these classifications : all of the things 
classified into a clan partake of the nature of the totem and its sacred 
character — The cosmological system of totemism — Totemism as the 
tribal religion . . . . . . . . . .148 


Totemic Beliefs — {end) 

The Individual Totem and the Sexual Totem 

I. — Individual totem as a forename ; its sacred character — Individual 
totem as personal emblem — Bonds between the man and his indi- 
vidual totem — Relations with the collective totem . . . .157 

II. — The totems of sexual groups — Resemblances and differences with the 

collective and individual totems — Their tribal nature . . . 165 


Origins of these Beliefs 

Critical Examination of Preceding Theories 

I. — Theories which derive totemism from a previous religion : from the 
ancestor cult (Wilken and Tylor) ; from the nature cult (Jevons) — 
Criticism of these theories . . . . . . . . 1 68 

II. — Theories which derive collective totemism from individual totemism — 
Origins attributed by these theories to the individual totem (Frazer, 
Boas, Hill Tout) — Improbability of these hypotheses — Reasons 
showing the priority of the collective totem . . . . .172 

III. — Recent theory of Frazer : conceptional and local totemism — The 
begging of the question upon which it rests — The religious character 
of the totem is denied — Local totemism is not primitive . . .180 

IV. — Theory of Lang : that the totem is only a name — Difficulties in 
explaining the religious character of totemic practices from this point 
of view ........... 184 

V. — All these theories explain totemism only by postulating other religious 

notions anterior to it . . . . . . . . .186 


Origins of these Beliefs — {continued) 

The Notion of the Totemic Principle, or M ana, and the Idea of Force 

L — The notion of the totemic force or principle — Its ubiquity — Its 

character at once physical and moral . . . . . .188 

viii Contents 


II. — Analogous conceptions in other inferior societies — The gods in Samoa, 
the wakan of the Sioux, the orenda of the Iroquigs, the mana of 
Melanesia — Connection of these notions with totemism — The Arun- 
kulta of the Arunta ......... 191 

III. — Logical priority of impersonal force over the difiEerent mjrthical 

personalities — Recent theories which tend to admit this priority . 198 

IV. — The notion of religious force is the prototype of that of force in 

general ........... 203 


Origins of these Beliefs — {end) 

Origin of the Idea of the Totemic Principle or Mana 

I. — The totemic principle is the clan, but thought of under a more empirical 

form ............ 205 

II. — General reasons for which society is apt to awaken the sensation of the 
sacred and the divine — Society as an imperative moral force ; the 
notion of moral authority — Society as a force which raises the indi- 
vidual outside of himself — Facts which prove that society creates the 
sacred ........... 206 

III. — Reasons peculiar to Australian societies — The two phases through 
which the life of these societies alternatively passes : dispersion, con- 
centration — Great collective effervescence during the periods of 
concentration — Examples — How the religious idea is bom out of this 
effervescence .......... 214 

Why collective force has been thought of under totemic forms : it is the 
totem that is the emblem of the clan — Explanation of the principal 
totemic beliefs . . . . . . . . . .219 

IV. — Religion is not the product of fear — It expresses something real — Its 
essential idealism — This idealism is a general characteristic of collective 
mentality — Explanation of the external character of religious forces 
in relation to their subjects — The principle that the part is equal to 
the whole ........... 223 

V. — Origin of the notion of emblem : emblems a necessary condition of 
collective representations — Why the clan has taken its emblems from 
the animal and vegetable kingdoms . . . . . .230 

VI. — The proneness of the primitive to confound the kingdoms and 
classes which we distinguish — Origins of these confusions — How 
they have blazed the way for scientific explanations — They do not 
exclude the tendency towards distinction and opposition . . . 

The Idea of the Soul 

I. — Analysis of the idea of the soul in the Australian societies . . . 240 

II. — Genesis of this idea — The doctrine of reincarnation according to 
Spencer and Gillen : it implies that the soul is a part of the totemic 
principle — Examination of the facts collected by Strehlow ; they 
confirm the totemic nature of the soul ...... 246 

III. — Generality of the doctrine of reincarnation — Diverse facts in support 

of the proposed genesis ........ 256 

IV. — Antithesis of the soul and the body : what there is objective in this — 
Relations of the individual soul with the collective soul — The idea of 
the soul is not chronologically after that of mana .... 262 

V. — Hypothesis to explain the belief in its survival .... 267 

VI. — The idea of a soul and the idea of a person ; impersonal eleipents in 

the personality .......... 269 

Contents ix 

The Idea of Spirits and Gods 


I. — Difference between a soul and a spirit — The souls of the mythical 
ancestors are spirits, having determined functions — Relations between 
the ancestral spirit, the individual soul and the individual totem — 
Explanation of this latter — Its sociological significance . . -273 

II. — Spirits and magic ......... 281 

III. — The civilizing heroes ......... 283 

IV, — The great gods — Their origin — Their relations with the totemic 

system — Their tribal and international character .... 285 

V. — Unity of the totemic system ........ 295 



The Negative Cult and its Functions 
The Ascetic Rites 

I. — The system of interdictions — Magic and religious interdictions — 
Interdictions between sacred things of different sorts — Interdictions 
between sacred and profane — These latter are the basis of the negative 
cult — Leading types of these interdictions ; their reduction to two 
essential types .......... 299 

II. — The observance of interdictions modifies the religious state of indivi- 
viduals — Cases where this efficacy is especially apparent : ascetic 
practices — The religious efficacy of sorrow — Social function of 
asceticism ........... 309 

III. — Explanation of the system of interdictions : antagonism of the 

sacred and the profane, contagiousness of the sacred . . -317 

IV. — Causes of this contagiousness — It cannot be explained by the laws 
of the association of ideas — It is because religious forces are outside 
of their subjects — Logical interest in this property of religious forces 321 


The Positive Cult 

I. — The Elements of Sacrifice 

The Intichiuma ceremony in the tribes of Central Australia — Different 

forms which it presents ........ 326 

I- — The Arunta Form — The two phases — Analysis of the first : visit to 
sacred places, scattering of sacred dust, shedding of blood, etc., to 
assure the reproduction of the totemic species .... 327 

H- — Second phase : ritual consumption of the totemic plant or animal . 333 

III. — Interpretation of the complete ceremony — The second rite consists 

in a communion meal — Reason for this communion . . . 336 

IV. — The rites of the first phase consists in oblations — Analogies with 
sacrificial oblations — The Intichiuma thus contains the two elements 
of sacrifice^ — Interest of these facts for the theory of sacrifice . . 340 

V. — On the pretended absurdity of sacrificial oblations — How they are 
explained : dependence of sacred beings upon their worshippers — 
Explanation of the circle in which sacrifice seems to move — Origin of 
the periodicity of positive rites ....... 344 

X Contents 

The Positive Cult — {continued) 
II. — Imitative Rites and the Principle of Causality p^^.^ 

T. — Nature of the imitative rites — Examples of ceremonies where they 

are employed to assure the fertility of the species .... 351 

II. — They rest upon the principle : like produces like — Examination of the 
explanation of this given by the anthropological school — Reasons why 
they imitate the animal of plant — Reasons for attributing a physical 
eâicacy to these gestures — Faith — In what sense it is founded upon 
experience — The principles of magic are bom in religion . . . 355 

III. — The preceding principle considered as one of the first statements 
of the principle of causality — Social conditions upon which this latter 
depends — The idea of impersonal force or power is of social origin — 
The necessity for the conception of causality explained by the 
authority inherent in social imperatives ...... 362 

The Positive Cult — (continued) 
III. — Representative or Commemorative Rites 
I. — Representative rites with physical efhcacy — Their relations with the 

ceremonies already described — Their action is wholly moral . . 371 

II. — Representative rites without physical efificacy — They confirm the 
preceding results — The element of recreation in religion : its impor- 
tance ; its reason for existence — The idea of a feast . . «376 
III. — Ambiguity of function in the various ceremonies studied ; they 
substitute themselves for each other — How this ambiguity confirms 
the theory proposed ......... 38^ 

PiAcuLAR Rites and the Ambiguity of the Notion of Sacredness 
Definition of the piacular rite ........ 38g 

I. — Positive rites of mourning — Description of these rites . . . 390 

II. — How they are explained — They are not a manifestation of private 
sentiments — The malice attributed to the souls of the dead cannot 
account for them either — They correspond to the state of mind in 
which the group happens to be — Analysis of this state — How it ends 
by mourning — Corresponding changes in the way in which the souls 
of the dead are conceived ........ 396 

III. — Other piacular rites ; after a public moiuming, a poor harvest, a 
drought, the southern lights — Rarity of these rites in Australia — 
How they are explained ........ 403 

IV. — The two forms of the sacred : the pure and the impure — Their 
antagonism — Their relationship — Ambiguity of the idea of the 
sacred — All rites present the same character ..... 409 


To what extent the results obtained may be generalized . . . .415 

I. — Religion rests upon an experience that is well founded but not 
privileged — Necessity of a science to reach the reality at the bottom 
of this experience — What is this reality ? — The human groups — 
Human meaning of religion — Concerning the objection which opposes 
the ideal society to the real society . . . . . .416 

How religious individualism and cosmopolitanism are explained in this 

theory ........... 424 

Contents xi 


II. — The eternal element in religion — Concerning the conflict between 
i.-'-^cience and religion ; it has to do solely with the speculative side of 

religion — What this side seems destined to become . . . .427 

III. — How has society been able to be the source of logical, that is to 
say conceptual, thought ? Definition of the concept : not to be con- 
founded with the general idea ; characterized by its impersonality and 
communicability — It has a collective origin — The analysis of its 
contents bears witness in the same sense — Collective representations 
as types of ideas which individuals accept — In regard to the objection 
that they are impersonal only on condition of being true — Con- 
ceptual thought is coeval with humanity ..... 431 

IV. — How the categories express social things — The chief category is the 
concept of totality which could be suggested only by society — Why 
the relations expressed by the categories could become conscious only 
in society — Society is not an a logical being — How the categories tend 
to detach themselves from geographically determined groups . . 439 

The unity of science on the one hand, and of morals and religion on the 
other — How the society accounts for this unity — Explanation of the 
rôle attributed to society : its creative power — Reactions of sociology 
upon the science of man ........ 445 





IN this book we propose to study the most primitive and simple 
religion which is actually known, to make an analysis of it, 
and to attempt an explanation of it. A religious system may be 
said to be the most primitive which we can observe when it fulfils 
the two following conditions : in the first place, when it is found 
in a society whose organization is surpassed by no others in 
simplicity ; ^ and secondly, when it is possible to explain it 
without making use of any element borrowed from a previous 

We shall set ourselves to describe the organization of this 
system with all the exactness and fidelity that an ethnographer 
or an historian could give it. But our task will not be limited to 
that : sociology raises other problems than history or ethno- 
graphy. It does not seek to know the passed forms of civilization 
with the sole end of knowing them and reconstructing them. But 
rather, like every pqsitive_sçience, it has as its object the ex- 
planation of some actual reality which is near to us, and which 
consequently is capable of affecting our ideas and our acts : this 
reality is man, and more precisely, the man of to-day, for there is 
nothing which we are more interested in knowing. Then we are 
not going to study a very archaic religion simply for the pleasure 
of telling its peculiarities and its singularities. If we have taken 

^ In the same way, we shall say of these societies that they are primitive, 
and we shall call the men of these societies primitives. Undoubtedly the expres- 
sion lacks precision, but that is hardly evitable, and besides, when we have 
taken pains to fix the meaning, it is not inconvenient. 

2 Elementary Forms of Religious Life 

it as the subject of our research, it is because it has seemed to us 
better adapted than any other to lead to an understanding of the 
reUgious nature of man, that is to say, to show us an essential and 
permanent aspect of humanity. 

But this proposition is not accepted before the raising of strong 
objections. It seems very strange that one must turn back, and 
be transported to the very beginnings of history, in order to arrive 
at an understanding of humanity as it is at present. This manner 
of procedure seems particularly paradoxical in the question which 
concerns us. In fact, the various religions generally pass as being 
quite unequal in value and dignity ; it is said that they do not all 
contain the same quota of truth. Then it seems as though one 
could not compare the highest forms of religious thought with the 
lowest, without reducing the first to the level of the second. If 
we admit that the crude cults of the Australian tribes can help us 
to understand Christianity, for example, is that not supposing 
that this latter religion proceeds from the same mentality as the 
former, that it is made up of the same superstitions and rests 
upon the same errors ? This is how the theoretical importance 
which has sometimes been attributed to primitive religions has 
come to pass as a sign of a systematic hostility to all religion, 
which, by prejudging the results of the study, vitiates them in 

There is no occasion for asking here whether or not there are 
scholars who have merited this reproach, and who have made 
religious history and ethnology a weapon against religion. In 
any case, a sociologist cannot hold such a point of view. In fact, 
it is an essential postulate of sociology that a human institution 
cannot rest upon an error and a lie, without which it could not 
exist. If it were not founded in the nature of things, it would 
have encountered in the facts a resistance over which it could 
never have triumphed. So when we commence the study of 
primitive religions, it is with the assurance that they hold to 
reality and express it ; this principle wiU be seen to re-enter again 
and again in the course of the analyses and discussions which 
follow, and the reproach which we make against the schools from 
which we have separated ourselves is that they have ignored it. 
When only the letter of the formulae is considered, these religious 
beliefs and practices undoubtedly seem disconcerting at times, 
and one is tempted to attribute them to some sort of a deep- 
rooted error. But one must know how to go underneath the 
symbol to the reality which it represents and which gives it its 
meaning. The most barbarous and the most fantastic rites and 
the strangest myths translate some human need, some aspect of 
hfe, either individual or social. The reasons with which the 

Subject of our Study 3 

faithful justify them may be, and generally are, erroneous ; but 
the true reasons do not cease to exist, and it is the duty of science 
to discover them. 

In reality, then, there are no rehgions which are false. All are 
true in their own fashion ; all answer, though in different ways, 
to the given conditions of human existence. It is undeniably 
possible to arrange them in a hierarchy. Some can be called 
superior to others, in the sense that they call into play higher 
mental functions, that they are richer in ideas and sentiments, 
that they contain more concepts with fewer sensations and 
images, and that their arrangement is wiser. But howsoever real 
this greater complexity and this higher ideality may be, they are 
not sufficient to place the corresponding religions in different 
classes. All are religions equally, just as all living beings are 
equally alive, from the most humble plastids up to man. So when 
we turn to primitive religions it is not with the idea of depreciating 
religion in general, for these religions are no less respectable than 
the others. They respond to the same needs, they play the same 
rôle, they depend upon the same causes ; they can also well serve 
to show the nature of the religious life, and consequently to resolve 
the problem which we wish to study. 

But why give them a sort of prerogative ? Why choose them 
in preference to all others as the subject of our study ? — It is 
merely for reasons of method. 

In the first place, we cannot arrive at an understanding of the 
most recent religions except by following the manner in which 
they have been progressively composed in history. In fact, 
h istorical analy sis is the only means of explanation which it is 
possible to apply to them. It alone enables us to resolve an 
institution into its constituent elements, for it shows them to us 
as they are born in time, one after another. On the other hand, 
by placing every one of them in the condition where it was born, 
it puts into our hands the only means we have of determining the 
causes which gave rise to it. Every time that we undertake to 
explain something human, taken at a given moment in history — 
be it a religious belief, a moral precept, a legal principle, an 
aesthetic style or an economic system — it is necessary to commence 
by going back to its most primitive and simple-form,_jtû try to 
account for the characteristics by which it was marked at that 
time, and then to show how it developed and became complicated 
httle by little, and how it became that which it is at the moment 
in question. One readily understands the importance which the 
determination of the point of departure has for this series of pro- 
gressive explanations, for all the others are attached to it. It was 

4 Elementary Forms of Religious Life 

one of Descartes's principles that thefiist ring has a predominating 
place in the chain of scientific truths But there is no question of 
placing at the foundation of the science of religions an idea 
elaborated after the cartesian manner, that is to say, a logical 
concept, a pure possibility, constructed simply by force of thought. 
What we must find is a concrete jeality, and historical and ethno- 
logical observation alone can reveal that to us. But even if this 
cardinal conception is obtained by a different process than that of 
Descartes, it remains true that it is destined to have a considerable 
influence on the whole series of propositions which the science 
establishes. Biological evolution has been conceived quite 
differently ever since it has been known that monocellular beings 
do exist. In the same way, the arrangement of religious facts is 
explained quite differently, according as we put naturism, 
animism or some other religious form at the beginning of the 
evolution. Even the most specialized scholars, if they are un- 
willing to confine themselves to a task of pure erudition, and if 
they desire to interpret the facts which they analyse, are obliged 
to choose one of these hypotheses, and make it their starting- 
point. Whether they desire it or not, the questions which they 
raise necessarily take the following form : how has naturism or 
animism been led to take this particular form, here or there, or to 
enrich itself or impoverish itself in such and such a fashion ? 
Since it is impossible to avoid taking sides on this initial problem, 
and since the solution given is destined to affect the whole science, 
it must be attacked at the outset : that is what we propose to do. 

Besides this, outside of these indirect reactions, the study of 
primitive religions has of itself an immediate interest which is of 
primary importance. 

If it is useful to know what a certain particular religion consists 
in, it is still more important to know what religion in general is. 
This is the problem which has aroused the interest of philosophers 
in all times ; and not without reason, for it is of interest to all 
humanity. Unfortunately, the method which they generally 
employ is purely dialectic : they confine themselves to analysing 
the idea which they make for themselves of religion, except as 
they illustrate the results of this mental analysis by examples 
borrowed from the religions which best realize their ideal. But 
even if this method ought to be abandoned, the problem remains 
intact, and the great service of philosophy is to have prevented 
its being suppressed by the disdain of scholars. Now it is possible 
to attack it in a different way. Since all religions can be com- 
pared to each other, and since all are specie s of the same^ ass, 
there are necessarily many elements which are common to all. 
We do not mean to speak simply of the outward and visible 

Subject of our Study 5 

characteristics which they all have equally, and which make it 
possible to give them a provisional definition from the very outset 
of our researches; the discovery of these apparent signs is relatively 
easy, for the observation which it demands does not go beneath 
the surface of things. But these external resemblances suppose' 
others which are profound. At the foundation of all systems o: 
beliefs and of all cults there ought necessarily to be a certain^ 
number of fundamental representations or conceptions and o: 
ritual attitudes which, in spite of the diversity of forms whic 
they have taken, have the same objective significance and ful: 
the same functions everywhere. These are the permanent 
elements which constitute that which is permanent and human 
in religion ; they form all the objective contents of the idea which 
is expressed when one speaks of religion in general. How is it 
possible to pick them out ? 

Surely it is not by observing the complex religions which appear 
in the course of history. Every one of these is made up of such 
a variety of elements that it is very difficult to distinguish what 
is secondary from what is principal, the essential from the 
accessory. Suppose that the religion considered is like that of 
Egypt, India or the classical antiquity. It is a confused mass of 
many cults, varying according to the locality, the temples, the 
generations, the dynasties, the invasions, etc. Popular super- 
stitions are there confused with the purest dogmas. Neither the 
thought nor the activity of the religion is evenly distributed 
among the believers ; according to the men, the environment 
and the circumstances, the beliefs as well as the rites are thought 
of in different ways. Here they are priests, there they are 
monks, elsewhere they are laymen ; there are mystics and 
rationalists, theologians and prophets, etc. In these conditions it 
is difficult to see what is common to all. In one or another of 
these systems it is quite possible to find the means of making a 
profitable study of some particular fact which is specially 
developed there, such as sacrifice or prophecy, monasticism or 
the mysteries ; but how is it possible to find the co mmon fo u nda- 
tion jDf the religious life underneath the luxuriant~vëgitâtion 
which covers it ? How is it possible to find, underneath the 
disputes of theology, the variations of ritual, the multiplicity of 
groups and the diversity of individuals, the fundamental states 
characteristic of religious mentality in general ? 

Things are quite different in the lower societies. The slighter 
development of individuality, the small extension of the group, 
the homogeneity of external circumstances, all contribute to 
reducing the differences and variations to a minimum. The 
group has an intellectual and moral conformity of which we find 

6 Elementary Forms of Religious Life 

but rare examples in the more advanced societies. Everything 
is common to all. Movements are stereotyped ; everybody 
performs the same ones in the same circumstances, and this 
conformity of conduct only translates the conformity of thought. 
Every mind being drawn into the same eddy, the individual 
type nearly confounds itself with that of the race. And while all 
is uniform, all is simple as well. Nothing is deformed like these 
myths, all composed of one and the same theme which is endlessly 
repeated, or like these rites made up of a small number of gestures 
repeated again and again. Neither the popular imagination nor 
that of the priests has had either the time or the means of refining 
and transforming the original substance of the religious ideas 
and practices ; these are shown in all their nudity, and offer them- 
selves to an examination, it requiring only the slightest effort to 
lay them open. That which is accessory or secondary, the develop- 
ment of luxury, has nof yet come to hide the principal elements. ^ 
All is reduced to that which is indispensable, to that without 
which there could be no religion. But that which is indispensable 
is also that which is essential, that is to say; that which we must 
know before all else. 

Primitive civilizations offer privileged cases, then, because they 
are simple cases. That is why, in all fields of human activity, the 
observations of ethnologists have frequently been veritable 
revelations, which have renewed the study of human institutions. 
For example, before the middle of the nineteenth century, every- 
body was convinced that the father was the essential element of 
the family ; no one had dreamed that there could be a family 
organization of which the paternal authority was not the key- 
stone. But the discovery of Bachofen came and upset this old 
conception. Up to very recent times it was regarded as evident 
that the moral and legal relations of kindred were only another 
aspect of the psychological relations which result from a common 
descent ; Bachofen and his successors, MacLennan, Morgan and 
many others still laboured under this misunderstanding. But 
since we have become acquainted with the nature of the primitive 
clan, we know that, on the contrary, relationships cannot be 
explained by consanguinity. To return to religions, the study 
of only the most familiar ones had led men to believe for a long 
time that the idea of god was characteristic of everything that is 
religious. Now the religion which we are going to study presently 

1 But that is not equivalent to saying that all luxury is lacking to the primitive 
cults. On the contrary, we shall see that in every religion there are beliefs and 
practices which do not aim at strictly utilitarian ends (Bk. Ill, ch. iv, § 2). This 
luxury is indispensable to the religious life ; it is at its very heart. But it is 
much more rudimentary in the inferior religions than in the others, so we are 
better able to determine its reason for existence here. 

Subject of our Study y 

is, in a large part, foreign to all idea of divinity ; the forces to 
which the rites are there addressed are very different from those 
which occupy the leading place in our modem religions, yet they 
aid us in understanding these latter forces. So nothing is more 
unjust than the disdain with which too many historians still 
regard the work of ethnographers. Indeed, it is certain that 
ethnology has frequently brought about the most fruitful revo- 
lutions in the different branches of sociology. It is for this same «^ 
reason that the discovery of unicellular beings, oTwhiçh we just ^.^^"^ 
spokéTTîas transformed the current idea of life. Since in these 
very simple beings, life is reduced to its essential traits, these are 
less easily misunderstopd. 

BuTprimitive religions do not merely aid us in disengaging the 
constituent elements of religion ; they also, have the great ad- 
vantage that they facilitate the explanation of it. Since the 
facts there are simpler, the relations between them are more 
apparent. The reasons with whteh then account for their acts 
have not yet been elaborated and denatured by studied reflection ; 
they are nearer and more closely related to the motives which 
have really determined, these acts. In order to understand an 
hallucination perfectly, and give it its most appropriate treat- 
ment, a physician must know its original point of departure. 
Now this event is proportionately easier to find if he can observe 
it near its beginnings. The longer the disease is allowed to de- 
velop, the more it evades observation ; that is because all sorts of 
interpretations have intervened as it advanced, which tend to 
force the original state into the background, and across which 
it is frequently difficult to find the initial one. Between a 
systematized hallucination and the first impressions which gave 
it birth, the distance is often considerable. It is the same thing 
with religious thought. In proportion as it progresses in history, 
the causes which called it into existence, though remaining active, 
are no longer perceived, except across a vast scheme of inter- '"' 

pretations which quite transform them. Popular mythologies 
and subtile theologies have done their work : they have super- 
imposed upon the primitive sentiments others which are quite 
different, and which, though holding to the first, of which they are 
an elaborated form, only allow their true nature to appear very 
imperfectly. The psychological gap between the cause and the 
effect, between the apparent cause and the effective cause, has 
become more considerable and more difficult for the mind to 
leap. The remainder of this book will be an illustration and a 
verification of this remark on method. It will be seen how, in 
the primitive religions, the religious fact still visibly carries the -^ 
mark of its origins : it would have been well-nigh impossible 

8 Elementary Forms of Religious Life 

to infer them merely from the study of the more developed 

The study which we are undertaking is therefore a way of 
taking up again, but under new conditions, the old problem of the 
origin of religion. To be sure, if by origin we are to understand 
the very first beginning, the question has nothing scientific about 
it, and should be resolutely discarded. There was no given 
moment when religion began to exist, and there is consequently 
no need pi finding a means of transporting ourselves thither in 
thought. " Like every human institution, religion did not com- 
mence anjrwhere. Therefore, all speculations of this sort are justly 
discredited ; they can only consist in subjective and arbitrary 
constnictions which are subject to no sort of control. ,-But the 
problem which we raise is quite another one. What vve want to 
do is to find a means of discerning the ever-present cause s upon 
which the most essential forms of religious thought and practice 
depend. Now for the reasons which were just set forth, these 
causes are proportionately more easily observable as the societies 
where they are observed are less complicated. That is why we 
try to get as near as possible to the origins.^ It is not that we 
ascribe particular virtues to the lower religions. On the cpntrary, 
they are rudimentary and gross ; we cannot make of them a sort 
of model which later religions only have to reproduce. But even 
their grossness makes them instructive, for they thus become 
convenient for experiments, as in them, the facts and their 
relations are easily seen. In order to discover the laws of the 
phenomena which he studies, the physicist tries to simplify these 
latter and rid them of their secondary characteristics. For that 
which concerns institutions, nature spontaneously makes the 
same sort of simplifications at the beginning of history. We 
merely wish to put these to profit. Undoubtedly we can only 
touch very elementary facts by this method. When we shall 
have accounted for them as far as possible, the novelties of every 
sort which have been produced in the course of evolution will 
not yet be explained. But while we do not dream of denying the 
importance of the problems thus raised, we think that they will 
profit by being treated in their turn, and that it is important to 
take them up only after those of which we are going to undertake 
the study at present. 

^ It is seen that we give a wholly relative sense to this word " origins," just 
as to the word ' primitive." By it we do not mean an absolute beginning, but 
the most simple social condition that is actually known or that beyond which 
we cannot go at present. When we speak of the origins or of the commencement 
of religious history or thought, it is in this sense that our statements should be 

Subject of our Study 


But our study is not of interest merely for the science of 
religion. In fact, every religion has one side by which it over- 
laps the circle of properly religious ideas, and there, the study 
of religious phenomena gives a means of renewing the problems 
which, up to the present, have only been discussed among 

For a long time it has been known that the first systems of 
representations with which men have pictured to themselves the 
world and themselves were of religious origin. There is no religion 
that is not a cosmology at the same time that it is a speculation — 
upon divine things. If philosophy and the sciences were bom 
of religion, it is because religion began by taking the place of ( 
the sciences and philosophy. But it has been less frequently- 
noticed that religion has not confined itself to enriching the 
human intellect, formed beforehand, with a certain number 
of ideas ; it has contributed to forming the intellect itself. Men 
owe to it not only a good part of the substance of their 
knowledge, but also the form in which this knowledge has been 

At the roots of all our judgments there are a certain number of 
essential ideas which dominate all our intellectual life ; they are 
what philosophers since A ristotle have called the categories of 
the understanding : ideas oftime, space, ^ class, number, cause, 
substance, personality, etc. They correspond to the most uni- 
versal properties of things. They are like the solid frame which 
encloses all thought ; this does not seem to be able to liberate 
itself from them without destroying itself, for it seems that we 
cannot think of objects that are not in time and space, which 
have no number, etc. Other ideas are contingent and unsteady ; 
we can conceive of their being unknown to a man, a society or 
an epoch ; but these others appear to be nearly inseparable from 
the normal working of the intellect. They are like the frame- 
worlj of the intelligence. Now when primitive religious belief s are 
systematically analysed, the principal categories are naturally 
found. They are born in religion and of religion ; they are 
a product of religious thought. This is a statement that we 
are going to have occasion to make many times in the course 
of this work. 

* We say that time and space are categories because there is no difference 
between the rôle played by these ideas in the intellectual life and that which 
falls to the ideas of class or cause (on this point see, Hamelin, Essai sur les- 
éléments principaux de la représentation, pp. 63, 76). 


10 Elementary Forms of Religious Life 

This remark has some interest of itself already ; but here is 
what gives it its real importance. 

The general conclusion of the book which the reader has before 
him is that religion-iS- Something eminently socia l. Religious 
representations are collective representations which express 
collective realities ; the rites are a manner of acting which take 
rise in the midst of the assembled groups and which are destined 
to excite, maintain or recreate certain mental states in these 
groups. So if the categories are of religious origin, they ought to 
participate in this nature common to all religious facts ; they too 
should be social affairs and the product of collective thought. At 
least — for in the actual condition of our knowledge of these matters, 
one should be careful to avoid all radical and exclusive statements 
— it is allowable to suppose that they are rich in social elements. 

Even at present, these can be imperfectly seen in some of them. 
For example, try to represent what the notion of tim e would be 
without the processes by which we divide it, measure it or express 
it with objective signs, a time which is not a succession of years, 
months, weeks, days and hours ! This is something nearly un- 
thinkable. We cannot conceive of time, except on condition of 
distinguishing its different moments. Now what is the origin of 
this differentiation ? Undoubtedly, the states of consciousness 
which we have already experienced can be reproduced in us in 
the same order in which they passed in the first place ; thus 
portions of our past become present again, though being clearly 
distinguished from the present. But howsoever important this 
distinction may be for our private experience, it is far from being 
enough to constitute the notion or category of time. This does 
not consist merely in a commemoration, either partial or integral, 
of our past life. It is an abstract and impersonal frame which 
surrounds, not only our individual existence, but that of all 
humanity. It is like an endless chart, where all duration is spread 
out before the mind, and upon which all possible events can be 
located in relation to fixed and determined guide lines. It is not 
my time that is thus arranged ; it is time in general, such as it is 
objectively thought of by everybody in a single civilization. That 
aJone is enough to give us a hint that such an arrangement ought 
to be collective. And in reality, observation proves that these 
indispensable guide lines, in relation to which all things are 
temporally located, are taken from social life. The divisions into 
days, weeks, months, years, etc., correspond to the periodical 
recurrence of rites, feasts, and public ceremonies.^ A calendar 

' See the support given this assertion in Hubert and Mauss, Mélanges 
d'Histoire des Religions [Travaux de l'Année Sociologique), chapter on La Repré- 
sentation du Temps dans la Religion. 

Subject of our Study ii 

expresses the rh3rthm of the collective activities, while at the same 
time its function is to assure their regularity. ^ 

It is the same thing with space. As Hamelin has shown, ^ 
space is not the vague and inaetermined medium which Kant 
imagined ; if purely and absolutely homogeneous, it would be 
of no use, and could not be grasped by the mind. Spatial 
representation consists essentially in a primary co-ordination 
of the data of sensuous experience. But this co-ordination would 
be impossible if the parts of space were qualitatively equivalent 
and if they were really interchangeable. To dispose things 
spatially there must be a possibility of placing them differently, of 
putting some at the right, others at the left, these above, those 
below, at the north of or at the south of, east or west of, etc., etc., 
just as to dispose states of consciousness temporally there must 
be a possibility of localizing them at determined dates. That 
is to say that space could not be what it is if it were not, like 
time, divided and differentiated. But whence come these 
divisions which are so essential ? By themselves, there are 
neither right nor left, up nor down, north nor south, etc. All 
these distinctions evidently come from the fact that different 
sympathetic values have been attributed to various regions. Since 
all the men of a single civilization represent space in the same way, 
it is clearly necessary that these sympathetic values, and the 
distinctions which depend upon them, should be equally universal, 
and that almost necessarily implies that they be of social origin.* 

Besides that, there are cases where this social character is 
made manifest. There are societies in Australia and North 
America where space is conceived in the form of an immense 
circle, because the camp has a circular form ;* and this spatial 
circle is divided 'up exactly like the tribal circle, and is in its 

^ Thus we see all the difierence which exists between the group of sensations 
and images which serve to locate us in time, and the category of time. The 
first are the summary of individual experiences, which are of value only for the 
person who experienced them. But what the category of time expresses is a 
time common to the group, a social time, so to speak. In itself it is a veritable 
social institution. Also, it is peculiar to man ; animals have no representations 
of this sort. 

This distinction between the category of time and the correspondmg sensa- 
tions could be made equally well in regard to space or cause. Perhaps this would 
aid in clearing up certain confusions which are maintained by the controversies 
of which these questions are the subject. We shall return to this point in the 
conclusion of the present work (§4). * Op. cit., pp. 75 ff. 

' Or else it would be necessary to admit that all individuals, in virtue of 
their organo-physical constitution, are spontaneously affected in the same 
manner by the different parts of space : which is more improbable, especially 
as in themselves the different regions are sympathetically indifferent. Also, 
the divisions of space vary with different societies, which is a proof that they are 
not founded exclusively upon the congenital nature of man. 

* See Durkheim and Mauss, De quelques formes primitives de classification, 
in Année Sociologique, VI, pp. 47 ff. 

12 Elementary Forms of Religious Life 

image. There are as many regions distinguished as there are 
clans in the tribe, and it is the place occupied by the clans inside 
the encampment which has determined the orientation of these 
regions. Each region is defined by the totem of the clan to which 
it is assigned. Among the Zuiïi, for example, the pueblo contains 
seven quarters ; each of these is a group of clans which has had 
a unity : in all probability it was originally a single clan which 
was later subdivided. Now their space also contains seven 
quarters, and each of these seven quarters of the world is in 
intimate connection with a quarter of the pueblo, that is to say 
with a group of clans. ^ " Thus," says Gushing, " one division 
is thought to be in relation with the north, another represents 
the west, another the south," etc.^ Each quarter of the pueblo 
has its characteristic colour, which symbolizes it ; each region 
has its colour, which is exactly the same as that of the corre- 
sponding quarter. In the course of history the number of 
fundamental clans has varied ; the number of the fundamental 
regions of space has varied with them. Thusj the social organi za- 
tion has j)e en the model for th e spatial organization and a ^re- 
prÔHÛHion_of_it. ^t~ls ï1ius"ëvên~up to the distraction between 
right "and left which, far from being inherent in the nature of 
man in general, is very probably the product of representations 
which are religious and therefore collective.^ 

Analogous proofs will be found presently in regard to the ideas 
of class, force, personality and efficacy. It is even possible to 
ask if the idea of contradiction does not also depend upon social 
conditions. What makes one tend to believe this is that the 
empire which the idea has exercised over human thought has 
varied with times and societies. To-day the principle of identity 
dominates scientific thought ; but there are vast systems of 
representations which have played a considerable rôle in the 
history of ideas where it has frequently been set aside : these 
are the mythologies, from the grossest up to the most reason- 
able.* There, we are continually coming upon beings which 

^ See Durkheim and Mauss, De quelques formes primitives de classification, in 
Année Sociologique, VI, p. 34. 

* Zuni Creation Myths, in 13/A Rep. of the Bureau of Amer. Ethnol., pp. 367 ff. 

* See Hertz, La prééminence de la main droite. /Aude de polarité religieuse, in 
the Revue Philosophique, Dec, 1909. On this same question of the relations between 
the representation of space and the form of the group, see the chapter in Ratzel, 
Politische Géographie, entitled Der Raum in Geist der Vôlker. 

* We do not mean to say that mythological thought ignores it, but that 
it contradicts it more frequently and openly than scientific thought does. 
Inversely, we shall show that science cannot escape violating it, though it 
holds to it far more scrupulously than religion does. On this subject, as on 
many others, there are only differences of degree between science and religion ; 
but if these differences should not be exaggerated, they must be noted, for they 
are significant. 

Subject of our Study 13 

have the most contradictory attributes simultaneously, who are 
at the same time one and many, material and spiritual, who can 
divide themselves up indefinitely without losing anything of 
their constitution ; in mythology it is an axiom that the part 
is worth the whole. These variations through which the rules 
which seem to govern our present logic have passed prove that, 
far from being engraven through all eternity upon the mental 
constitution of men, they depend, at least in part, upon factors 
that are historical and consequently social. We do not know 
exactly what they are, but we may presume that they exist. ^ 

This hypothesis once admitted, the problem of knowledge is 
posed in new terms. 

Up to the present there have been only two doctrines in thè\ 
field. For some, the catego ries cannot be derived from experience :/ 
they are logically prior to it and condition it. They are repre-j 
sented as so many simple and irreducible data, imminent in the> 
human mind by virtue of its inborn constitution. For this reason 
they are said to be a priori. Others, however, hold that they are! 
constructed and made up of pieces and bits, and that the indi-j 
vidual is the artisan of this construction. ^ 

But each solution raises grave difficulties. 

Is the empirical thesis the one adopted ? Then it is necessary 
to deprive the categories of all their characteristic properties. 
As a matter of fact they are distinguished from all other know- 
ledge by their universality and necessity. They are the most 
general concepts which exist, because they are applicable to all 
that is real, and since they are not attached to any particular 
object they are independent of every particular subject ; they 
constitute the common field where all minds meet. Further, 
they must meet there, for reason, which is nothing more than 
all the fundamental categories taken together, is invested with 
an authority which we could not set aside if we would. When 
we attempt to revolt against it, and to free ourselves from some 

* This hypothesis has already been set forth by the founders of the Vdlker- 
psychologie. It is especially remarked in a short article by Windelbrand entitled 
Die Erkenntnisslehre unter dem Volkerpsychologischen Gesichtspunke, in the 
Zeitsch. f. Viilkerpsychologie, viii, pp. i66 fE. Cf. a note of Steinthal on the same 
subject, ibid., pp. 178 ff. 

* Even in the theory of Spencer, it is by individual experience that the 
categories are made. The only difference which there is in this regard between 
ordinary empiricism and evolutionary empiricism is that according to this 
latter, the results of individual experience are accumulated by heredity. But 
this accumulation adds nothing essential to them ; no element enters into their 
composition which does not have its origin in the experience of the individual. 
According to this theory, also, the necessity with which the categories actually 
impose themselves upon us is the product of an illusion and a superstitious 
prejudice, strongly rooted in the organism, to be sure, but without foundation in 
the nature of things. 

14 Elementary Forms of Religious Life 

of these essential ideas, we meet with great resistances. They 
do not merely depend upon us, but they impose themselves 
upon us. Now empirical data present characteristics which are 
diametrically opposed to these. A sensation or an image always 
relies upon a determined object, or upon a collection of objects 
of the same sort, and expresses the momentary condition of a 
particular consciousness ; it is essentially individual and sub- 
jective. We therefore have considerable liberty in dealing with 
the representations of such an origin. It is true that when our 
sensations are actual, they impose themselves upon us in fact. 
But by right we are free to conceive them otherwise than they 
really are, or to represent them to ourselves as occurring in a 
different order from that where they are really produced. In 
regard to them nothing is forced upon us except as considerations 
of another sort intervene. Thus we find that we have here two 
sorts of knowledge, which are like the two opposite poles of the 
intelligence. Under these conditions forcing reason back upon 
experience causes it to disappear, for it is equivalent to reducing 
the universality and necessity which characterize it to pure 
appearance, to an illusion which may be useful practically, but 
which corresponds to nothing in reality ; consequently it is 
denying all objective reality to the logical life, whose regulation 
and organization is the function of the categories. Classical 
empiricism results in irrationalism ; perhaps it would even be 
fitting to designate it by this latter name. 

In spite of the sense ordinarily attached to the name, the 
apriorists have more respect for the facts. Since they do not 
'admit it as a truth established by evidence that the categories 
are made up of the same elements as our sensual representations, 
they are not obliged to impoverish them systematically, to draw 
from them all their real content, and to reduce them to nothing 
more than verbal artifices. On the contrary, they leave them aU 
their specific characteristics. The apriorists are the r ationali sts ; 
they believe that the world has a logical âspect^wHîch the reason 
expresses excellently. But for all that, it is necessary for them 
to give the mind a certain power of transcending experience and 
of adding to that which is given to it directly ; and of this sin- 
gular power they give neither explanation nor justification. 
For it is no explanation to say that it is inherent in the nature of 
the human intellect. It is necessary to show whence we hold 
this surprising prerogative and how it comes that we can see 
certain relations in things which the examination of these things 
cannot reveal to us. Saying that only on this condition is 
experience itself possible changes the problem perhaps, but does 
not answer it. For the real question is to know how it comes 

Subject of our Study 15 

that experience is not sufficient unto itself, but presupposes 
certain conditions which are exterior and prior to it, and how 
it happens that these conditions are reahzed at the moment and 
in the manner that is desirable. To answer these questions it 
has sometimes been assumed that above the reason of individuals 
there is a superior and perfect reason from which the others 
emanate and from which they get this marvellous power of theirs, 
by a sort of mystic participation : this is the divine reason. 
But this hypothesis has at least the one grave disadvantage of 
being deprived of all experimental control ; thus it does not 
satisfy the conditions demanded of a scientific hypothesis. 
More than that, the categories of human thought are never fixed 
in any one definite form ; they are made, unmade and remade 
incessantly ; they change with places and times. On the other 
hand, the divine reason is immutable. How can this immuta- 
bility give rise to this incessant variability ? 

Such are the two conceptions that have been pitted against 
each other for centuries ; and if this debate seems to be eternal, 
it is because the arguments given are really about equivalent. 
If reason is only a form of individual experience, it no longer 
exists. On the other hand, if the powers which it has are recog- 
nized but not accounted for, it seems to be set outside the con- 
fines of nature and science. In the face of these two opposed 
objections the mind remains uncertain. But if the social origin 
of the categories is admitted, a new attitude becomes possible, 
which we believe will enable us to escape both of the opposed 

The fundamental proposition of the apriorist theory is that 
knowledge is made up of two sorts of elements, which cannot 
be reduced into one another, and which are like two distinct 
layers superimposed one upon the other. ^ Our hypothesis keeps 
this principle intact. In fact, that knowledge jwhich is jcalled 
empirical, the only knowledge of which the^theorists-of empiricism 
hayejiiajdeaise4H-^nstructing the reason, is that which is brought 
into ourlninds by the direct acLion_s£_objects. It is composed 
of individual states which are completely explained ^ by the 
psychical nature of the individual. If, on the other hand, the 
categories are, as we believe they are, essentially collective 

^ Perhaps some will be surprised that we do not define the apriorist theory 
by the hypothesis of innateness. But this conception really plays a secondary 
part in the doctrine. It is a simple way of stating the impossibility of reducing 
rational knowledge to empirical data. Saying that the former is innate is only 
a positive way of saying that it is not the product of experience, such as it is 
ordinarily conceived. 

* At least, in so far as there are any representations which are individual 
and hence wholly empirical. But there are in fact probably none where the two 
elemçnts are not found closely united. 

1 6 Elementary Forms of Religious Life 

representations, before all else, they should show the mental 
states of the group ; they should depend upon the way in which 
this is founded and organized, upon its morphology, upon its 
religious, moral and economic institutions, etc. So betv/een these 
two sorts of representations there is all the difference which 
exists between the individual and the social, and one can no 
more derive the second from the first than he can deduce society 
from the individual, the whole from the part, the complex from 
the simple. 1 Society is a reality sui generis ; it has its own 
peculiar characteristics, which are not found elsewhere and 
which are not met with again in the same form in all the rest of 
the universe. The representations which express it have a wholly 
different contents from purely individual ones and we may rest 
assured in advance that the first add something to the second. 

Even the manner in which the two are formed results in dif- 
ferentiating them. Collective representations are the result of 
an immense co-operation, which stretches out not only into space 
but into time as well ; to make them, a multitude of minds have 
associated, united and combined their ideas and sentiments ; 
for them, long generations have accumulated their experience 
and their knowledge. A special intellectual activity is therefore 
concentrated in them which is infinitely richer and complexer 
than that of the individual. From that one can understand how 
the reason has been able to go beyond the limits of empirical 
knowledge. It does not owe this to any vague mysterious virtue 
but simply to the fact that according to the well-known formula, 
man is double. There are two beings in him : an individuaT^, 
being which has its foundation in the organism and the circle of 
whose activities is therefore strictly limited, and a social being 
which represents the highest reality in the intellectual and 
moral order that we can know by observation — I mean society. 
This duality of our nature has as its consequence in the practical 
order, the irreducibility of a moral ideal to a utilitarian motive, 
and in the order of thought, the irreducibility of reason to 
individual experience. In so far as he belongs to society, the 

^ This irreducibility must not be taken in any absolute sense. We do not 
wish to say that there is nothing in the empirical representations which shows 
rational ones, nor that there is nothing in the individual which could be taken 
as a sign of social life. If experience were completely separated from all that 
is rational, reason could not operate upon it ; in the same way, if the psychic 
nature of the individual were absolutely opposed to the social life, society would 
be impossible. A complete analysis of the categories should seek these germs of 
rationality even in the individual consciousness. We shall have occasion to 
come back to this point in our conclusion. All that we wish to establish here 
is that between these indistinct germs of reason and the reason properly so called, 
there is a difference comparable to that which separates the properties of the 
mineral elements out of which a living being is composed from the characteristic 
attributes of life after this has once been constituted. 

Subject of our Study 17 

individual transcends himself, both when he thinks and when 
he acts. 

This same social character leaf's to an understanding of the 
origin of the necessity of the categories. ^Itis.^(ilhat an idea 
isjiecGoo ary when it imp oses- itself- upon the mind by some sorF 
of_virtue of its o>wii, with out_ being accompanied by any proof. 
It contains within it something which constrains the intelligence"" 
and which leads to_its_acç eptance w ithout preliminary examina- 
tion. The apriorist postulates this singular quality, Hbut does not 
account for it ; for saying that the categories are necessary 
because they are indispensable to the functioning of the intellect 
is simply repeating that they are necessary. But if they really 
have the origin which we attribute to them, their ascendancy no 
longer has anything surprising in it. They represent the most 
general relations which exist between things ; surpassing all 
our other ideas in extension, they dominate all the details of 
our intellectual life. If men did not agree upon these essential 
ideas at every moment, if they did not have the same conception 
of time, space, cause, number, etc., all contact between their 
minds, _would be impossibles and with that, all "life together.' 
Thus socj_ety_c ould not a bandon the categories to the free choice 
ofthe individual without abandoning itself. If it is to live there 
is hot merelyuLeg d^of a satisfactory moral con formity, but also 
there is a minimum of logical conformity beyond which it cannot 
safely go. For this reason it uses all its authority upon its 
members to forestall such dissidences. Does a mind ostensibly 
free itself from these forms of thought ? It is no longer considered 
a human mind in the full sense of the word, and is treated accord- 
ingly. That is why we feel that we are no longer completely free 
and that something resists, both within and outside ourselves, 
when we attempt to rid ourselves of these fundamental notions, 
even in our own conscience. Outside of us there is public opinion 
which judges us ; but more than that, since society is also repre- 
sented inside of us, it sets itself against these revolutionary 
fancies, even inside of ourselves ; we have the feeling that we 
cannot abandon them if our whole thought is not to cease being 
really human. This seems to be the origin of the exceptional 
authority which Is inherent In the reason and which makes us 
accept its suggestion s with confidence. It is the very authority 
of society, 1 transferring itself to a certain manner of thought 
which is the indispensable condition of all common action. The 
necessity with which the categories are imposed upon us is not 

* It has frequently been remarked that social disturbances result in multi- 
plying mental disturbances. This is one more proof that logical discipline is 
a special aspect of social discipline. The first gives way as the second is weakened. 

1 8 Elementary Forms of Religious Life 

the effect of simple habits whose yoke we could easily throw off 
with a little effort ; nor is it a physical or metaphysical necessity, 
since the categories change in different places and times ; it is 
a special sort of moral necessity which is to the intellectual life 
what moral obligation is to the will.^ 

But if the categories originally only translate social .states., 
does it not follow that they can be applied to the rest of nature 
only as metaphors ? If they were made merely to express 
social conditions, it seems as though they could not be extended 
to other realms except in this sense. Thus in so far as they aid 
us in thinking of the physical or biological world, they have only 
the value of artificial symbols, useful practically perhaps, but 
having no connection with reality. Thus we come back, by a 
different road, to nominalism and empiricism. 

But when we interpret a sociological theory of knowledge in 
this way, we forget that even if society is a specific reality it is 
not an empire within an empire ; it is a part of nature, and in- 
deed its highest representation. The social realm is a natural 
realm which differs from the others only by a greater complexity. 
Now it is impossible that nature should differ radically from 
itself in the one case and the other in regard to that which is 
most essential. The fundamental relations that exist between 
things — just that which it is the function of the categories to 
express — cannot be essentially dissimilar in the different realms. 
If, for reasons which we shall discuss later, ^ they are more clearly 
disengaged in the social world, it is nevertheless impossible that 
they should not be found elsewhere, though in less pronounced 
forms. Society makes them more manifest but it does not have 
a monopoly upon them. That is why ideas which have been 
elaborated on the model of social things can aid us in thinking 
of another department of nature. It is at least true that if these 
ideas play the rôle of symbols when they are thus turned aside 
from their original signification, they are well-founded symbols. 
If a sort of artificiality enters into them from the mere fact that 

* There is an analogy between this logical necessity and moral obligation but 
there is not an actual identity. To-day society treats criminals in a diUcrent 
fashion than subjects whose intelligence only is abnormal ; that is a proof that 
the authority attached to logical rules and that inherent in moral rules are not 
of the same nature, in spite of certain similarities. They are two species of the 
same class. It would be interesting to make a study on the nature and origin of 
this difference, which is probably not primitive, for during a long time, the 
public conscience has poorly distinguished between the deranged and the 
delinquent. We confine ourselves to signalizing this question. By this example, 
one may see the number of problems which are raised by the analysis of these 
notions which generally pass as being elementary and simple, but which are 
really of an extreme complexity. 

• This question will be treated again in the conclusion of this work. 

Subject of our Study 19 

they are constructed concepts, it is an artificiality which follows 
nature very closely and which is constantly approaching it still 
more closely. ^ From the fact that the ideas of time, space, class, 
cause or personality are constructed out of social elements, it is 
not necessary to conclude that they are devoid of all objective 
value. On the contrary, their social origin rather leads to the 
belief that they are not without foundation in the nature of 
things. 2 

Thus renovated, the theory of knowle dge seems destined to 
unite the opposing advantages of the two rival theories, without 
incurring their inconveniences. It keeps allt he essential prin - 
ciples of the aprioristsj_but at the 3c[TTiêTîme~it is inspired by 
t hat^p ositiy£lspirit--whiçlLJhê^mprncist^ to satis fy, 

it leavesthe reason its specific power/but it accounts for it and 
does so without leaving the world of observable phenomena. 
It affirms the duality of our intellectual life, but it explains it, 
and with natural causes. The categories are no longer con- 
sidered as primary and unanalysable facts, yet they keep a 
complexity which falsifies any analysis as ready as that with 
which the empiricists content themselves. They no longer appear 
as very simple notions which the first comer can very easily 
arrange from his own personal observations and which the 
popular imagination has unluckily complicated, but rather they 
appear as priceless instruments of thought which the human 
groups have laboriously forged through the centuries and where 
they have accumulated the best of their intellectual capital.^ 
A complete section of the history of humanity is resumed therein. 
This is equivalent to saying that to succeed in understanding \^ 
them and judging them, it is necessary to resort to other means \ 

1 The rationalism which is imminent in the sociological theory of knowledge \ 

is thus midway between the classical empiricism and apriorism. For the first, I 

the categories are purely artificial constructions ; for the second, on the contrary, J 

they are given by nature ; for us, they are in a sense a work of art, but of an art^„^^ 
which imitates nature with a perfection capable of increasing unlimitedly. 

"' l^or example, that which is at the foundation of the category of time is the 

ihth of social life-; but if there is a rhythm in collective life, one may rest 
assured that there is another in the life of the individual, and more generally, 
in that of the universe. The first is merely more marked and apparent than the 
others. In the same way, we shall see that the notion of class is founded on that 
of the human group. But if men form natural groups, it can be assumed that 
among things there exists groups which are at once analogous and different. 
Classes and species are natural groups of things. 

If it seems to many minds that a social origin cannot be attributed to the 
categories without depriving them of all speculative value, it is because society 
is still too frequently regarded as something that is not natural ; hence it is 
concluded that the representations which express it express nothing in nature. 
But the conclusion is not worth more than the premise. 

* This is how it is legitimate to compare the categories to tools ; for on its 
side, a tool is material accumulated capital. There is a close relationsliip between 
the three ideas of tool, category and institution. 

20 Elementary Forms of Religious Life 

than those which have been in use up to the present. To know 
what these conceptions which we have not made ourselves are 
really made of, it does not suffice to interrogate our own con- 
sciousnesses ; we must look outside of ourselves, it is hisLory 
thaL_WÊ,-ï»ust-observe, there is a whole science wfiîcîTTnust be 
formed, a complex science which can advance but slowly and by 
collective labour, and to which the present work brings some 
fragmentary contributions in the nature of an attempt. With- 
out making these questions the direct object of our study, we shall 
profit by all the occasions which present themselves to us of 
catching at their very birth some at least of these ideas which, 
while being of religious origin, still remain at the foundation of 
the human intelligence. 




IF we are going to look for the most primitive and simple 
religion which we can observe, it is necessary to begin by 
defining what is meant by a religion ; for without this, we would 
run the risk of giving the name to a system of ideas and practices 
which has nothing at all religious about it, or else of leaving 
to one side many religious facts, without perceiving their true 
nature. That this is not an imaginary danger, and that nothing 
is thus sacrificed to a vain formalism of method, is well shown 
by the fact that owing to his not having taken this precaution, 
a certain scholar to whom the science of comparative religions 
owes a great deal, Professor Frazer, has not been able to recog- 
nize the profoundly religious character of the beliefs and rites 
which will be studied below, where, according to our view, the 
initial germ of the religious life of humanity is to be found. 
So this is a prejudicial question, which must be treated before 
all others. It is not that we dream of arriving at once at the 
profound characteristics which really explain religion : these 
can be determined only at the end of our study. But that which 
is necessary and possible, is to indicate a certain number of 
external and easily recognizable signs, which will enable us to 
recognize religious phenomena wherever they are met with, and 
which will deter us from confounding them with others. We shall 
proceed to this preliminary operation at once. 

But to attain the desired results, it is necessary to begin by 
freeing the mind of every preconceived idea. Men have been 
obliged to make for themselves a notion of what religion is, 
long before the science of religions started its methodical com- 
parisons. The necessities of existence force all of us, believers 
and non-believers, to represent in some way these things in 

1 We have already attempted to define religious phenomena in a paper 
which was published in the Année Sociologique (Vol. Il, pp. i fï.). The defini- 
tion then given differs, as will be seen, from the one we give to-day. At the end 
of this chapter (p. 47, n. i), we shall explain the reasons which have led us to 
these modifications, but which imply no essential change in the conception of 
the facts. 


24 Elementary Forms of Religious Life 

the midst of which we live, upon which we must pass judgment 
constantly, and which we must take into account in all our 
conduct. However, since these preconceived ideas are formed 
without any method, according to the circumstances and chances 
of life, they have no right to any credit whatsoever, and must 
be rigorously set aside in the examination which is to follow. 
It is not from our prejudices, passions or habits that we should 
demand the elements of the definition which we must have ; 
it is from the reality itself which we are going to define. 

Let us set ourselves before this reality. Leaving aside all con- 
ceptions of religion in general, let us consider the various re- 
ligions in their concrete reality, and attempt to disengage that 
which they have in common ; for religion cannot be defined 
except by the characteristics which are found wherever religion 
itself is found. In this comparison, then, we shall make use of 
all the religious systems which we can know, those of the present 
and those of the past, the most primitive and simple as well as 
the most recent and refined ; for we have neither the right nor 
the logical means of excluding some and retaining others. For 
those who regard re^ligion as only a natural manifestation of 
human activity, all religions, without any exception whatsoever, 
are instructive ; for all, after their manner, express man, and 
thus can aid us in better understanding this aspect of our nature. 
Also, we have seen how far it is from being the best way of study- 
ing religion to consider by preference the forms which it presents 
among the most civilized peoples.^ 

But to aid the mind in freeing itself from these usual con- 
ceptions which, owing to their prestige, might prevent it from 
seeing things as they really are, it is fitting to examine some of 
the most current of the definitions in which these prejudices 
are commonly expressed, before taking up the question on our 
own account. 


' One idea which generally passes as characteristic of all that 
is religious, is that of the supernatu ral. By this is understood 
all sorfs of things which surpass the limits of our knowledge ; 
the supernatural is the world of the mysterious, of the unknow- 
able, of the un-uuderstandable. Thus religion would be a sort 
of speculation upon all that which evades science or distinct 
thought in general, " Religions diametrically opposed in their 
overt dogmas," said Spencer, " are perfectly at one in the tacit 

1 See above, p. 3. We shall say nothing more upon the necessity of these 
preliminary definitions nor upon the method to be followed to attain them. 
That is exposed in our Règles de la Méthode sociologique, pp. 43 ff. Cf. Le Suicide, 
pp. I fi. (Paris, F. Alcan). 

Definition of Religious Phenomena and of Religion 25 

conviction that the existence of the world, with all it contains c 
and all which surrounds it, is a mystery calling for an explana- I 
tion " ; he thus makes them consist essentially in " the belief ' 
in the omnipresence of something which is inscrutable." ^ In \ 
the same manner, Max Muller sees in religion " a struggle to 
conceive the inconceivable, to utter the unutterable, a longing 
after the Infinite." 2 _j 

It is certain that the sentiment of mystery has not been without 
a considerable importance in certain religions, notably in Chris- 
tianity, It must also be said that the importance of this senti- 
ment has varied remarkably at different moments in the history 
of Christianity. There are periods when this notion passes to an 
inferior place, and is even effaced. For example, for the Christians 
of the seventeenth century, dogma had nothing disturbing for 
the reason ; faith reconciled itself easily with science and 
philosophy, and the thinkers, such as Pascal, who really felt that 
there is something profoundly obscure in things, were so little in 
harmony with their age that they remained misunderstood by 
their contemporaries.^ It would appear somewhat hasty, there- 
fore, to make an idea subject to parallel eclipses, the essential 
element of even the Christian religion. 

In^all events, it is certain that this idea does not appear until ^ 
late in the history of religions ; it is completely foreign, not only . 
to those peoples who are called primitive, but also to all others 
who have not attained a considerable degree of intellectual 
culture. When we see them attribute extraordinary virtues to 
insignificant objects, and people the universe with singular 
principles, made up of the most diverse elements and endowed 
with a sort of ubiquity which is hardly representable, we are 
undoubtedly prone to find an air of mystery in these conceptions. 
It seems to us that these men would have been willing to resign 
themselves to these ideas ,_so_ disturbing for our modern reason, 
only because of their inability to find others which were more 
rational. But, as a matter of fact, these explanations which 
surprise us so much, appear to the primitive man as the simplest -^i. 
in the world. He does not regard them as a sort of ultima ratio 
to which the intellect resigns itself only in despair of others, but 
rather as the most obvious manner of representing and under- 
standing what he sees about him. For him there is nothing strange 
in the fact that by a mere word or gesture one is able to command 

^ First Principles, p. 37. 

• Introduction to the Science of Religions, p. 18. Cf. Origin and Development 
of Religion, p. 23. 

' This same frame of mind is also found in the scholastic period, as is witnessed 
by the formula with which philosophy was defined at this time : Fides qucerens 


20 Elementary Forms of Religious Life 

the elements, retard or precipitate the motion of the stars, bring 
rain or cause it to cease, etc. The rites which he employs to assure 
the fertility of the soil or the fecundity of the animal species on 
which he is nourished do not appear more irrational to his eyes 
than the technical processes of which our agriculturists make 
use, for the same object, do to ours. The powers which he puts 
into play by these diverse means do not seem to him to have 
anything especially mysterious about them. Undoubtedly these 
forces are different from those which the modern scientist thinks 
of, and whose use he teaches us ; they have a different way of 
acting, and do not allow themselves to be directed in the same 
manner ; but for those who believe in them, they are no more 
unintelligible than are gravitation and electricity for the 

! physicist of to-day. Moreover, we shall see, in the course of this 
work, that the idea of physical forces is very probably derived 
from that of religious forces ; then there cannot exist between 
the two the abyss which separates the rational from the irrational. 
Even the fact that religious forces are frequently conceived under 
the form of spiritual beings or conscious wills, is no proof of their 
irrationality. The reason has no repugnance a priori to ad- 
mitting that the so-called inanimate bodies should be directed by 
intelligences, just as the human body is, though contemporary 
science accommodates itself with difficulty to this hypothesis. 
When Leibniz proposed to conceive the external world as an 
immense society of minds, between which there were, and could 
be, only spiritual relations, he thought he was working as a 
rationalist, and saw nothing in this universal animism which 

[_could be offensive to the intellect. 

[" Moreover, the idea of the supernatural, as we understand it, 
dates only from to-day ; in fact, it presupposes the contrary idea, 
of which it is the negation ; but this idea is not at all primitive. 
In order to say that certain things are supernatural, it is necessary 
to have the sentiment that a natural order of things exists, that is 
to say, that the phenomena of the universe are bound together 
by necessary relations, called laws. When this principle has once 
been admitted, all that is contrary to these laws must necessarily 
appear to be outside of nature, and consequently, of reason ; for 
what is natural in this sense of the word, is also rational, these 
necessary relations only expressing the manner in which things 
are logically related. But this idea of universal determinism is 
of recent origin ; even the greatest thinkers of classical antiquity 
never succeeded in becoming fully conscious of it. It is a conquest 
of the positive sciences ; it is the postulate upon which they 
repose and which they have proved by their progress. Now as 
long as this was lacking or insufficiently established, the most 

Definition of Religious Phenomena and of Religion 27 

marvellous events contained nothing which did not appear 
perfectly conceivable. So long as men did not know the im- 
mutability and the inflexibility of the order of things, and so 
long as they saw there the work of contingent wills, they found 
it natural that either these wills or others could modify them 
arbitrarily. That is why the miraculous interventions which 
the ancients attributed to their gods were not to their eyes 
miracles in the modern acceptation of the term. For them, 
they were beautiful, rare or terrible spectacles, or causes of 
surprise and marvel {davfxaTa, mirahilia, miracula) ; but they 
never saw in them glimpses of a mysterious world into which the 
reason cannot penetrate. ^ 

We can understand this mentality the better since it has not 
yet completely disappeared from our midst. If the principle of 
determinism is soHdly established to-day in the physical and 
natural sciences, it is only a century ago that it was first intro- 
duced into the social sciences, and its authority there is still 
contested. There are only a small number of minds which are 
strongly penetrated with~this-4dear^at societies are subject to 
natural laws and form a kingdom of nature. It follows that 
veritable miracles are believed to be possible there. It is ad- 
mitted, for example, that a legislator can create an institution 
out of nothing by a mere injunction of its will, or transform one 
social system into another, just as the believers in so many 
religions have held that the divine will created the world out of 
nothing, or can arbitrarily transmute one thing into another. As 
far as social facts are concerned, we still have the mentaUty of 
primitives. However, if so many of our contemporaries still 
retain this antiquated conception for sociological affairs, it is not 
because the life of societies appears obscure and mysterious to 
them ; on the contrary, if they are so easily contented with these 
explanations, and if they are so obstinate in their illusions which 
experience constantly belies, it is because social events seem to 
them the clearest thing in the world ; it is because they have not 
yet reaUzed their real obscurity ; it is because they have not yet 
recognized the necessity of resorting to the laborious methods of 
the natural sciences to gradually scatter the darkness. The same 
state of mind is found at the root of many religious beliefs which 
surprise us by their pseudo-simplicity. It is science and not~^ 
religion which has taught men that things are complex and 
difficult to understand. ^ 

But the human mind, says Jevons,^ has no need of a properly 
scientific culture to notice that determined sequences, or a constant 
order of succession, exist between facts, or to observe, on the 

* Introduction to the History of Religions, pp. 15 fi. 


28 Elementary Forms of Religious Life 

other hand, that this order is frequently upset. It sometimes 
happens that the sun is suddenly eclipsed, that rain fails at the 

^time when it is expected, that the moon is slow to reappear after 
its periodical disappearance, etc. Since these events are outside 
the ordinary course of affairs, they are attributed to extraordinary 
exceptional causes, that is to say, in fine, to extra-natural causes. 
It is under this form that the idea of the supernatural is born at 
the very outset of history, and from this moment, according to 
this author, religious thought finds itself provided with its proper 


^ But in the first place, the supernatural cannot be reduced to 
the unforeseen. The new is a part of nature just as well as its 
contrary. If we state that in general, phenomena succeed one 
another in a determined order, we observe equally well that this 
order is only approximative, that it is not always precisely the 
same, and that it has all kinds of exceptions. If we have ever 
so little experience, we are accustomed to seeing our expectations 
fail, and these deceptions return too often to appear extraordinary 
to us. A certain contingency is taught by experience just as well 
as a certain uniformity ; then we have no reason for assigning 
the one to causes and forces entirely different from those upon 
which the other depends. In order to arrive at the idea of the 
supernatural, it is not enough, therefore, to be witnesses to un- 
expected events ; it is also necessary that these be^onceived as 
impossible, that is to say, irreconcilable with an order which, 
rightly or wrongly, appears to us to be implied in the nature of 
things. Now this idea of a necessary order has been constructed 
little by little by the positive sciences, and consequently the 

^contrary notion could not have existed before them, 
r Also, in whatever manner men have represented the novelties 
and contingencies revealed by experience, there is nothing in 
these representations which could serve to characterize religion. 
For religious conceptions have as their object, before everything 
else, to express and explain, not that which is exceptional and 
abnormal in things, but, on the contrary, that which is constant 
and regular. Very frequently, the gods serve less to account for 
the monstrosities, fantasies and anomalies than for thejregular 
march of the universe, for the movement of the stars, the rhythm 
of the seasons, the annual growth of vegetation, the perpetuation 
of species, etc. It is far from being true, then, that the notion of 
the religions coincides with that of the extraordinary or the 
[^unforeseen. Jevons replies that this conception of religious 
forces is not primitive. Men commenced by imagining them to 
account for disorders and accidents, and it was only afterwards 
that they began to utilize them in explaining the uniformities of 

Definition of Religious Phenomena and of Religion 29 

nature. 1 But it is not clear what could have led men to attribute 
such manifestly contradictory functions to them. More than 
that, the hypothesis according to which sacred beings were at 
first restricted to the negative function of disturbers is quite 
arbitrary. In fact, we shall see that, even with the most simple"" 
religions we know, their essential task is t o ma mt ain, in a positive 
manner, the normal course of life.^ -^ -' 

So the idea of mystery is not of primitive origin. It was not 
given to man ; it is man who has forged it, with his own hands, 
along with the contrary idea. This is why it has a place only in 
a very small number of advanced religions. It is impossible to 
make it the characteristic mark of religious phenomena without 
excluding from the definition the majority of the facts to be 


Another idea by which the attempt to define religion is often 
made, is that of divinity. " Religion," says M. Réville, ^ " is the 
-determination of human life by the sentiment of a bond uniting 
the human mind to that mysterious mind whose domination of 
the world and itself it recognizes, and to whom it delights in 
feeling itself united." It is certain that if the word divinity is 
taken in a precise and narrow sense, this definition leaves aside 
a multitude of obviously religious facts. The souls of the dead 
and the spirits of all ranks and classes with which the religious 
imagination of so many different peoples has populated nature, 
are always the object of rites and sometimes even of a regular 
cult ; yet they are not gods in the proper sense of the term. But 
in order that the definition may embrace them, it is enough to 
substitute for the term " gods " the more comprehensive one of 
" spiritual beings." This is what Tylor does. " The first 
requisite in a systematic study of the religions of the lower 
races," he says, '* is to lay down a rudimentary definition of 
religion. By requiring in this definition the belief in a supreme 
deity . . ., no doubt many tribes may be excluded from the category 
of religious. But such narrow definition has the fault of identify- 
ing religion rather with particular developments. ... It seems 
best . . . simply to claim as a minimum definition of Religion, 
the beljeljn Spiritua j_Beings."^ By spiritual beings must be 
understood conscious~su5Jects gifted with powers superior to 
those possessed by common men ; this qualification is found 

* Introduction to the History of Religions, p. 23. 
" See below, Bk. Ill, ch. ii. 

^ Prolegomena to the History of Religions, p. 25 (tr. by Squire). 

* Primitive Culture, I, p. 424. (Fourth edition, 1903.) 

30 Elementary Forms of Religious Life 

in the souls of the dead, geniuses or demons as well as in divinities 
properly so-called. It is important, therefore, to give our 
attention at once to the particular conception of religion which 
is implied in this definition. The relations which we can have 
with beings of this sort are determined by the nature attributed 
to them. They are conscious beings ; then^we can act upon 
them only in the same way that we act upon consciousnesses in 
general, that is to say, by psychological processes, attempting 
to convince them or move them, either with thé aid of words 
(invocations, prayers), or by offerings and, sacrifices. And 
since the object of religion is to regulate our relations with these 
special beings, there can be no religion except where there are 
prayers, sacrifices, propitiatory rites, etc. Thus we have a very 
simple critérium which permits us to distinguish that which is 
religious from that which is not. It is to this critérium that 
Frazer,^ and with him numerous ethnographers,^ systematically 
makes reference. 

But howsoever evident this definition may appear, thanks to 
the mental habits which we owe to our religious education, 
there are many facts to which it is not applicable, but which 
appertain to the field of religion nevertheless. 

In the first place, there are great religions from which the idea 
of^ds and spirits is absent, or at least, where it plays only a 
secondary and minor rôle. This is the case with Buddhism. 
Buddhism, says Burnouf, " sets itself in opposition to Brah- 
manism as a moral system without god and an atheism without 
Nature . " ^ " As it recognizes not a god upon whom man depends , ' ' 
says Barth, " its doctrine is absolutely atheistic,"^ while Oldfen- 
berg, in his turn, calls it " a faith without a gor* "^ In fact, 
all that is essential to Buddhism is found in the four propositions 
which the faithful call the four noble truths.* The first states 
the existence of suffering as the accompaniment to the perpetual 
change of things ; the second shows desire to be the cause of 
suffering ; the third makes the suppression of desire the only 
means of suppressing sorrow ; the fourth enumerates the three 
stages through which one must pass to attain this suppression : 
they are uprightness, meditation, and finally wisdom, the full 

^ Beginning with the first edition of the Golden Bough, I, pp. 30-32. 

" Notably Spencer and Gillen and even Preuss, who gives the name magic 
to all non-individualized religious forces. 

^ Burnouf, Introduction à l'histoire du bouddhisme indien, sec. edit., p. 464. 
The last word of the text shows that Buddhism does not even admit the existence 
of an eternal Nature. 

* Barth, The Religions of India, p. no (tr. by Wood). 

' Oldenberg, Buddha, p. 53 (tr. by Hoey). 

" Oldenberg, ibid., pp. 313 ff. Cf. Kern, Histoire du bouddhisme dans l'Inde, 
I. pp. 389 ff. 

Definition of Religious Phenomena and of Religion 31 

possession of the doctrine. These three stages once traversed, 
one arrives at the end of the road, at the deUverance, at salvation 
by the Nirvana. 

Now in none of these principles is there question of a divinity. 
The Buddhist is not interested in knowing whence came the 
world in which he lives and suffers ; he takes it as a given fact,^ 
and his whole concern is to escape it. On the other hand, in this 
work of salvation, he can count only upon himself ; " he has 
no god to thank, as he had previously no god to invoke during 
his struggle. "2 Instead of praying, in the ordinary sense of the 
term, instead of turning towards a superior being and imploring 
his assistance, he relies upon himself and meditates. This is not 
sajnng " that he absolutely denies the existence of the beings 
called Indra, Agni and Varuna ;^ but he believes that he owes 
them nothing and that he has nothing to do with them," for 
their power can only extend over the goods of this world, which 
are without value for him. Then he is an atheist, in the sense 
that he does not concern himself with the question whether gods 
exist or not. Besides, even if they should exist, and with what- 
ever powers they might be armed, the saint or the emancipated 
man regards himself superior to them ; for that which causes 
the dignity of beings is not the extent of the action they exercise 
over things, but merely the degree of their advancement upon 
the road of salvation.'' 

It is true that Buddha, at least in some divisions of the Budd- 
hist Church, has sometimes been considered as a sort of god. 
He has his temples ; he is the object of a cult, which, by the 
way, is a very simple one, for it is reduced essentially to the 
offering of flowers and the adoration of consecrated relics or 
images. It is scarcely more than a comemorative cult. But 
more than that, this divinization of Buddha, granting that the 
term is exact, is peculiar to the form known as Northern 
Buddhism. " The Buddhist of the South," says Kern, " and 
the less advanced of the Northern Buddhists can be said, accord- 
ing to data known to-day, to speak of their founder as if he 
were a man."^ Of course, they attribute extraordinary powers 
to Buddha, which are superior to those possessed by ordinary 
mortals ; but it was a very ancient belief in India, and one that 

* Oldenberg, p. 250; Barth, p. no. 

* Oldenberg, p. 314. 

' Barth, p. 109. In the same way, Bumouf says, " I have the profound 
conviction that if Çâkya had not found about him a Pantheon already peopled 
with the gods just named, he would have felt no need of inventing them " 
(Introd. à I' hist, du bouddhisme indien, p. 119). 

* Bumouf, op. cit., p. 117. 
' Kern, op. cit., I, p. 289. 

32 Elementary Forms of Religious Life 

is also very general in a host of different religions, that a great 
saint is endowed with exceptional virtues ;i yet a. saint is not 
a god, any more than a priest or magician is, in spite of the 
superhuman faculties frequently attributed to them. On the 
other hand, according to the most authorized scholars, all this 
theism and the complicated mythology which generally accom- 
panies it, are only derived and deviated forms of Buddhism^ 
At first, Buddha was only regarded as " the wisest of men."^ 
Burnouf says " the conception of a Buddha who is something 
more than a man arrived at the highest stage of holiness, is out- 
side the circle of ideas which form the foundation of the simple 
Sutras " ;3 and the same author adds elsewhere that " his 
humanity is a fact so incontestably recognized by all that the 
myth-makers, to whom miracles cost so little, have never even 
had the idea of making a god out of him since his death."'* 
So we may well ask if he has ever really divested himself com- 
pletely of all human character, and if we have a right to make 
him into a god completely ;^ in any case, it would have to be 
a god of a very particular character and one whose rôle in no 
way resembles that of other divine personalities. For a god is 
before all else a living being, with whom man should reckon, 
and upon whom he may count ; but Buddha is dead, he has 
entered into the Nirvana, and he can no longer influence the 
march of human events.* 

Finally, whatever one may think of the divinity of Buddha, 
it remains a fact that this is a conception wholly outside the 
essential part of Buddhism. Buddhism consists primarily in 
the idea of salvation, and salvation supposes only that one know 
the good doctrine and practise it. To be sure, this could never 
have been known if Buddha had not come to reveal it ; but 
when this revelation had once been made, the work of Buddha 
was accomplished. From that moment he ceased to be a factor 
necessary to the religious life. The practice of the four holy 
truths would be possible, even if the memory of him who revealed 

^ " The belief, universally admitted in India, that great holiness is necessarily 
accompanied by supernatural faculties, is the only support which he (Çâkya) 
should find in spirits " (Burnouf, p. 119). 

* Burnouf, p. 120. 

' Ibid., p. 107. * Ibid., p. 302. 

^ This is what Kern expresses in the following terms : "In certain regards, 
he is a man ; in certain others, he is not a man ; in others, he is neither the one 
nor the other " [op. cit., I, p. 290). 

* " The conception " " was foreign to Buddhism " " that the divine Head of 
the Community is not absent from his people, but that he dwells powerfully in 
their midst as their lord and king, so that all cultus is nothing else but the 
expression of this continuing living fellowship. Buddha has entered into 
Nirvana ; if his believers desired to invoke him, he could not hear them " 
(Oldenberg, p. 309). 

Définition of Religious Phenomena and of Religion 33 

them were completely obliterated.^ It is quite another matter 
with Christianity, which is inconceivable without the ever- 
present idea of Christ and his ever-practised cult ; for it is by 
the ever-living Christ, sacrificed each day, that the community 
of believers continues to communicate with the supreme source 
of the spiritual life.^ 

All that precedes can be applied equally well to another great 
religion of India, Jainism. The two doctrines have nearly the 
same conception of the world and of life. " Like the Buddhists," 
says Barth, " the Jainas are atheists. They admit of no creator ; 
the world is eternal ; they explicitly deny the possibility of 
a perfect being from the beginning. The Jina became perfect ; 
he was not always so." 

Just as the Buddhists in the north, the Jainists, or at least 
certain of them, have come back to a sort of deism ; in the 
inscriptions of Dekhan there is mention of a Jinapati, a sort of 
supreme Jina, who is called the primary creator ; but such 
language, says the same author, is "in contradiction to the 
most explicit declarations extracted from their most authorized 

•• Moreover, if this indifference for the divine is developed to 
such a point in Buddhism and Jainism, it is because its germ 
existed already in the Brahmanism from which the two were 
derived. In certain of its forms at least, Brahmic speculation 
ended in " a frankly materialistic and atheistic interpretation 
of the universe."^ In time, the numerous divinities which the 
people of India had originally learned to adore, came to merge 
themselves into a sort of principal deity, impersonal and abstract, 
the essence of all that exists. This supreme reality, which no 
longer has anything of a divine personality about it, is contained 
within man himself, or rather, man is but one with it, for nothing 
exists apart from it. To find it, and unite himself to it, one does 
not have to search some external support outside himself ; it is 
enough to concentrate upon himself and meditate. "If in 
Buddhism," says Oldenberg, " the proud attempt be made to 
conceive a deliverance in which man himself delivers himself, 
to create a faith without a god, it is Brahmanical speculation 
which has prepared the way for this thought. It thrusts back 
the idea of a god step by step ; the forms of the old gods have 

^ " Buddhist doctrine might be in all its essentials what it actually is, even 
if the idea of Buddha remained completely foreign to it " (Oldenberg, p. 322). — 
And whatever is said of the historic Buddha can be applied equally well to the 
mythological Buddhas. 

* For the same idea, see Max Miiller, Natural Religion, pp. 103 fî. and 190. 
' Op. cit., p. 146. 

* Barth, in Encyclopédie des sciences religieuses, VI, p. 548. 

34 Elementary Forms of Religious Life 

faded away, and besides the Brahma, which is enthroned in its 
everlasting quietude, highly exalted above the destinies of the 
human world, there is left remaining, as the sole really active 
person in the great work of deliverance, man himself. "^ Here, 
then, we find a considerable portion of religious evolution which 
has consisted in the progressive recoil of the idea of a spiritual 
being from that of a deity. Here are great religions where 
invocations, propitiations, sacrifices and prayers properly so- 
called are far, from holding a preponderating place, and which 
consequently do not present that distinctive sign by which some 
claim to recognize those manifestations which are properly 
called religious. 

But even within deistic religions there are many rites which 
are completely independent of all idea of gods or spiritual beings. 
In the first place, there are a multitude of interdictions. For 
example, the Bible orders that a woman live isolated during a 
determined period each month ;^ a similar isolation is obligatory 
during the lying-in at child-birth ;^ it is forbidden to hitch an 
ass and a horse together, or to wear a garment in which the 
hemp is mixed with flax ; * but it is impossible to see the part 
which belief in Jahveh can have played in these interdictions, 
for he is wholly absent from all the relations thus forbidden, 
and could not be interested in them. As much can be said for 
the majority of the dietetic regulations. These prohibitions are 
not peculiar to the Hebrews, but they are found under diverse 
forms, but with substantially the same character, in innumerable 

It is true that these rites are purely negative, but they do not 
cease being religious for that. Also there are others which 
demand active and positive services of the faithful, but which 
are nevertheless of the same nature. They work by themselves, 
and their efficacy depends upon no divine power ; they mechani- 
cally produce the effects which are the reason for their existence. 
They do not consist either in prayers or offerings addressed to 
a being upon whose goodwill the expected result depends ; 
this result is obtained by the automatic operation of the ritual. 
Such is notably the case with the sacrifice of the Vedic religion. 
" The sacrifice exercises a direct influence upon the celestial 
phenomena," says Bergaigne ;* it is all-powerful of itself, and 
without any divine influence. It is this, for example, which 
broke open the doors of the cavern where the dawn was im- 
prisoned and which made the light of day burst forth. ^ In the 

^ Oldenberg, op. cit., p. 53. * i Sam. xxi., 6. 

' Levit. xii. * Deut. xxii., 10 and 11. 

* La religion védique, I, p. 12a. ' Ibid., p. 133. 

Definition of Religious Phenomena and of Religion 35 

same way there are special hymns which, by their direct action, 
made the waters of heaven fall upon the earth, and even in 
spite of the gods^ The practice of certain austerities has the 
same power. More than that, " the sacrifice is so fully the origin 
of things par excellence, that they have attributed to it not only 
the origin of man, but even that of the gods. . . . Such a con- 
ception may well appear strange. It is explained, however, as 
being one of the ultimate consequences of the idea of the omni- 
potence of sacrifice. "2 Thus, in the entire first part of his work, 
M. Bergaigne speaks only of sacrifices, where divinities play no 
rôle whatsoever. 

Nor is this fact peculiar to the Vedic religion, but is, on the 
contrary, quite general. In every cult there are practices which 
act by themselves, by a virtue which is their own, without the 
intervention of any god between the individual who practises 
the rite and the end sought after. When, in the so-called Feast 
of the Tabernacles, the Jew set the air in motion by shaking 
willow branches in a certain rhythm, it was to cause the wind 
to rise and the rain to fall ; and it was believed that the desired 
phenomenon would result automatically from the rite, provided 
it were coiTectly performed. ^ This is the explanation of the 
fundamental importance laid by nearly all cults upon the material 
portion of the ceremonies. This religious formalism — very 
probably the first form of legal formalism — comes from the fact 
that since the formula to be pronounced and the movements 
to be made contain within themselves the source of their efficacy, 
they would lose it if they did not conform absolutely to the 
type consecrated by success. 

Thus there are rites without gods, and even rites from which 
gods are derived. All religious powers do not emanate from 
divine personalities, and there are relations of cult which have 
other objects than uniting man to a deity. Religion is more 
than the idea of gods or spirits, and consequently cannot be 
defined exclusively in relation to these latter. 

^ " No text," says Bergaigne, " bears better witness to the consciousness of 
a magic action by man upon the waters of heaven than verse x, 32, 7, where 
this belief is expressed in general terms, applicable to an actual man, as well as 
to his real or mythological ancestors : ' The ignorant man has questioned the 
wise ; instructed by the wise, he acts, and here is the profit of his instruction : 
he obtains the flowing of streams ' " (p. 137)- 

* Ibid., p. 139. 

* Examples will also be found in Hubert, art. Magia in the Dictionnaire des 
Antiquités, VI, p. 1509. 

36 Elementary Forms of Religious Life 


These definitions set aside, let us set ourselves before the 

First of all, let us remark that in all these formulae it is the 
nature of religion as a whole that they seek to express.^ They 
proceed as if it were a sort of indivisible entity, while, as a matter 
of fact, it is made up of parts ; it is a more or less complex 
system of myths, dogmas, rites and ceremonies. Now a whole 
cannot be defined except in relation to its parts. It will be 
more methodical, then, to try to characterize the various 
elementary phenomena of which all religions are made up, 
before we attack the system produced by their union. This 
method is imposed still more forcibly by the fact that there are 
religious phenomena which belong to no determined religion. 
Such are those phenomena which constitute the matter of folk- 
lore. In general, they are the debris of passed religions, in- 
organized survivals ; but there are some which have been 
formed spontaneously under the influence of local causes. In 
our European countries Christianity has forced itself to absorb 
and assimilate them ; it has given them a Christian colouring. 
Nevertheless, there are many which have persisted up until a 
recent date, or which still exist with a relative autonomy : 
celebrations of May Day, the summer solstice or the carnival, 
beliefs relative to genii, local demons, etc., are cases in point. 
If the religious character of these facts is now diminishing, 
their religious importance is nevertheless so great that they have 
enabled Mannhardt and his school to revive the science of 
religions. A definition which did not take account of them 
would not cover all that is religious. 

Religious phenomena are naturally arranged in two funda- 
mental categories : beliefs and rites. The first are states of 
opinion, and consist in represeiitations ; the second are deter- 
mined modes of action. Between these two classes of facts there 
is all the difference which separates thought from action. 

The rites can be defined and distinguished from other human 
practices, moral practices, for example, only by the special 
nature of their object. A moral rule prescribes certain manners 
of acting to us, just as a rite does, but which are addressed to 
a different class of objects. So it is the object of the rite which 
must be characterized, if we are to characterize the rite itself. 
Now it is in the beliefs that the special nature of this object is 
expressed. It is possible to define the rite only after we have 
defined the belief. 

Definition of Religious Phenomena and of Religion 37 

All known religious beliefs, whether simple or complex, present ' 
one common characteristic : they presuppose a classification of 
all the things, real and ideal, of which men think, into two classes 
or opposed groups, generally designated by two distinct terms 
which are translated well enough by the words profane and sacred 
{profane, sacré). This division of the world into two domains, 
the one containing all that is sacred, the other all that is profane, 
is the distinctive trait of religious thought ; the beliefs, myths, 
dogmas and legends are either representations or systems of 
representations which express the nature of sacred things, the 
virtues and powers which are attributed to them, or their 
relations with each other and with profane things. But by 
sacred things one must not understand simply those personal 
beings which are called gods or spirits ; a rock, a tree, a spring, 
a pebble, a piece of wood, a house, in a word, anything can be 
sacred. A rite can have this character ; in fact, the rite does 
not exist which does not have it to a certain degree. There are 
words, expressions and formulae which can be pronounced only 
by the mouths of consecrated persons ; there are gestures and 
movements which everybody cannot perform. If the Vedic 
sacrifice has had such an efficacy that, according to mythology, 
it was the creator of the gods, and not merely a means of winning 
their favour, it is because it possessed a virtue comparable to 
that of the most sacred beings. The circle of sacred objects 
cannot be determined, then, once for all. Its extent varies in- 
finitely, according to the different religions. That is how Budd- 
hism is a religion : in default of gods, it admits the existence of 
sacred things, namely, the four noble truths and the practices 
derived from them.^ 

Up to the present we have confined ourselves to enumerating 
a certain number of sacred things as examples : we must now 
show by what general characteristics they are to be distinguished 
from profane things. 

One might be tempted, first of all, to define them by the place 
they are generally assigned in the hierarchy of things. They 
are naturally considered superior in dignity and power to profane 
things, and particularly to man, when he is only a man and has 
nothing sacred about him. One thinks of himself as occupying 
an inferior and dependent position in relation to them ; and surely 
this conception is not \vithout some truth. Only there is nothing 
in it which is really characteristic of the sacred. It is not enough 
that one thing be subordinated to another for the second to be 
sacred in regard to the first. Slaves are inferior to their masters, 

^ Not to mention the sage and the saint who practise these truths and who 
for that reason are sacred. 

38 Elementary Forms of Religious Life 

subjects to their king, soldiers to their leaders, the miser to his 
gold, the man ambitious for power to the hands which keep it 
from him ; but if it is sometimes said of a man that he makes a 
religion of those beings or things whose eminent value and 
superiority to himself he thus recognizes, it is clear that in any 
case the word is taken in a metaphorical sense, and that there 
is nothing in these relations which is really religious.^ 

On the other hand, it must not be lost to view that there are 
sacred things of every degree, and that there are some in relation 
to which a man feels himself relatively at his ease. An amulet 
has a sacred character, yet the respect which it inspires is nothing 
exceptional. Even before his gods, a man is not always in such 
a marked state of inferiority ; for it very frequently happens 
that he exercises a veritable physical constraint upon them to 
obtain what he desires. He beats the fetich with which he is 
not contented, but only to reconcile himself with it again, if in 
the end it shows itself more docile to the wishes of its adorer. ^ 
To have rain, he throws stones into the spring or sacred lake 
where the god of rain is thought to reside ; he beheves that by 
this means he forces him to come out and show himself.^ More- 
over, if it is true that man depends upon his gods, this dependence 
is reciprocal. The gods also have need of man ; without offerings 
and sacrifices they would die. We shall even have occasion to 
show that this dependence of the gods upon their worshippers is 
maintained even in the most idealistic religions. 

But if a purely hierarchic distinction is a critérium at once too 
general and too imprecise, there is nothing left with which to 
characterize the sacred in its relation to the profane except 
their heterogeneity. However, this heterogeneity is sufficient to 
characterize this classification of things and to distinguish it 
from all others, because it is very particular : it is absolute. 
In all the history of human thought there exists no other example 
of two categories of things so profoundly differentiated or so 
radically opposed to one another. The traditional opposition 
of good and bad is nothing beside this ; for the good and the bad 
are only two opposed species of the same class, namely morals, 
just as sickness and health are two different aspects of the same 
order of facts, life, while the sacred and the profane have always 
and everywhere been conceived by the human mind as two 
distinct classes, as two worlds between which there is nothing in 

* This is not saying that these relations cannot take a religious character. 
But they do not do so necessarily. 

' Schultze, Fetichisntus , p. 129. 

* Examples of these usages will be found in Frazer, Golden Bough, 2 edit., 
I, pp. 81 fi. 

Définition of Religious Phenomena and of Religion 39 

common. The forces which play in one are not simply those 
which are met with in the other, but a little stronger ; they are 
of a different sort. In different religions, this opposition has 
been conceived in different ways. Here, to separate these two 
sorts of things, it has seemed sufficient to localize them in different 
parts of the physical universe ; there, the first have been put into 
an ideal and transcendental world, while the material world is 
left in full possession of the others. But howsoever much the 
forms of the contrast may vary,^ the fact of the contrast is uni- 

This is not equivalent to saying that a being can never pass from 
one of these worlds into the other : but the manner in which this 
passage is effected, when it does take place, puts into relief the 
essential duality of the two kingdoms. In fact, it implies a 
veritable metamorphosis. This is notably demonstrated by the 
initiation rites, such as they are practised by a multitude of 
peoples. This initiation is a long series of ceremonies with the 
object of introducing the young man into the religious life : for 
the first time, he leaves the purely profane world where he 
passed his first infancy, and enters into the world of sacred 
things. Now this change of state is thought of, not as a simple 
and regular development of pre-existent germs, but as a trans- 
formation iotius suhstantiae — of the whole being. It is said that 
at this moment the young man dies, that the person that he was 
ceases to exist, and that another is instantly substituted for it. 
He is re-bom under a new form. Appropriate ceremonies are 
felt to bring about this death and re-birth, which are not under- 
stood in a merely symbolic sense, but are taken literally. ^ Does 
this not prove that between the profane being which he was and 
the religious being which he becomes, there is a break of con- 
tinuity ? 

This heterogeneity is even so complete that it frequently 
degenerates into a veritable antagonism. The two worlds are 
not only conceived of as separate, but as even hostile and jealous 
rivals of each other. Since men cannot fully belong to one except 

* The conception according to which the profajie is opposed to the sacred, 
just as the irrational is to the rational, or the intelligible is to the mysterious, 
is only one of the forms under which this opposition is expressed. Science being 
once constituted, it has taken a profane character, especially in the eyes of the 
Christian religions ; from that it appears as though it could not be applied to 
sacred things. 

* See Frazer, On Some Ceremonies of the Central Australian Tribes m Australian 
Association for the Advancement of Science, igoi, pp. 313 it. This conception is 
also of an extreme generality. In India, the simple participation in the sacrificial 
act has the same effects ; the sacrificer, by the mere act of entering within the 
circle of sacred things, changes his personality. (See, Hubert and Mauss, Essai 
sur le Sacrifice in the Année Sociologique, II, p. loi.) 

40 Elementary Forms of Religious Life 

on condition of leaving the other completely, they are exhorted 
to withdraw themselves completely from the profane world, in 
order to lead an exclusively religious life. Hence comes the 
monasticism which is artificially organized outside of and apart 
from the natural environment in which the ordinary man leads 
the life of this world, in a different one, closed to the first, and 
nearly its contrary. Hence comes the mystic asceticism whose 
object is to root out from man all the attachment for the profane 
world that remains in him. From that come all the forms of 
religious suicide, the logical working-out of this asceticism ; for 
the only manner of fully escaping the profane life is, after all, 
to forsake all life. 

The opposition of these two classes manifests itself outwardly 
with a visible sign by which we can easily recognize this very 
special classification, wherever it exists. Since the idea of the 
sacred is always and everywhere separated from the idea of the 
profane in the thought of men, and since we picture a sort of 
logical chasm between the two, the mind irresistibly refuses to 
allow the two corresponding things to be confounded, or even to 
be merely put in contact with each other ; for such a promiscuity, 
or even too direct a contiguity, would contradict too violently 
the dissociation of these ideas in the mind. The sacred thing is 
par excellence that which the profane should not touch, and cannot 
touch with impunity. To be sure, this interdiction cannot go so 
far as to make all communication between the two worlds im- 
possible ; for if the profane could in no way enter into relations 
with the sacred, this latter could be good for nothing. But, in 
addition to the fact that this establishment of relations is always 
a delicate operation in itself, demanding great precautions and 
a more or less complicated initiation, ^ it is quite impossible, unless 
the profane is to lose its specific characteristics and become sacred 
after a fashion and to a certain degree itself. The^two classes 
cannot even approachjeacjh^other an d keep their own nature at 
the same time. 

Thus we arrive at the first critérium of religious beliefs. Un- 
doubtedly there are secondary species within these two funda- 
mental classes which, in their turn, are more or less incompatible 
with each other. ^ But the real characteristic of religious pheno- 
mena is that they always suppose a bipartite division of the whole 
universe, known and knowable, into two classes which embrace 
all that exists, but which radically exclude each other. Sacred 

* See what was said of the initiation above, p. 39. 

* We shall point out below how, for example, certain species of sacred things 
exist, between which there is an incompatibility as all-exclusive as that between 
the sacred and the profane (Bk. Ill, ch. v, § 4). 

Definition of Religious Phenomena and of Religion 41 

things are those which the interdictions protect and isolate ; 
profane things, those to which these interdictions are applied 
and which must remain at a distance from the first. Religious 
beliefs are the representations which express the nature of sacred 
things and the relations which they sustain, either with each 
other or with profane things. Finally, rites are the rules of 
conduct which prescribe how a man should comport himself in 
the presence of these sacred objects. 

When a certain number of sacred things sustain relations of 
co-ordination or subordination with each other in such a way as 
to form a system having a certain unity, but which is not com- 
prised within any other system of the same sort, the totality of 
these beliefs and their corresponding^ rites-eonstitutes a religion. 
FronPthis définition it is seen that a religion is not necessarily 
contained within one sole and single idea, and does not proceed 
from one unique principle which, though varying according to 
the circumstances under which it is applied, is nevertheless at 
bottom always the same : it is rather a whole made up of distinct 
and relatively individualized parts. Each homogeneous group 
of sacred things, or even each sacred thing of some importance, 
constitutes a centre of organization about which gravitate a 
group of beliefs and rites, or a particular cult ; there is no religion, 
howsoever unified it may be, which does not recognize a plurality 
of sacred things. Even Christianity, at least in its Catholic form, 
admits, in addition to the divine personality which, incidentally, 
is triple as well as one, the Virgin, angels, saints, souls of the dead, 
etc. Thus ji^religion^annotjbe reduced to one single cult generally, 
but rather consists in a system of cults, each endowed with a 
certain autonomy. Also, this autonomy is variable. Sometimes 
they are arranged in a hierarchy, and subordinated to some pre- 
dominating cult, into which they are finally absorbed ; but some- 
times, also, they are merely rearranged and united. The religion 
which we are going to study will furnish us with an example of 
just this latter sort of organization. 

At the same time we find the explanation of how there can be 
groups of religious phenomena which do not belong to any special 
religion ; it is because they have not been, or are no longer, a 
part of any religious system. If, for some special reason, one of 
the cults of which we just spoke happens to be maintained while 
the group of which it was a part disappears, it survives only in a 
disintegrated condition. That is what has happened to many 
agrarian cults which have survived themselves as folk-lore. In 
certain cases, it is not even a cult, but a simple ceremony or 
particular rite which persists in this way.^ 

^ This is the case with certain marriage and funeral rites, for example. 

42 Elementary Forms of Religious Life 

Although this definition is only preliminary, it permits us to 
see in what terms the problem which necessarily dominates the 
science of religions should be stated. When we believed that 
sacred beings could be distinguished from others merely by the 
greater intensity of the powers attributed to them, the question 
of how men came to imagine them was sufficiently simple : it was 
enough to demand which forces had, because of their exceptional 
_energy, been able to strike the human imagination forcefully 
• enough to inspire religious sentiments. But if, as we have sought 
to establish, sacred things differ in nature from profane things, 
if they have a wholly different essence, then the problem is 
more complex. For we mu st first of all ask what has been able 
to lead men to see Jn ihe^-W.arld_two -heterogeneous and incom- 
patible worlds, though nothing in sensible experience seems able 
-to suggest the idea of so radical a duality to them. 


However, this definition is not yet complete, for it is equally 
applicable to two sorts of facts which, while being related to each 
other, must be distinguished nevertheless : these are magic and 

Magic, too, is made up of beliefs and rites. Like religion, it 
has its myths and its dogmas ; only they are more elementary, 
undoubtedly because, seeking technical and utilitarian ends, it 
does not waste its time in pure speculation. It has its ceremonies, 
sacrifices, lustrations, prayers, chants and dances as well. The 
beings which the magician invokes and the forces which he 
throws in play are not merely of the same nature as the forces 
and beings to which religion addresses itself ; very frequently, 
they are identically the same. Thus, even with the most inferior 
societies, the souls of the dead are essentially sacred things, and 
the object of religious rites. But at the same time, they play a 
considerable rôle in magic. In Australia ^ as well as in Melanesia, * 
in Greece as well as among the Christian peoples,^ the souls of 
the dead, their bones and their hair, are among the intermediaries 
used the most frequently by the magician. Demons are also a 
common instrument for magic action. Now these demons are 
also beings surrounded with interdictions ; they too are separated 
and live in a world apart, so that it is frequently difficult to 

* See Spencer and Gillen, Native Tribes of Central Australia, pp. 534 ff. ; 
Northern Tribes of Central Australia, p. 463 ; Howitt, Native Tribes of S.E. 
Australia, pp. 359-361. 

* See Codrington, The Melanesians, ch. xii. 

* See Hubert, art. Magia in Dictionnaire des Antiquités. 

Definition of Religious Phenomena and of Religion 43 

distinguish them from the gods properly so-called.^ Moreover, in 
Christianity itself, is not the devil a fallen god, or even leaving 
aside all question of his origin, does he not have a religious char- 
acter from the mere fact that the hell of which he has charge is 
something indispensable to the Christian religion ? There are 
even some regular and official deities who are invoked by the 
magician. Sometimes these are the gods of a foreign people ; 
for example, Greek magicians called upon Egyptian, Assyrian or 
Jewish gods. Sometimes, they are even national gods : Hecate 
and Diana were the object of a magic cult ; the Virgin, Christ 
and the saints have been utilized in the same way by Christian 
magicians. 2 

Then will it be necessary to say that magic is hardly dis- 
tinguishable from religion ; that magic is full of religion just as 
rehgion is full of magic, and consequently that it is impossible 
to separate them and to define the one without the other ? It is 
difficult to sustain this thesis, because of the marked repugnance 
of religion for magic, and in return, the hostility of the second 
towards the first. Magic takes a sort of professional pleasure in 
profaning holy things ; ^ in its rites, it performs the contrary of 
the rehgious ceremony.'* On its side, religion, when it has not 
condemned and prohibited magic rites, has always looked upon 
them with disfavour. As Hubert and Mauss have remarked, there 
is something thoroughly anti-religious in the doings of the 
magician.^ Whatever relations there may be between these two 
sorts of institutions, it is difficult to imagine their not being 
opposed somewhere ; and it is still more necessary for us to find 
where they are differentiated, as we plan to limit our researches 
to religion, and to stop at the point where magic commences. 

Here is how a line of demarcation can be traced between these 
two domains. 

The really religious beliefs are always common to a determined 
group, which makes profession of adhering to them and of prac- 
tising the rites connected with them. They are not merely 
received individually by all the members of this group ; they 
are something belonging to the group, and they make its unity. 
The individuals which compose it feel themselves united to each 
other by the simple fact that they have a common faith. A 

^ For example, in Melanesia, the tindalo is a spirit, now religious, now magic 
(Codrington, pp. 125 ff., 194 S.). 

* See Hubert and Mauss, Théorie Générale de la Magie, in Artnée Sociologique, 
vol. VII, pp. 83-84. 

* For example, the host is profaned in the black mass. 

* One turns his back to the altar, or goes around the altar commencing by the 
left instead of by the right. 

' Loc. cit., p. 19. 

44 Elementary Forms of Religious Life 

society whose members are united by the fact that thçy think in 
the same way in regard to the sacred world and its relations with 
the profane world, and by the fact that they translate these 
common ideas into common practices, is what is called a Church. 
In all history, we do not find a single religion without a Church. 
Sometimes the Church is strictly national, sometimes it passes 
the frontiers ; sometimes it embraces an entire people (Rome, 
Athens, the Hebrews), sometimes it embraces only a part of them 
(the Christian societies since the advent of Protestantism) ; 
sometimes it is directed by a corps of priests, sometimes it is 
almost completely devoid of any official directing body.^ But 
wherever we observe the religious life, we find that it has a definite 
group as its foundation. Even the so-called private cults, such 
as the domestic cult or the cult of a corporation, satisfy this 
condition ; for they are always celebrated by a group, the family 
or the corporation. Moreover, even these particular religions 
are ordinarily only special forms of a more general religion 
which embraces all ; ^ these restricted Churches are in reality 
only chapels of a vaster Church which, by reason of this very 
extent, merits this name still more.^ 

It is quite another matter with magic. To be sure, the belief 
in magic is always more or less general ; it is very frequently 
diffused in large masses of the population, and there are even 
peoples where it has as many adherents as the real religion. But 
it does not result in binding together those who adhere to it, nor 
in uniting them into a group leading a common life. There is no 
Church of magic. Between the magician and the individuals- 
who consult him, as between these individuals themselves, there 
are no lasting bonds which make them members of the same moral 
community, comparable to that formed IBy The believers in the 
'same god or the observers of the same cult. The magician has a 
clientele and not a Church, and it is very possible that his clients 
have no other relations between each other, or even do not know 
each other ; even the relations which they have with him are 
generally accidental and transient ; they are just like those of a 
sick man with his physician. The official and public character 

^ Undoubtedly it is rare that a ceremony docs not have some director at the 
moment when it is celebrated ; even* in the most crudely organized societies, 
there are generally certain men whom the importance of their social position 
points out to exercise a directing influence over the religious life (for example, 
the chiefs of the local groups of certain Australian societies). But this attribution 
of functions is still very uncertain. 

* At Athens, the gods to whom the domestic cult was addressed were only 
specialized forms of the gods of the city (Zei's ^-T?;(7ios, Zei;s ép/cetos). In the 
same way, in the Middle Ages, the patrons of the guilds were saints of the 

' For the name Church is ordinarily applied only to a group whose common 
beliefs refer to a circle of more special affairs. 

Definition of Religious Phenomena and of Religion 45 

with which he is sometimes invested changes nothing in this 
situation ; the fact that he works openly does not unite him 
more regularly or more durably to those who have recourse to 
his services. 

It is true that in certain cases, magicians form societies among 
themselves : it happens that they assemble more or less 
periodically to celebrate certain rites in common ; it is well 
known what a place these assemblies of witches hold in European 
folk-lore. But it is to be remarked that these associations are in 
no way indispensable to the working of the magic ; they are even 
rare and rather exceptional. The magician has no need of uniting 
himself to his fellows to practise his art. More frequently, he is 
a recluse ; in general, far from seeking society, he flees it. " Even 
in regard to his colleagues, he always keeps his personal inde- 
pendence." ^ Religion, on the other hand, is inseparable from 
the idea of a Church. From this point of view, there is an essential 
difference between magic and religion. But what is especially 
important is that when these societies of magic are formed, 
they do not include all the adherents to magic, but only the 
magicians ; the laymen, if they may be so called, that is to say, 
those for whose profit the rites are celebrated, in fine, those who 
represent the worshippers in the regular cults, are excluded. 
Now the magician is for magic what the priest is for religion, but 
a college of priests is not a Church, any more than a religious 
congregation which should devote itself to some particular saint 
in the shadow of a cloister, would be a particular cult. A Ch urch 
is not a fraternity of priests ; itJS-ajiimuIxuininunity-foFmed by 
all the believers in a sjngle faith, layme n as w^ell as priests. But 
magic lacks any such cQijii^mruty.^ 

But if the idea of a Church is made to enter into the definition 
of religion, does that not exclude the private religions which the 
individual establishes for himself and celebrates by himself ? 
There is scarcely a society where these are not found. Every 
Ojibway, as we shall see below, has his own personal manitou, 
which he chooses himself and to which he renders special religious 
services ; the Melanesian of the Banks Islands has his tamaniu ; ' 
the Roman, his genius ; * the Christian, his patron saint and 
guardian angel, etc. By definition all these cults seem to be 

* Hubert and Mauss, loc. cit., p. i8. 

* Robertson Smith has already pointed out that magic is opposed to religion, 
as the individual to the social {The Religion of the Semites, 2 edit., pp. 264-265). 
Also, in thus distinguishing magic from religion, we do not mean to establish a 
break of continuity between them. The frontiers between the two domains are 
frequently uncertain. 

' Codrington, Trans, and Proc. Roy. Soc. of Victoria, XVI, p. 136. 

* Negrioli, Dei Gem: presso i Romani. 

46 . Elementary Forms of Religious Life 

independent of all idea of the group. Not only are these indi- 
pvidual religions very frequent in history, but nowadays many 
I are asking if they are not destined to be the pre-eminent form of 
I the religious life, and if the day will not come when there will be 
no other cult than that which each man will freely perform within 
L himself. 1 

But if we leave these speculations in regard to the future aside 
for the moment, and confine ourselves to religions such as they 
are at present or have been in the past, it becomes clearly evident 
that these individual cults are not distinct and autonomous 
religious systems, but merely aspects of the common religion of 
the whole Church, of which the individuals are members. The 
patron saint of the Christian is chosen from the official Ust of 
saints recognized by the Catholic Church ; there are even canonical 
rules prescribing how each Catholic should perform this private 
cult. In the same way, the idea that each man necessarily has a 
protecting genius is found, under different forms, at the basis of 
a great number of American religions, as well as of the Roman 
religion (to cite only these two examples) ; for, as will be seen 
later, it is very closely connected with the idea of the soul, and 
this idea of the soul is not one of those which can be left entirely 
to individual choice. In a word, it is the Church of which he is 
a member which teaches the individual what these personal gods 
are, what their function is, how he should enter into relations 
with them and how he should honour them. When a methodical 
analysis is made of the doctrines of any Church whatsoever, 
sooner or later we come upon those concerning private cults. 
So these are not two religions of different types, and turned in 
opposite directions ; both are made up of the same ideas and the 
same principles, here applied to circumstances which are of 
interest to the group as a whole, there to the life of the individual. 
This solidarity is even so close that among certain peoples, ^ the 
ceremonies by which the faithful first enter into communication 
with their protecting geniuses are mixed with rites whose public 
character is incontestable, namely the rites of initiation.' 

^ Tliis is the conclusion reached by Spencer in his Ecclesiastical Institutions 
(ch. xvi), and by Sabatier in his Outlines of a Philosophy of Religion, based on 
Psychology and History (tr. by Seed), and by all the school to which he belongs. 

* Notably among numerous Indian tribes of North America. 

^ This statement of fact does not touch the question whether exterior and 
public religion is not merely the development of an interior and personal 
rehgion which was the primitive fact, or whether, on the contrary, the second 
is not the projection of the first into individual consciences. The problem 
will be directly attacked below (Bk. II, ch. v, § 2, cf. the same book, ch. vi and 
vii, § i^. For the moment, we confine ourselves to remarking that the individual 
cult is presented to the observer as an element of, and something dependent upon, 
the collective cult. 

Definition of Religious Phenomena and of Religion 47 

There still remain those contemporary aspirations towards a 
religion which would consist entirely in internal and subjective 
states, and which would be constructed freely by each of us. 
But howsoever real these aspirations may be, they cannot affect 
our definition, for this is to be applied only to facts already 
realized, and not to uncertain possibilities. One can define 
religions such as they are, or such as they have been, but not 
such as they more or less vaguely tend to become. It is possible 
that this religious individualism is destined to be realized in facts ; 
but before we can say just how far this may be the case, we must 
first know what religion is, of what elements it is made up, from 
what causes it results, and what function it fulfils — all questions 
whose solution cannot be foreseen before the threshold of our 
study has been passed. It is only at the close of this study that 
we can attempt to anticipate the future. 

Thus we arrive at the following definition : A religion is a 
unified system of beliefs and practices relative to sacred things, thatf 
is to say, things set apart and forbidden — beliefs and practices which^ 
unite into one single moral community called a Church, all those who 
adhere to them. The second element which thus finds a place in. 
our definition is no less essential than the first ; for by showing 
that the idea of religion is inseparable from that of the Church, it 
makes it clear that religion should be an eminently collective 

* It is by this that our present definition is connected to the one we have 
already proposed in the Année Sociologique. In this other work, we defined 
religious beliefs exclusively by their obligatory character ; but, as we shall show, 
this obligation evidently comes from the fact that these beliefs are the possession 
of a group which imposes them upon its members. The two definitions are thus 
in a large part the same. If we have thought it best to propose a new one, it is 
because the first was too formal, and neglected the contents of the religious 
representations too much. It will be seen, in the discussions which follow, how 
important it is to put this characteristic into evidence at once. Moreover, if 
their imperative character is really a distinctive trait of religious beliefs, it allows 
of an infinite number of degrees ; consequently there are even cases where it is 
not easily perceptible. Hence come difficulties and embarrassments which are 
avoided by substituting for this critérium the one we now employ. 



I. — Animism 

ARMED with this definition, we are now able to set out in 
L search of this elementary religion which we propose to 

Even the crudest religions with which history and ethnology 
make us acquainted are already of a complexity which corre- 
sponds badly with the idea sometimes held of primitive mentality. 
One finds there not only a confused system of beliefs and rites, 
but also such a plurality of different principles, and such a 
richness of essential notions, that it seems impossible to see in 
them anything but the late product of a rather long evolution. 
Hence it has been concluded that to discover the truly original 
form of the religious life, it is necessary to descend by analysis 
beyond these observable religions, to resolve them into their 
common and fundamental elements, and then to seek among 
these latter some one from which the others were derived. 

To the problem thus stated, two contrary solutions have 
been given. 

There is no religious system, ancient or recent, where one 
does not meet, under different forms, two religions, as it were, 
side by side, which, though being united closely and mutually 
penetrating each other, do not cease, nevertheless, to be dis- 
tinct. The one addresses itself to the phenomena of nature, 
either the great cosmic forces, such as winds, rivers, stars or the 
sky, etc., or else the objects of various sorts which cover the 
surface of the earth, such as plants, animals, rocks, etc. ; for 
this reason it has been given the name of naiurism. The other 
has spiritual beings as its object, spirits, souls, geniuses, demons, 
divinities properly so-called, animated and conscious agents 
like man, but distinguished from him, nevertheless, by the 
nature of their powers and especially by the peculiar characteristic 
that they do not affect the senses in the same way : ordinarily 
they are not visible to human eyes. This religion of spirits is 
called animism. Now, to explain the universal co-existence of 


Leading Conceptions of the Elementary Religion 49 

these two sorts of cults, two contradictory theories have been 
proposed. For some, animism is the primitive rehgion, of which 
naturism is only a secondary and derived form. For the others, 
on the contrary, it is the nature cult which was the point of 
departure for religious evolution ; the cult of spirits is only a 
peculiar case of that. 

These two theories are, up to the present, the only ones by 
which the attempt has been made to explain rationally^ the 
origins of religious thought. Thus the capital problem raised 
by the history of religions is generally reduced to asking which 
of these two solutions should be chosen, or whether it is not 
better to combine them, and in that case, what place must be 
given to each of the two elements. ^ Even those scholars who do 
not admit either of these hypotheses in their systematic form, 
do not refuse to retain certain propositions upon which they rest.^ 
Thus we have a certain number of theories already made, which 
must be submitted to criticism before we take up the study of 
the facts for ourselves. It will be better understood how in- 
dispensable it is to attempt a new one, when we have seen the 
insufficiency of these traditional conceptions. 


It is Tylor who formed the animist theory in its essential 
outlines.* Spencer, who took it up after him, did not reproduce 
it without introducing certain modifications.^ But in general 
the questions are posed by each in the same terms, and the 
solutions accepted are, with a single exception, identically the 
same. Therefore we can unite these two doctrines in the exposi- 
tion which follows, if we mark, at the proper moment, the place 
where the two diverge from one another. 

^ We thus leave aside here those theories which, in whole or in part, make 
use of super-experimental data. This is the case with the theory which Andrew 
Lang exposed in his book. The Making of Religion, and which Father Schmidt 
has taken up again, with variations of detail, in a series of articles on The Origin 
of the Idea of God [Anthropos, 1908, 1909). Lang does not set animism definitely 
aside, but in the last analysis, he admits a sense or intuitibn of the divine directly. 
Also, if we do not consider it necessary to expose and discuss this conception m 
the present chapter, we do not intend to pass it over in silence ; we shall come 
to it again below, when we shall ourselves explain the facts upon which it is 
founded (Bk. II, ch. ix, § 4). 

* This is the case, for example, of Fustel de Coulanges who accepts the two 
conceptions together {The Ancient City, Bk. I and Bk. Ill, ch. ii). 

* This is the case with Jevons, who criticizes the animism taught by Tylor, 
but accepts his theories on the origin of the idea of the soul and the anthropo- 
morphic instinct of man. Inversely, Usener, in his Gotternamen, rejects certain 
hypotheses of Max Miiller which will be described below, but admits the principal 
postulates of naturism. 

* Primitive Culture, chs. xi-xviii. 

* Principles of Sociology, Parts I and VI, 

50 Elementary Forms of Religious Life 

In order to find the elementary form of the reUgious life in 
these animistic beliefs and practices, three desiderata must be 
satisfied : first, since according to this hypothesis, the idea of 
the soul is the cardinal idea of religion, it must be shown how 
this is formed without taking any of its elements from an anterior 
religion ; secondly, it must be made clear how souls become the 
object of a cult and are transformed into spirits ; and thirdly 
and finally, since the cult of these spirits is not all of any religion, 
it remains to be explained how the cult of nature is derived 
from it. 

According to this theory, the idea of the soul was first suggested 
to men by the badly understood spectacle of the double life they 
ordinarily lead, on the one hand, when awake, on the other, 
when asleep. In fact, for the savage, ^ the mental representa- 
tions which he has while awake and those of his dreams are 
said to be of the same value : he objectifies the second like the 
first, that is to say, that he sees in them the images of external 
objects whose appearance they more or less accurately reproduce. 
So when he dreams that he has visited a distant country, he 
believes that he really was there. But he could not have gone 
there, unless two beings exist within him : the one, his body, 
which has remained lying on the ground and which he finds in 
the same position on awakening ; the other, during this time, 
has travelled through space. Similarly, if he seems to talk with 
one of his companions who he knows was really at a distance, 
he concludes that the other also is composed of two beings : 
one which sleeps at a distance, and another which has come to 
manifest himself by means of the dream. From these repeated 
experiences, he little by little arrives at the idea that each of us 
has a double, another self, which in determined conditions has 
the power of leaving the organism where it resides and of going 
roaming at a distance. 

Of course, this double reproduces all the essential traits of the 
perceptible being which serves it as external covering ; but at 
the same time it is distinguished from this by many character- 
istics. It is more active, since it can cover vast distances in 
an instant. It is more malleable and plastic ; for, to leave the 
body, it must pass out by its apertures, especially the mouth and 
nose. It is represented as made of matter, undoubtedly, but of 
a matter much more subtile and etherial than any which we 

^ This is the word used by Tylor. It has the inconvenience of seeming to 
imply that men, in the proper sense of the term, existed before there was a 
civilization. However, there is no proper term for expressing the idea ; that 
of primitive, which we prefer to use, lacking a better, is, as we have said, far from 

Leading Conceptions of the Elementary Religion 51 

know empirically. This double is the soul. In fact, it cannot be 
doubted that in numerous societies the soul has been conceived 
in the image of the body ; it is believed that it reproduces even 
the accidental deformities such as those resulting from wounds 
or mutilations. Certain Australians, after having killed their 
enemy, cut off his right thumb, so that his soul, deprived of its 
thumb also, cannot throw a javelin and revenge itself. But 
while it resembles the body, it has, at the same time, something 
half spiritual about it. They say that "it is the finer or more 
aeriform part of the body," that " it has no flesh nor bone nor 
sinew ' ' ; that when one wishes to take hold of it, he feels nothing ; 
that it is " like a purified body."^ 

Also, other facts of experience which affect the mind in the 
same way naturally group themselves around this fundamental 
fact taught by the dream : fainting, apoplexy, catalepsy, 
ecstasy, in a word, all cases of temporary insensibility. In fact, 
they all are explained very well by the hypothesis that the 
principle of life and feeling is able to leave the body momentarily. 
Also, it is natural that this principle should be confounded with 
the double, since the absence of the double during sleep daily 
has the effect of suspending thought and life. Thus diverse 
observations seem to agree mutually and to confirm the idea 
of the constitutional duality of man.^ 

4_But the soul is not a spirit. It is attached to a body which it 
can leave only by exception ; in so far as it is nothing more 
than that, it is not the object of any cult. The spirit, on the 
other hand, though generally having some special thing as its 
residence, can go away at will, and a man can enter into relations 
with it only by observing ritual precautions, s^he soul can 
become a spirit, then, only by transforming itself : the simple 
application of these preceding ideas to the fact of death pro- 
duced this metamorphosis quite naturally. For a rudimentary 
intelligence, in fact, death is not distinguished from _ a long 
fainting swoon or a prolonged sleep ; it has all their aspects. 
Thus it seems that it too consists in a separation of the soul and 
the body, analogous to that produced every night ; but as in 
such cases, the body is not reanimated, the 'dea is formed of a 
separation without an assignable limit of time. When the body 
is once destroyed — and funeral rites have the object of hastening 
this destruction — the separation is taken as final. Hence come 
spirits detached from any organism and left free in space.-' As 

* Tylor, op. cit., I, pp. 455 f . 

• See Spencer, Principles of Sociology, I, pp. 143 fi., and Tylor, op. cit., I, 
pp. 434 fi., 445 fi. 

52 Elementary Forms of Religious Life 

their number augments with time, a population of souls forms 
around the living population. These souls of m^i have the 
needs and passions of men ; they seek to concern themselves 
with the life of their companions of yesterday, either to aid 
them or to injure them, according to the sentiments which they 
have kept towards them. According to the circumstances, their 
nature makes them either very precious auxiliaries or very 
redoubtable adversaries. Owing to their extreme fluidity, they 
can even enter into the body, and cause all sorts of disorders 
there, or else increase its vitality. Thus comes the habit of 
attributing to them all those events of life which vary slightly 
from the ordinary : there are very few of these for which they 
cannot account. Thus they constitute a sort of ever- ready 
supply of causes which never leaves one at a loss when in search 
of explanations. Does a man appear inspiied, does he speak 
with energy, is it as though he were lifted outside himself and 
above the ordinary level of men ? It is because a good spirit is 
in him and animates him. Is he overtaken by an attack or 
seized by madness ? It is because an evil spirit has entered into 
him and brought him all this trouble. There are no maladies 
which cannot be assigned to some influence of this sort. Thus 
the power of souls is increased by all that men attribute to 
them, and in the end men find themselves the prisoners of this 
imaginary world of which they are, however, the authors and 
the models. They fall into dependence upon these spiritual 
forces which they have created with their owti hands and in 
their own image. For if souls are the givers of health and sick- 
ness, of goods and evils to this extent, it is wise to conciliate 
their favour or appease them when they are irritated ; hence 
come the offerings, prayers, sacrifices, in a word, all the apparatus 
of religious observances. ^ 

Here is the soul transformed. From a simple vital principle 
animating the body of a man, it has become a spirit, a good or 
evil genius, or even a deity, according to the importance of 
the effects with which it is charged. But since it is death which 
brought about this apotheosis, it is to the dead, to the souls of 
ancestors, that the first cult known to humanity was addressed. 
Thus the first rites were funeral rites ; the first sacrifices were 
food offerings destined to satisfy the needs of the departed ; 
the first altars were tombs. ^ 

But since these spirits were of human origin, they interested 
themselves only in the life of men and were thought to act only 
upon human events. It is still to be explained how other spirits 

1 Tylor, II, pp. 113 fi. » Tylor, I, pp. 481 fi. 

Leading Conceptions of the Elementary Religion 53 

were imagined to account for the other phenomena of the universe 
and how the cult of nature was subsequently formed beside that 
of the ancestors. 

For Tylor, this extension of animism was due to the particular 
mentality of the primitive who, like an infant, cannot distinguish 
the animate and the inanimate. Since the first beings of which 
the child commences to have an idea are men, that is, himself 
and those around him, it is upon this model of human nature 
that he tends to think of everything. The toys with which he 
plays, or the objects of every sort which affect his senses, he 
regards as living beings like himself. Now the primitive thinks 
like a child. Consequently, he also is inclined to endow all 
things, even inanimate ones, with a nature analogous to his 
own. Then if, for the reasons exposed above, he once arrives 
at the idea that man is a body animated by a spirit, he must 
necessarily attribute a duality of this sort and souls like his own 
even to inert bodies themselves. Yet the sphere of action of 
the two could not be the same. The souls of men have a direct 
influence only upon the world of men : they have a marked 
preference for the human organism, even when death has given 
them their liberty. On the other hand, the souls of things reside 
especially in these things, and are regarded as the productive 
causes of all that passes there. The first account for health and 
sickness, skilfulness or unskilfulness, etc. ; by the second are 
explained especially the phenomena of the physical world, the 
movement of water-courses or the stars, the germination of 
plants, the reproduction of animals, etc. Thus the first philosophy 
of man, which is at the basis of the ancestor-cult, is completed 
by a philosophy of the world. 

In regard to these cosmic spirits, man finds himself in a state 
of dependence still more evident than that in regard to the 
wandering doubles of his ancestors. For he could have only 
ideal and imaginary relations with the latter, but he depends 
upon things in reality ; to live, he has need of their concurrence ; 
he then believes that he has an equaJ need of the spirits which 
appear to animate these things and to determine their diverse 
manifestations. He implores their assistance, he solicits them 
with offerings and prayers, and the religion of man is thus 
completed in a religion of nature. 

Herbert Spencer objects against this explanation that the 
hypothesis upon which it rests is contradicted by the facts. 
It is held, he says, that there is a time when men do not realize 
the differences which separate the animate from the inanimate. 
Now, as one advances in the animal scale, he sees the ability 
to make this distinction develop. The superior animals do not 

54 Elementary Forms of Religious Life 

confound an object which moves of itself and whose movements 
are adapted to certain ends, with those which are mechanically 
moved from without. " Amusing herself with a mouse she has 
caught, the cat, if it remains long stationary, touches it with her 
paw to make it run. Obviously the thought is that a living thing 
disturbed will try to escape."^ Even the primitive men could not 
have an intelligence inferior to that of the animals which preceded 
them in evolution ; then it cannot be for lack of discernment that 
they passed from the cult of ancestors to the cult of things. 

According to Spencer, who upon this point, but upon this 
point only, differs from Tylor, this passage was certainly due 
to a confusion, but to one of a different sort. It was, in a large 
part at least, the result of numerous errors due to language. 
In many inferior societies it is a very common custom to give to 
each individual, either at his birth or later, the name of some 
animal, plant, star or natural object. But as a consequence of 
the extreme imprecision of his language, it is very difficult for 
a primitive to distinguish a metaphor from the reality. He soon 
lost sight of the fact that these names were only figures, and 
taking them literally, he ended by believing that an ancestor 
named " Tiger " or " Lion " was really a tiger or a lion. Then 
the cult of which the ancestor was the object up to that time, 
was changed over to the animal with which he was thereafter 
confounded ; and as the same substitution went on for the 
plants, the stars and all the natural phenomena, the religion 
of nature took the place of the old rehgion of the dead. Besides 
this fundamental confusion, Spencer signalizes others which 
aided the action of the first from time to time. For example, 
the animals which frequent the surroundings of the tombs or 
houses of men have been taken for their reincarnated souls, and 
adored under this title ;2 or again, the mountain which tradition 
made the cradle of the race was finally taken for the ancestor 
of the race ; it was thought that men were descended from it 
because their ancestors appeared coming from it, and it was 
consequently treated as an ancestor itself.^ But according to 
the statement of Spencer, these accessory causes had only a 
secondary influence ; that which principally determined the 
institution of naturism was " the literal interpretation of 
metaphorical names."* 

We had o mention this theory to have our exposition of 
animism complete ; but it is too inadequate for the facts, and 
too universally abandoned to-day to demand that we stop any 
longer for it. In order to explain a fact as general as the religion 

* Principles of Sociology, I, p. 126. 

» Ibid., pp. 322 ff. • Ibid., pp. 366-367. * Ibid., p. 346. Cf. p. 384. 

Leading Conceptions of the Elementary Religion 55 

of nature by an illusion, it would be necessary that the illusion 
invoked should have causes of an equal generality. Now even 
if misunderstandings, such as those of which Spencer gives some 
rare illustrations, could explain the transformation of the cult 
of ancestors into that of nature, it is not clear why this should 
be produced with a sort of universality. No psychical mechanism 
necessitated it. It is true that because of its ambiguity, the word 
might lead to an equivocation ; but on the other hand, all the 
personal souvenirs left by the ancestor in the memories of men 
should oppose this confusion. Why should the tradition which 
represented the ancestor such as he really was, that is to say, 
as a man who led the life of a man, everywhere give way before 
the prestige of a word ? Likewise, one should have a little 
difficulty in admitting that men were bom of a mountain or a 
star, of an animal or a plant ; the idea of a similar exception to 
the ordinary conceptions of generation could not fail to raise 
active resistance. Thus, it is far from true that the error found 
a road all prepared before it, but rather, all sorts of reasons should 
have kept it from being accepted. It is difficult to understand 
how, in spite of all these obstacles, it could have triumphed so 


The theory of Tylor, whose authority is always great, still 
remains. His hypotheses on the dream and the origin of the 
ideas of the soul and of spirits are still classic ; it is necessary, 
therefore, to test their value. 

First of all, it should be recognized that the theorists of 
animism have rendered an important service to the science of 
religions, and even to the general history of ideas, by submitting 
the idea of the soul to historical analysis. Instead of following 
so niany philosophers and making it a simple and immediate 
object of consciousness, they have much more correctly viewed 
it as a complex whole, a product of history and mythology. It 
cannot be doubted that it is something essentially religious in 
its nature, origin and functions. It is from religion that the 
philosophers received it ; it is impossible to understand the form 
in which it is represented by the thinkers of antiquity, if one 
does not take into account the mythical elements which served 
in its formation. 

But if Tylor has had the merit of raising this problem, the 
solution he gives raises grave difficulties. 

First of all, there are reservations to be made in regard to the 
very principle which is at the basis of this theory. It is taken 

56 Elementary Forms of Religious Life 

for granted that the soul is entirely distinct from the body, that 
it is its double, and that within it or outside of it, it normally lives 
its own autonomous life. Now we shall see^ that this conception 
is not that of the primitive, or at least, that it only expresses 
one aspect of his idea of the soul. For him, the soul, though 
being under certain conditions independent of the organism 
which it animates, confounds itself with this latter to such an 
extent that it cannot be radically separated from it : there are 
organs which are not only its appointed seat, but also its outward 
form and material manifestation. The notion is therefore more 
complex than the doctrine supposes, and it is doubtful conse- 
quently whether the experiences mentioned are sufficient to 
account for it ; for even if they did enable us to understand how 
men have come to believe themselves double, they cannot 
explain how this duality does not exclude, but rather, implies 
a deeper unity and an intimate interpénétration of the two beings 
thus differentiated. 

But let us admit that the idea of the soul can be reduced to 
the idea of a double, and then see how this latter came to be 
formed. It could not have been suggested to men except by the 
experience of dreams. That they might understand how they 
could see places more or less distant during sleep, while their 
bodies remained lying on the ground, it would seem that they 
were led to conceive of themselves as two beings : on the one 
hand, the body, and on the other, a second self, able to leave 
the organism in which it lives and to roam about in space. But 
if this hypothesis of a double is to be able to impose itself upon 
men with a sort of necessity, it should be the only one possible, 
or at least, the most economical one. Now as a matter of fact, 
there are more simple ones which, it would seem, might have 
occurred to the mind just as naturally. For example, why 
should the sleeper not imagine that while asleep he is able to 
see things at a distance ? To imagine such a power would 
demand less expense to the imagination than the construction 
of this complex notion of a double, made of some etherial, semi- 
invisible substance, and of which direct experience offers no 
example. But even supposing that certain dreams rather 
naturally suggest the animistic explanation, there are certainly 
many others which are absolutely incompatible with it. Often 
our dreams are concerned with passed events ; we see again the 
things which we saw or did yesterday or the day before or even 
during our youth, etc. ; dreams of this sort are frequent and 
hold a rather considerable place in our nocturnal life. But the 
idea of a double cannot account for them. Even if the double 

1 See below, Bk. II, ch. viii. 

Leading Conceptions of the Elementary Religion 57 

can go from one point to another in space, it is not clear how it 
could possibly go back and forth in time. Howsoever rudi- 
mentary his intelligence may be, how could a man on awakening 
believe that he had really been assisting at or taking part in 
events which he knows passed long before ? How could he 
imagine that during his sleep he lived a life which he knows has 
long since gone by ? It would be much more natural that he 
should regard these renewed images as merely what they really 
are, that is, as souvenirs like those which he has during the day, 
but ones of a special intensity. 

Moreover, in the scenes of which we are the actors and witnesses 
while we sleep, it constantly happens that one of our contem- 
poraries has a rôle as well as ourselves : we think we see and hear 
him in the same place where we see ourselves. According to the 
animists, the primitive would explain this by imagining that his 
double was visited by or met with those of certain of his com- 
panions. But it would be enough that on awakening he question 
them, to find that their experiences do not coincide with his. 
During this same time, they too have had dreams, but wholly 
different ones. They have not seen themselves participating in the 
same scene ; they believe that they have visited wholly different 
places. Since such contradictions should be the rule in these 
cases, why should they not lead men to believe that there 
had probably been an error, that they had merely imagined it, 
that they had been duped by illusions ? This blind credulity 
which is attributed to the primitive is really too simple. It is 
not true that he must objectify all his sensations. He cannot 
live long without perceiving that even when awake his senses 
sometimes deceive him. Then why should he believe them more 
infallible at night than during the day ? Thus we find that 
there are many reasons opposing the theory that he takes his 
dreams for the reality and interprets them by means of a double 
of himself. 

But more than that, even if every dream were well explained 
by the hypothesis of a double, and could not be explained 
otherwise, it would remain a question why men have attempted 
to explain them. Dreams undoubtedly constitute the matter 
of a possible problem. But we pass by problems every day 
which we do not raise, and of which we have no suspicion until 
some circumstance makes us feel the necessity of raising them. 
Even when the taste for pure speculation is aroused, reflection 
is far from raising all the problems to which it could eventually 
apply itself ; only those attract it which present a particular 
interest. Especially, when it is a question of facts which always 
take place in the same manner, habit easily numbs cliriosity, and 

58 Elementary Forms of Religious Life 

we do not even dream of questioning them. To shake off this 
torpor, it is necessary that practical exigencies, or at least a very 
pressing theoretical interest, stimulate our attention and turn 
it in this direction. That is why, at every moment of history, 
there have been so many things that we have not tried to under- 
stand, without even being conscious of our renunciation. Up 
until very recent times, it was believed that the sun was only 
a few feet in diameter. There is something incomprehensible 
in the statement that a luminous disc of such slight dimensions 
could illuminate the world : yet for centuries men never thought 
of resolving this contradiction. The fact of heredity has been 
known for a long time, but it is very recently that the attempt 
has been made to formulate its theory. Certain beliefs were even 
admitted which rendered it wholly unintelligible : thus in many 
Australian societies of which we shall have occasion to speak, 
the child is not physiologically the offspring of its parents.^ 
This intellectual laziness is necessarily at its maximum among 
the primitive peoples. These weak beings, who have so much 
trouble in maintaining life against all the forces which assail it, 
have no means for supporting any luxury in the way of specula- 
tion. They do not reflect except when they are driven to it. 
Now it is difficult to see what could have led them to make 
dreams the theme of their meditations. What does the dream 
amount to in our lives ? How little is the place it holds, especially 
because of the very vague impressions it leaves in the memory, 
and of the rapidity with which it is effaced from remembrance, 
and consequently, how surprising it is that a man of so rudi- 
mentary an intelligence should have expended such efforts to 
find its explanation ! Of the two existences which he successively 
leads, that of the day and that of the night, it is the first which 
should interest him the most. Is it not strange that the second 
should have so captivated his attention that he made it the 
basis of a whole system of complicated ideas destined to have 
so profound an influence upon his thought and conduct ? 

Thus all tends to show that, in spite of the credit it still enjoys, 
the animistic theory of the soul must be revised. It is true that 
to-day the primitive attributes his dreams, or at least certain 
of them, to displacements of his double. But that does not say 
that the dream actually furnished the materials out of which 
the idea of the double or the soul was first constructed ; it might 
have been applied afterwards to the phenomena of dreams, 
ecstasy and possession, without having been derived from them. 
It is very frequent that, after it has been formed, an idea is 

^ See Spencer and Gillen, The Native Tribes of Central Australia, pp. 123-127 ; 
Strehlow. Die Aranda- und Loritja-Stamme in Zentral Australien, II, pp. 52 Q. 

Leading Conceptions of the Elementary Religion 59 

employed to co-ordinate or illuminate — with a light frequently 
more apparent than real — certain facts with which it had no 
relation at first, and which would never have suggested it them- 
selves. God and the immortality of the soul are frequently 
proven to-day by showing that these beliefs are implied in the 
fundamental principles of morality ; as a matter of fact, they 
have quite another origin. The history of religious thought 
could furnish numerous examples of these retrospective justi- 
fications, which can teach us nothing of the way in which the 
ideas were formed, nor of the elements out of which they are 

It is also probable that the primitive distinguishes between 
his dreams, and does not interpret them all in the same way. 
In our European societies the still numerous persons for whom 
sleep is a sort of magico-religious state in which the mind, being 
partially relieved of the body, has a sharpness of vision which 
it does not enjoy during waking moments, do not go to the point 
of considering all their dreams as so many mystic intuitions : 
on the contrary, along with everybody else, they see in the 
majority of their dreams only profane conditions, vain plays 
of images, or simple hallucinations. It might be supposed that 
the primitive should make analogous distinctions. Codrington 
says distinctly that the Melanesians do not attribute all their 
dreams indiscriminately to the wanderings of their souls, but 
merely those which strike their imagination forcibly i^ un- 
doubtedly by that should be understood those in which the 
sleeper imagines himself in relations with religious beings, good 
or evil geniuses, souls of the dead, etc. Similarly, the Dieri in 
Australia sharply distinguish ordinary dreams from those 
nocturnal visions in which some deceased friend or relative 
shows himself to them. In the first, they see a simple fantasy 
of their imagination ; they attribute the second to the action 
of an evil spirit. 2 All the facts which Howitt mentions as 
examples to show how the Australian attributes to the soul the 
power of leaving the body, have an equally mystic character. 
The sleeper believes himself transported into the land of the 
dead or else he converses with a dead companion.^ These 
dreams are frequent among the primitives.* It is probably 

* The Melanesians, pp. 249-250. 

* Howitt, The Native Tribes of South-Eastern Australia, p. 358. 
» Ibid., pp. 434-442. 

* Of the negroes of southern Guinea, Tylor says that " their sleeping hours are 
characterized by almost as much intercour-se with the dead as their waking are 
with the living " (Primitive Culture, I, p. 443). In regard to these peoples, the 
same author cites this remark of an observer : "All their dreams are construed 
into visits from the spirits of their deceased friends " (ibid., p. 443). This state- 
ment is certainly exaggerated ; but it is one more proof of the frequency of 

6o Elementary Forms of Religious Life 

upon these facts that the theory is based. To account for them, 
it is admitted that the souls of the dead come back to the hving 
during their sleep. This theory was the more readily accepted 
because no fact of experience could invalidate it. But these 
dreams were possible only where the ideas of spirits, souls and 
a land of the dead were already existent, that is to say, where 
religious evolution was relatively advanced. Thus, far from 
having been able to furnish to religion the fundamental notion 
upon which it rests, they suppose a previous religious system, 
upon which they depended.^ 


We now arrive at that which constitutes the very heart of 
the doctrine. 

Wherever this idea of a double may come from, it is not 
sufficient, according to the avowal of the animists themselves, 
to explain the formation of the cult of the ancestors which they 
would make the initial type of all religions. If this double is to 
become the object of a cult, it must cease to be a simple repro- 
duction of the individual, and must acquire the characteristics 
necessary to put it in the rank of sacred beings. It is death, 
they say, which performs this transformation. But whence 
comes the virtue which they attribute to this ? Even were the 
analogy of sleep and death sufficient to make one believe that 
the soul survives the body (and there are reservations to be made 
on this point), why does this soul, by the mere fact that it is now 
detached from the organism, so completely change its nature ? 
If it was only a profane thing, a wandering vital principle, 
during life, how does it become a sacred thing all at once, and 

mystic dreams among the primitives. The etymology which Strehlow proposes 
for the Arunta word altjiverama, which means " to dream," also tends to confirm 
this theory. This word is composed of altjira, which Strehlow translates by 
" god " and rama, which means " see." Thus a dream would be the moment 
when a man is in relations with sacred beings (Die Aranda- und Loritja-Stamme, 

I, p. 2). 

^ Andrew Lang, who also refuses to admit that the idea of the soul was 
suggested to men by their dream experiences, believes that he can derive it 
from other empirical data : these are the data of spiritualism (telepathy, distance- 
seeing, etc.). We do not consider it necessary to discuss the theory such as it 
has been exposed in his book The Making of Religion. It reposes upon the 
hypothesis that spiritualism is a fact of constant observation, and that distance- 
seeing is a real faculty of men, or at least of certain men, but it is well known 
how much this theory is scientifically contested. What is still more contestable 
is that the facts of spiritualism are apparent enough and of a sufficient frequency 
to have been able to serve as the basis for all the religious beliefs and practices 
which are connected with souls and spirits. The examination of these questions 
would carry us too far from what is the object of our study. It is still less 
necessary to engage ourselves in this examination, since the theory of Lang 
remains open to many of the objections which we shall address to that of Tylor 
in the paragraphs which follow. 

Leading Conceptions of the Elementary Religion 6i, 

the object of religious sentiments ? Death adds notKmg essential 
to it, except a greater liberty of movement. Being no longer 
attached to a special residence, from now on, it can do at any 
time what it formerly did only by night ; but the action of which 
it is capable is always of the same sort. Then why have the 
living considered this uprooted and vagabond double of their" 
former companion as anything more than an equal ? It was 
a fellow-creature, whose approach might be inconvenient ; 
it was not a divinity.^ 

It seems as though death ought to have the effect of weakening 
vital energies,, instead of strengthening them. It is, in fact, a 
very common belief in the inferior societies that the soul 
participates actively in the life of the body. If the body is 
wounded, it is wounded itself and in a corresponding place. 
Then it should grow old along with the body. In fact, there are 
peoples who do not render funeral honours to men arrived at 
senility ; they are treated as if their souls also had become 
senile.- It even happens that they regularly put to death, before 
they arrive at old age, certain privileged persons, such as kings 
or priests, who are supposed to be the possessors of powerful 
spirits whose protection the community wishes to keep. They 
thus seek to keep the spirit from being affected by the physical 
decadence of its momentary keepers ; with this end in view, 
they take it from the organism where it resides before age can 
have weakened it, and they transport it, while it has as yet 
lost nothing of its vigour, into a younger body where it will be 
able to keep its vitality intact.^ So when death results from 
sickness or old age, it seems as though the soul could retain only 
a diminished power ; and if it is only its double, it is difficult to 
see how it could survive at all, after the body is once definitely 
dissolved. From this point of view, the idea of survival is 
intelligible only with great difficulty. There is a logical and 
psychological gap between the idea of a double at liberty and 
that of a spirit to which a cult is addressed. 

This interval appears still more considerable when we realize y 
what an abyss separates the sacred world from the profane ; ,^ 

^ Jevons has made a similar remark. With Tylor, he admits that the idea of 
the soul comes from dreams, and that after it was created, men projected it into 
things. But, he adds, the fact that nature has been conceived as animated like 
men does not explain how it became the object of a cult. " The man who believes 
the bowing tree or the leaping flame to be a living thing like himself, does not 
therefore believe it to be a supernatural being — rather, so far as it is like himself, 
it, like himself, is not supernatural " {Introduction to the History of Religions, 

P- 55)- 

* See Spencer and Gillen, Nor. Tr., p. 506, and Nat. Tr., p. 512. 

' This is the ritual and mythical theme which Frazer studies in his Golden 

62 Elementary Forms of Religious Life 

it becomes evident that a simple change of degree could not be 
enough to make something pass from one category into the 
other. Sacred beings are not distinguished from profane ones 
merely by the strange or disconcerting forms which they take 
or by the greater powers which they enjoy ; between the two 
there is no common measure. Now there is nothing in the 
notion of a double which could account for so radical a hetero- 
geneity. It is said that when once freed from the body, the spirit 
can work all sorts of good or evil for the living, according to the 
way in which it regards them. But it is not enough that a being 
should disturb his neighbourhood to seem to be of a wholly 
different nature from those whose tranquillity it menaces. To 
be sure, in the sentiment which the believer feels for the things 
he adores, there always enters in some element of reserve and 
fear ; but this is a fear sui generis, derived from respect more 
than from fright, and where the dominating emotion is that 
which la majesté inspires in men. The idea of majesty is essen- 
tially religious. Then we have explained nothing of religion 
until we have found whence this idea comes, to what it corre- 
sponds and what can have aroused it in the mind. Simple souls 
of men cannot become invested with this character by the 
simple fact of being no longer incarnate. 

This is clearly shown by an example from Melanesia. The 
Melanesians believe that men have souls which leave the body 
at death ; it then changes its name and becomes what they call 
a tindalo, a natmat, etc. Also, they have a cult of the souls of 
the dead : they pray to them, invoke them and make offerings 
and sacrihces to them. But every tindalo is not the object of 
these ritual practices ; only those have this honour which come 
from men to whom public opinion attributed, during life, the 
fvery special virtue which the Melanesians call the mana. Later 
on, we shall have occasion to fix precisely the meaning which 
this word expresses ; for the time being, it will suffice to say 
that it is the distinctive character of every sacred being. As 
Codrington says, "it is what works to effect anything which 
is beyond the ordinary power of men, outside the common 
processes of nature."^ A priest, a sorcerer or a ritual formula 
have mana as well as a sacred stone or spirit. Thus the 
only tindalo to which religious services are rendered are those 
which were already sacred of themselves, when their proprietor 
was still alive. In regard to the other souls, which come 
from ordinary men, from the crowd of the profane, the same 
author says that they are " nobodies alike before and after 
death. "2 By itself, death has no deifying virtue. Since it 

1 The Melanesians, p. 119. * Ibid., p. 125. 


Leading Conceptions of the Elementary Religion 63 

brings about in a more or less complete and final fashion the 
separation of the soul from profane things, it can well reinforce 
the sacred character of the soul, if this already exists, but it 
cannot create it. — » 

Moreover, if, as the hypothesis of the animists supposes, the 
first sacred beings were really the souls of the dead and the 
first cult that of the ancestors, it should be found that the lower 
the societies examined are, the more the place given to this cult 
in the religious life. But it is rather the contrary which is true. 
The ancestral cult is not greatly developed, or even presented 
under a characteristic form, except in advanced societies like 
those of China, Egypt or the Greek and Latin cities ; on the 
other hand, it is completely lacking in the Australian societies 
which, as we shall see, represent the lowest and simplest form 
of social organization which we know. It is true that funeral 
rites and rites of mourning are found there ; but these practices 
do not constitute a cult, though this name has sometimes wrong- 
fully been given them. In reality, a cult is not a simple group 
of ritual precautions which a man is held to take in certain 
circumstances ; it is a system of diverse rites, festivals and 
ceremonies which all have this characteristic, that they reappear 
periodically. They fulfil the need which the believer feels of 
strengthening and reaffirming, at regular intervals of time, the 
bond which unites him to the sacred beings upon which he 
depends. That is why one speaks of marriage rites but not of a 
marriage cult, of rites of birth but not of a cult of the new-bom 
child ; it is because the events on the occasion of which these 
rites take place imply no periodicity. In the same way, there 
is no cult of the ancestors except when sacrifices are made on 
the tombs from time to time, when libations are poured there 
on certain more or less specific dates, or when festivals are 
regularly celebrated in honour of the dead. But the Australian 
has no relations of this sort with his dead. It is true that he 
must bury their remains according to a ritual, mourn for them 
during a prescribed length of time and in a prescribed manner, 
and revenge them if there is occasion to.^ But when he has once 
accomplished these pious tasks, when the bones are once dry 
and the period of mourning is once accomplished, then all is 
said and done, and the survivors have no more duties towards 
their relatives who exist no longer. It is true that there is a 
way in which the dead continue to hold a place in the lives of 

* There are sometimes, as it seems, even funeral offerings. (See Roth, 
Superstition, Magic and Medzcine. in North Queensland Ethnog., Bulletin No. 5, 
§6g c, and Burial Customs, in tbtd.. No. 10, in Records of the Australian Museum, 
Vol. VI, No. 5, p. 395). But these olïerings are not periodical. 

64 Elementary Forms of Religious Life 

their kindred, even after the mourning is finished. It is some- 
times the case that their hair or certain of their bones are kept, 
because of special virtues which are attached to them.^ But 
by that time they have ceased to exist as persons, and have 
fallen to the rank of anonymous and impersonal charms. In 
this condition they are the object of no cult ; they serve only 
for magical purposes. 

However, there are certain Australian tribes which periodically 
celebrate rites in honour of fabulous ancestors whom tradition 
places at the beginning of time. These ceremonies generally 
consist in a sort of dramatic representation in which are 
rehearsed the deeds which the myths ascribe to these legendary 
heroes. 2 But the personages thus represented are not men who, 
after living the life of men, have been transformed into a sort 
of god by the fact of their death. They are considered to have 
exercised superhuman powers while alive. To them is attributed 
all that is grand in the history of the tribe, or even of the whole 
world. It is they who in a large measure made the earth such 
as it is, and men such as they are. The haloes with which they 
are still decorated do not come to them merely from the fact 
that they are ancestors, that is to say, in fine, that they are dead, 
but rather from the fact that a divine character is and always 
has been attributed to them ; to use the Melanesian expression, 
it is because they are constitutionally endowed with mana. 
Consequently, there is nothing in these rites which shows that 
death has the slightest power of deification. It cannot even be 
correctly said of certain rites that they form an ancestor-cult, 
since they are not addressed to ancestors as such. In order to 
have a real cult of the dead, it is necessary that after death 
real ancestors, the relations whom men really lose every day, 
become the object of the cult ; let us repeat it once more, there 
are no traces of any such cult in Australia. 

Thus the cult which, according to this hypothesis, ought to 
be the predominating one in inferior societies, is really non- 
existent there. In reality, the Australian is not concerned with 
his dead, except at the moment of their decease and during the 
time which immediately follows. Yet these same peoples, as 
we shall see, have a very complex cult for sacred beings of a 
wholly different nature, which is made up of numerous cere- 
monies and frequently occupjdng weeks or even entire months. 
It cannot be admitted that the few rites which the Australian 
performs when he happens to lose one of his relatives were the 
origin of these permanent cults which return regularly every 

^ Spencer and Gillen, Nat. Tr., pp. 538, 553, and Nor. Tr., pp. 463, 543, 547. 
* See especially, Spencer and GiUen, Northern Tribes, ch. vi, vii. ix. 

Leading Conceptions of the Elementary Religion 65 

year and which take up a considerable part of his existence. 
The contrast between the two is so great that we may even ask 
whether the first were not rather derived from the second, and 
if the souls of men, far from having been the model upon which 
the gods were originally imagined, have not rather been con- 
ceived from the very first as emanations from the divinity. 


From the moment that the cult of the dead is shown not to 
be primitive, animism lacks a basis. It would then seem useless 
to discuss the third thesis of the system, which concerns the 
transformation of the cult of the dead into the cult of nature. 
But since the postulate upon which it rests is also found in 
certain historians of religion who do not admit the animism 
properly so-called, such as Brinton,^ Lang,^ Ré ville, ^ and even 
Robertson Smith himself/ it is necessary to make an examination 
of it. 

This extension of the cult of the dead to all nature is said to 
come from the fact that we instinctively tend to represent all 
things in our own image, that is to say, as living and thinking 
beings. We have seen- that Spencer has already contested the 
reality of this so-called instinct. Since animals clerrly dis- 
tinguish living bodies from dead ones, it seemed to him impossible 
that man, the heir of the animals, should not have had this same 
faculty of discernment from the very first. But howsoever 
certain the facts cited by Spencer may be, they have not the 
demonstrative value which he attributes to them. His reasoning 
supposes that all the faculties, instincts and aptitudes of the 
animal have passed integrally into man ; now many errors 
have their origin in this principle which is wrongfully taken as 
a proven truth. For example, since sexual jealousy is generally 
very strong among the higher animals, it has been concluded 
that it ought to be found among men with the same intensity 
from the very beginnings of history.^ But it is well known 
to-day that men can practise a sexual communism which would 
be impossible if this jealousy were not capable of attenuating 
itself and even of disappearing when necessary.* The fact is 

^ The Religions of Primitive Peoples, pp. 47 ff. 

* Myth, Ritual and Religions, p. 123. 

* Les Religions des peuples non civilisés, II, Conclusion. 

* The Religion of the Semites, 2 éd., pp. 126, 132. 

^ This is the reasoning of Westermarck (Origins of Human Marriage, p. 6). 

* By sexual communism we do not mean a state of promiscuity where man 
knows no matrimonial rules : we believe that such a state has never existed. 
But it has frequently happened that groups of men have been regularly united 
to one or several women. 

66 Elementary Forms of Religious Life 

that man is not merely an animal with certain additional quali- 
ties : he is something else. Human nature is the result of a sort 
of recasting of the animal nature, and in the course of the various 
complex operations which have brought about this recasting, 
there have been losses as well as gains. How many instincts 
have we not lost ? The reason for this is that men are not only 
in relations with the physical environment, but also with a 
social envirohment infinitely more extended, more stable and 
more active than the one whose influence animals undergo. 
To live, they must adapt themselves to this. Now in order to 
maintain itself, society frequently finds it necessary that we 
should see things from a certain angle and feel them in a certain 
way ; consequently it modifies the ideas which we would 
ordinarily make of them for ourselves and the sentiments to 
which we would be inclined if we listened only to our animal 
nature ; it alters them, even going so far as to put the contrary 
sentiments in their place. Does it not even go so far as to make 
us regard our own individual lives as something of little value, 
while for the animal this is the greatest of things ? ^ Then it is 
a vain enterprise to seek to infer the mental constitution of the 
primitive man from that of the higher animals. 

But if the objection of Spencer does not have the decisive 
value which its author gives it, it is equally true that the animist 
theory can draw no authority from the confusions which children 
seem to make. When we hear a child angrily apostrophize an 
object which he has hit against, we conclude that he thinks of 
it as a conscious being like himself ; but that is interpreting 
his words and acts very badly. In reality, he is quite a stranger 
to the very complicated reasoning attributed to him. If he 
lays the blame on the table which has hurt him, it is not because 
he supposes it animated and intelligent, but because it has hurt 
him. His anger, once aroused by the pain, must overflow ; 
so it looks for something upon which to discharge itself, and 
naturally turns toward the thing which has provoked it, even 
though this has no effect. The action of an adult in similar 
circumstances is often as slightly reasonable. When we are 
violently irritated, we feel the need of inveighing, of destroying, 
though we attribute no conscious ill-will to the objects upon 
which we vent our anger. There is even so little confusion that 
when the emotion of a child is calmed, he can very well dis- 
tinguish a chair from a person : he does not act in at all the 
same way towards the two. It is a similar reason which explains 
his tendency to treat his playthings as if they were living beings. 
It is his extremely intense need of playing which thus finds a 

^ See our Suicide, pp. 233 ff. 

Leading Conceptions of the Elementary Religion 67 

means of expressing itself, just as in the other case the violent 
sentiments caused by pain created an object out of nothing. 
In order that he may consciously play with his jumping- jack, 
he imagines it a living person. This illusion is the easier for him 
because imagination is his sovereign mistress ; he thinks almost 
entirely with images, and we know how pliant images are, 
bending themselves with docility before every exigency of the 
will. But he is so little deceived by his own fiction that he would 
be the first to be surprised if it suddenly became a reality, and 
his toy bit him '.^ 

Let us therefore leave these doubtful analogies to one side. 
To find out if men were primitively inclined to the confusions 
imputed to them, we should not study animals or children of 
to-day, but the primitive beliefs themselves. If the spirits and 
gods of nature were really formed in the image of the human 
soul, they should bear traces of their origin and bring to mind 
the essential traits of their model. The most important char- 
acteristic of the soul is that it is conceived as the internal 
principle which animates the organism : it is that which moves 
it and makes it live, to such an extent that when it withdraws 
itself, life ceases or is suspended. It has its natural residence 
in the body, at least while this exists. But it is not thus with 
the spirits assigned to the different things in nature. The god 
of the sun is not necessarily in the sun, nor is the spirit of a 
certain rock in the rock which is its principal place of habitation. 
A spirit undoubtedly has close relations with the body to which 
it is attached, but one employs a very inexact expression when 
he says that it is its soul. As Codrington says,^ " there does not 
appear to be anywhere in Melanesia a belief in a spirit which 
animates any natural object, a tree, waterfall, storm or rock, 
so as to be to it what the soul is believed to be to the body of 
man. Europeans, it is true, speak of the spirits of the sea or, of 
the storm or of the forest ; but the native idea which they 
represent is that ghosts haunt the sea and the forest, having 
power to raise storms and strike a traveller with disease." While 
the soul is essentially within the body, the spirit passes the 
major portion of its time outside the object which serves as its 
base. This is one difference which does not seem to show that 
the second idea was derived from the first. 

From another point of view, it must be added that if men 
were really forced to project their own image into things, then 
the first sacred beings ought to have been conceived in their like- 
ness. Now anthropomorphism, far from being primitive, is 

^ Spencer, Principles oj Sociology, I, pp. 129 f. 
- Tke Melanesians, p. 123. 

68 Elementary Forms of Religious Life 

rather the mark of a relatively advinced civilization. In the 
beginning, sacred beings are concei"' cd in the form of an animal 
or vegetable, from which the human form is only slowly dis- 
engaged. It will be seen below that in Australia, it is animals 
and plants which are the first sacred beings. Even among the 
Indians of North America, the great cosmic divinities, which 
commence to be the object of a cult there, are very frequently 
represented in animal forms. ^ " The difference between the 
animal, man and the divine being," says Réville, not without 
surprise, " is not felt in this state of mind, and generally it might 
be said that it is the animal form which is the fundamental one."^ 
To find a god made up entirely of human elements, it is necessary 
to advance nearly to Christianity. Here, God is a man, not only 
in the physical aspect in which he is temporarily made manifest, 
but also in the ideas and sentiments which he expresses. But 
even in Greece and Rome, though the gods were generally 
represented with human traits, many mythical personages 
still had traces of an animal origin : thus there is Dionysus, 
who is often met with in the form of a bull, or at least with the 
horns of a bull ; there is Demeter, who is often represented with 
a horse's mane, there are Pan and Silenus, there are the Fauns, 
etc.^ It is not at all true that man has had such an inclination 
to impose his own form upon things. More than that, he even 
commenced by conceiving of himself as participating closely in 
the animal nature. In fact, it is a belief almost universal in 
Australia, and very widespread among the Indians of North 
America, that the ancestors of men were beasts or plants, or at 
least that the first men had, either in whole or in part, the 
distinctive characters of certain animal or vegetable species. 
Thus, far from seeing beings like themselves everywhere, men 
commenced by believing themselves to be in the image of some 
beings from which they differed radically. 


Finally, the animistic theory implies a consequence which is 
perhaps its best refutation. 

If it were true, it would be necessary to admit that religious 
beliefs are so many hallucinatory representations, without any 
objective foundation whatsoever. It is supposed that they are 
all derived from the idea of the soul because one sees only a 

* Dorsey, A Study of Siouan Cults, in Xltk Annual Report of the Bureau of 
Amer. Ethnology, pp. 431 ff., and passim. 

* La religion des peuples non civilisés, I, p. 248. 

3 V. W. de Visser, De Graecorum dits non referentibus speciem humanam. 
Cf. P. Perdrizet, Bulletin de correspondance hellénique, 1899, p. 035. 

Leading Conceptions of the Elementary Religion 69 

magnified soul in the spirits and gods. But according to Tylor 
and his disciples, the idea of the soul is itself constructed entirely 
out of the vague and inconsistent images which occupy our 
attention during sleep : for the soul is the double, and the double 
is merely a man as he appears to himself while he sleeps. From 
this point of view, then, sacred beings are only the imaginary 
conceptions which men have produced during a sort of delirium 
which regularly overtakes them every day, though it is quite 
impossible to see to what useful ends these conceptions serve, 
nor what they answer to in reality. If a man prays, if he makes 
sacrifices and offerings, if he submits to the multiple privations 
which the ritual prescribes, it is because a sort of constitutional 
eccentricity has made him take his dreams for perceptions, 
death for a prolonged sleep, and dead bodies for living and 
thinking beings. Thus not only is it true, as many have held, 
that the forms under which religious powers have been repre- 
sented to the mind do not express them exactly, and that 
the symbols with the aid of which they have been thought of 
partially hide their real nature, but more than that, behind 
these images and figures there exists nothing but the nightmares 
of primitive minds. In fine, religion is nothing but a dream, 
systematized and lived, but without any foundation in reality.^ 
Thence it comes about that the theorists of animism, when looking 
for the origins of religious thought, content themselves with a 
small outlay of energy. When they think that they have ex- 
plained how men have been induced to imagine beings of a 
strange, vaporous form, such as those they see in their dreams, 
they think the problem is resolved. 

In reality, it is not even approached. It is inadmissible that 
systems of ideas like religions, which have held so considerable 
a place in history, and to which, in all times, men have come 
to receive the energy which they must have to live, should be 
made up of a tissue of illusions. To-day we are beginning to 

* However, according to Spencer, there is a germ of truth in the belief in 
spirits : this is the idea that " the power which manifests itself inside the 
consciousness is a different form of power from that manifested outside the 
consciousness " {Ecclesiastical Institutions, § 659). Spencer understands by this 
that the notion of force in general is the sentiment of the force which we have 
extended to the entire universe ; this is what animism admits implicitly when it 
peoples nature with spirits analogous to our own. But even if this hypothesis in 
regard to the way in which the idea of force is formed were true — and it requires 
important reservations which we shall make (Bk. Ill, ch. iii, § 3) — it has nothing 
religious about it ; it belongs to no cult. It thus remains that the system of 
religious symbols and rites, the classification of things into sacred and profane, 
all that which is really religious in religion, corresponds to nothing in reality. 
Also, this germ of truth, of which he speaks, is still more a germ of error, for if it 
be true that the forces of nature and those of the mind are related, they are 
profoundly distinct, and one exposes himself to grave misconceptions in identify- 
ing them. 

70 Elementary Forms of Religious Life 

realize that law, morals and even scientific thought itself were 
bom of religion, were for a long time confounded with it, and 
have remained penetrated with its spirit. How could a vain 
fantasy have been able to fashion the human consciousness so 
strongly and so durably ? Surely it ought to be a principle of 
the science of religions that religion expresses nothing which 
does not exist in nature ; for there are sciences only of natural 
phenomena. The only question is to learn from what part of 
nature these realities come and what has been able to make 
men represent them under this singular form which is peculiar 
to religious thought. But if this question is to be raised, it is 
necessary to commence by admitting that they are real things 
which are thus represented. When the philosophers of the 
eighteenth century made religion a vast error imagined by the 
priests, they could at least explain its persistence by the interest 
which the sacerdotal class had in deceiving the people. But if 
the people themselves have been the artisans of these systems 
of erroneous ideas at the same time that they were its dupes, 
how has this extraordinary dupery been able to perpetuate 
itself all through the course of history ? 

One might even demand if under these conditions the words 
of science of religions can be employed without impropriety. 
A science is a discipline which, in whatever manner it is 
conceived, is always applied to some real data. Physics and 
chemistry are sciences because physico-chemical phenomena 
are real, and of a reality which does not depend upon the truths 
which these sciences show. There is a psychological science 
because there are really consciousnesses which do not hold their 
right of existence from the psychologist. But on the contrary, 
religion could not survive the animistic theory and the day 
when its truth was recognized by men, for they could not fail 
to renounce the errors whose nature and origin would thus be 
revealed to them. What sort of a science is it whose principal 
discovery is that the subject of which it treats does not exist ? 





II. — Naiurism 

THE spirit of the naturistic school is quite different. 
In the first place, it is recruited in a different environ- 
ment. The animists are, for the most part, ethnologists or anthro- 
pologists. The religions which they have studied are the crudest 
which humanity has ever known. Hence comes the extra- 
ordinary importance which they attribute to the souls of the 
dead, to spirits and to demons, and, in fact, to all spiritual 
beings of the second order : it is because these religions know 
hardly any of a higher order. ^ On the contrary, the theories 
which we are now going to describe are the work of scholars who 
have concerned themselves especially with the great civilizations 
of Europe and Asia. 

Ever since the work of the Grimm brothers, who pointed out 
the interest that there is in comparing the different mythologies 
of the Indo-European peoples, scholars have been struck by the 
remarkable similarities which these present. Mythical personages 
were identified who, though having different names, symbolized 
the same ideas and fulfilled the same functions ; even the names 
were frequently related, and it has been thought possible to 
establish the fact that they are not unconnected with one another. 
Such resemblances seemed to be explicable only by a common 
origin. Thus they were led to suppose that these conceptions, 
so varied in appearance, really came from one common source, 
of which they were only diversified forms, and which it was not 
impossible to discover. By the comparative method, they 
believed one should be able to go back, beyond these great 
religions, to a much more ancient system of ideas, and to the 
really primitive religion, from which the others were derived. 

The discovery of the Vedas aided greatly in stimulating these 
ambitions. In the Vedas, scholars had a written text, whose 
antiquity was undoubtedly exaggerated at the moment of its 

^ This is undoubtedly what explains the sympathy which folk-lorists like 
Mannhardt have felt for animistic ideas. In popular religions as in inferior 
religions, these spiritual beings of a second order hold the first place. 


72 Elementary Forms of Religious Life 

discovery, but which is surely one of the most ancient which we 
have at our disposition in an Indo-European language. Here 
they were enabled to study, by the ordinary methods of philology, 
a literature as old as or older than Homer, and a religion which 
was believed more primitive than that of the ancient Germans. 
A document of such value was evidently destined to throw a new 
light upon the religious beginnings of humanity, and the science 
of religions could not fail to be revolutionized by it. 

The conception which was thus born was so fully demanded by 
the state of the science and by the general march of ideas, that it 
appeared almost simultaneously in two different lands. In 1856, 
Max Millier exposed its principles in his Oxford Essays.^ Three 
years later appeared the work of Adalbert Kuhn on The Origin 
of Fire and the Drink of the Gods,^ which was clearly inspired by 
the same spirit. When once set forth, the idea spread very 
rapidly in scientific circles. To the name of Kuhn is closely 
associated that of his brother-in-law Schwartz, whose work on 
The Origin of Mythology,^ followed closely upon the preceding 
one. Steinthal and the whole German school of Volkerpsychologie 
attached themselves to the same movement. The theory was 
introduced into France in 1863 by M. Michel Bréal.* It met so 
little resistance that, according to an expression of Gruppe,^ " a 
time came when, aside from certain classical philologists, to 
whom Vedic studies were unknown, all the mythologists had 
adopted the principles of Max Miiller or Kuhn as their point of 
departure." * It is therefore important to see what they really 
are, and what they are worth. 

Since no one has presented them in a more systematic form 
than Max Miiller, it is upon his work that we shall base the 
description which follows.' 

^ In the essay entitled Comparative Mythology (pp. 47 ff). 

* Herabkunft des Feuers und Gôttertranks, Berlin, 1859 (a new edition was 
given by Ernst Kuhn in 1 886) . Cf. Der Schuss des Wilden Jàgers auf den Sonnen- 
hirsck, Zeitschrift f. d. Phil., I, 1869, pp. 89-169. Entwickelungsstufen des 
Mythus. Abhandl. d. Berl. Akad., 1873. 

' Der TJr sprung der Mythologie, Berlin, i860. 

* In his book Hercule et Cocus. Étude de mythologie comparée. Max Miiller's 
Comparative Mythology is there signalized as a work " which marks a new epoch 
in the history of Mythology " (p. 12). 

' Die Griechischen Kulte und Mythen, I, p. 78. 

* Among others who have adopted this conception may be cited Renan. See 
his Nouvelles études d'histoire religieuse, 1884, p. 31. 

' Aside from the Comparative Mythology, the works where Max Miiller has 
exposed his general theories on religion are : Hibbert Lectures (1878) under the 
title The Origin and Development of Beligion ; Natural Religion (1889) ; Physical 
Religion (1890) ; Anthropological Religion (1892) ; Theosophy, or Psychological 
Religion (1893) ; Contributions to the Science of Mythology (1897). Since his 
mythological theories are closely related to his philosophy of language, these 
works should be consulted in connection with the ones consecrated to language or 
logic, especially Lectures on the Science of Language, and The Science of Thought. 

Leading Conceptions of the Elementary Religion 73 

We have seen that the postulate at the basis of animism is that 
rehgion, at least in its origin, expresses no physical reality. But 
Max Millier commences with the contrary principle. For him, 
it is an axiom that religion reposes upon an experience, from 
which it draws all its authority. " Religion," he says, " if it is 
to hold its place as a legitimate element of our consciousness, 
must, like all other knowledge, begin with sensuous experience." ^ 
Taking up the old empirical adage, " Nihil est in intellectu quod 
noil ante fiierit in sensu," he applies it to religion and declares 
that there can be nothing in beliefs which was not first perceived. 
So here is a doctrine which seems to escape the grave objection 
which we raised against animism. From this point of view, it 
seems that religion ought to appear, not as a sort of vague and 
confused dreaming, but as a system of ideas and practices well 
founded in reality. 

But which are these sensations which give birth io religious 
thought ? That is the question which the study of the Vedas is 
supposed to aid in resolving. 

The names of the gods are generally either common words, still 
employed, or else words formerly common, whose original sense 
it is possible to discover. Now both designate the principal 
phenomena of nature. Thus A gni, the name of one of the principal 
divinities of India, originally signified only the material fact of 
fire, such as it is ordinarily perceived by the senses and without 
any mythological addition. Even in the Vedas, it is still employed 
with this meaning ; in any case, it is well shown that this significa- 
tion was primitive by the fact that it is conserved in other Indo- 
European languages : the Latin ignis, the Lithuanian ugnis, the 
old Slav ogny are evidently closely related to Agni. Similarly, 
the relationship of the Sanskrit Dyans, the Greek Zeus, the Latin 
Jovis and the Zio of High German is to-day uncontested. This 
proves that these different words designate one single and the 
same divinity, whom the different Indo-European peoples recog- 
nized as such before their separation. Now Dyaus signifies the 
bright sky. These and other similar facts tend to show that 
among these peoples the forms and forces of nature were the 
first objects to which the religious sentiment attached itself : 
they were the first things to be deified. Going one step farther 
in his generalization. Max Miiller thought that he was prepared 
to conclude that the religious evolution of humanity in general 
had the same point of departure. 

^ Natural Religion, p. 114. 

74 Elementary Forms of Religious Life 

It is almost entirely by considerations of a psychological sort 
that he justifies these inferences. The varied spectacles which 
nature offers man seemed to him to fulfil all the conditions 
necessary for arousing religious ideas in the mind directly. 
In fact, he says, " at first sight, nothing seemed less natural than 
nature. Nature was the greatest surprise, a terror, a marvel, a 
standing miracle, and it was only on account of their permanence, 
constancy, and regular recurrence that certain features of that 
standing miracle were called natural, in the sense of foreseen, 
common, intelligible. ... It was that vast domain of surprise, of 
terror, of marvel, of miracle, the unknown, as distinguished from 
the known, or, as I like to express it, the infinite, as distinct from 
the finite, which supplied from the earliest times the impulse to 
religious thought and language." ^ In order to illustrate his idea, 
he applies it to a natural force which holds a rather large place 
in the Vedic religion, fire. He says, " if you can for a moment 
transfer yourselves to that early stage of life to which we must 
refer not only the origin, but likewise the early phases of Physical 
Religion, you can easily understand what an impression the first 
appearance of fire must have made on the human mind. Fire 
was not given as something permanent or eternal, like the sky, 
or the earth, or the water. In whatever way it first appeared, 
whether through lightning or through the friction of the branches 
of trees, or through the sparks of flints, it came and went, it had 
to be guarded, it brought destruction, but at the same time, it 
made life possible in winter, it served as a protection during the 
night, it became a weapon of defence and offence, and last, not 
least, it changed man from a devourer of raw flesh into an eater 
of cooked meat. At a later time it became the means of working 
metal, of making tools and weapons, it became' an indispensable 
factor in all mechanical and artistic progress, and has remained 
so ever since. What should we be without fire even now ? " ^ 
The same author says in another work that a man could not enter 
into relations with nature without taking account of its immensity, 
of its infiniteness. It surpasses him in every way. Beyond the 
distances which he perceives, there are others which extend 
without limits ; each moment of time is preceded and followed 
by a time to which no limit can be assigned ; the flowing river 
manifests an infinite force, since nothing can exhaust it.^ There 
is no aspect of nature which is not fitted to awaken within us 
this overwhelming sensation of an infinity which surrounds us and 
dominates us.* It is from this sensation that religions are derived.^ 

^ Physical Religion, pp. iig-120. * Ibid., p. I2i ; cf. p. 304. 

' Natural Religion, pp. 121 fï., and 149-155. 

* " The overwhelming pressure of the infinite " {ibid., p. 138). 

' Ibid., pp. 195-196. 

Leading Conceptions of the Elementary Religion 75 

However, they are there only in germ.i ReUgion really 
commences only at the moment when these natural forces are no 
longer represented in the mind in an abstract form. They must 
be transformed into personal agents, living and thinking beings, 
spiritual powers or gods ; for it is to beings of this sort that the 
cult is generally addressed. We have seen that animism itself 
has been obliged to raise this question, and also how it has 
answered it : man seems to have a sort of native incapacity for 
distinguishing the animate from the inanimate and an irresistible 
tendency to conceive the second under the form of the first. 
Max Muller rejects any such solution. ^ According to him it is 
language which has brought about this metamorphosis, by the 
action which it exercises upon thought. 

It is easily explained how men, being perplexed by the mar- 
vellous forces upon which they feel that they depend, have been 
led to reflect upon them, and how they have asked themselves 
what these forces are and have made an effort to substitute for 
the obscure sensation which they primitively had of them, a 
clearer idea and a better defined concept. But as our author very 
justly says,^ this idea and concept are impossible without the 
word. Language is not merely the external covering of a thought ; 
it also is its internal framework. It does not confine itself to 
expressing this thought after it has once been formed ; it also 
aids in making it. However, its nature is of a different sort, so 
its laws are not those of thought. Then since it contributes to 
the elaboration of this latter, it cannot fail to do it violence to 
some extent, and to deform it. It is a deformation of this sort 
which is said to have created the special characteristic of religious 

Thinking consists in arranging our ideas, and consequently In 
classifying them. To think of fire, for example, is to put it into 
a certain category of things, in such a way as to be able to say 
that it is this or that, or this and not that. But classifying is also 
naming, for a general idea has no existence and reality except in 
and by the word which expresses it and which alone makes its 
individuality. Thus the language of a people always has an 
influence upon the manner in which new things, recently learned, 
are classified in the mind and are subsequently thought of ; these 
new things are thus forced to adapt themselves to pre-existing 
forms. For this reason, the language which men spoke when they 

* Max Mijller even goes so far as to say that until thought has passed this 
first stage, it has very few of the characteristics which we now attribute to 
religion (Physic. Rel., p. 120). 

* Physic. Rel., p. 12S. 

' The Science 0/ Thought, p. 30. 

76 Elementary Forms of Religious Life 

undertook to construct an elaborated representation of the 
universe marked the system of ideas which was then born with 
an indehble trace. 

Nor are we without some knowledge of this language, at least 
in so far as the Indo-European peoples are concerned. Howsoever 
distant it may be from us, souvenirs of it remain in our actual 
languages which permit us to imagine what it was : these are the 
roots. These stems, from which are derived all the words which we 
employ and which are found at the basis of all the Indo-European 
languages, are regarded by Max Miiller as so many echoes of the 
language which the corresponding peoples spoke before their 
separation, that is to say, at the very moment when this religion 
of nature, which is to be explained, was being formed. Now these 
roots present two remarkable characteristics, which, it is true, 
have as yet been observed only in this particular group of 
languages, but which our author believes to be present equally 
in the other linguistic families. ^ 

In the first place, the roots are general ; that is to say that 
they do not express particular things and individuals, but types, 
and even types of an extreme generality. They represent the 
most general themes of thought ; one finds there, as though 
fixed and crystallized, those fundamental categories of the in- 
tellect which at every moment in history dominate the entire 
mental life, the arrangement of which philosophers have many 
times attempted to reconstruct. ^ 

Secondly, the types to which they correspond are types of 
action, and not of objects. They translate the most general 
manners of acting which are to be observed among living beings 
and especially among men ; they are such actions as striking, 
pushing, rubbing, l3àng down, getting up, pressing, mounting, 
descending, walking, etc. In other words, men generalized and 
named their principal ways of acting before generalizing and 
naming the phenomena of nature.^ 

Owing to their extreme generality, these words could easily 
be extended to all sorts of objects which they did not originally 
include ; it is even this extreme suppleness which has permitted 
them to give birth to the numerous words which are derived 
from them. Then when men, turning towards things, undertook 
to name them, that they might be able to think about them, 
they applied these words to them, though they were in no way 
designed for them. But, owing to their origin, these were able to 
designate the forces of nature only by means of their manifestations 

^ Natural Religion, pp. 393 ff. 

* Physic. Rel., p. 133 ; The Science of Thought, p. 219 ; Lectures on the Science 
of Language, II, pp. i fi. 

' The Science of Thought, p. 272. 

Leading Conceptions of the Elementary Religion yy 

which seemed the nearest to human actions : a thunderbolt 
was called something that tears up the soil or that spreads 
fire ; the wind, something that sighs or whistles ; the sun, some- 
thing that throws golden arrows across space ; a river, something 
that flows, etc. But since natural phenomena were thus compared 
to human acts, this something to which they were attached was 
necessarily conceived under the form of personal agents, more or 
less like men. It was only a metaphor, but it was taken literally ; 
the error was inevitable, for science, which alone could dispel the 
illusion, did not yet exist. In a word, since language was made 
of human elements, translating human states, it could not be 
applied to nature without transforming it.^ Even to-day, re- 
marks M. Bréal, it forces us in a certain measure to represent 
things from this angle. " We do not express an idea, even one 
designating a simple quality, without giving it a gender, that is 
to say, a sex ; we cannot speak of an object, even though it be 
considered in a most general fashion, without determining it by 
an article ; every subject of a sentence is presented as an active 
being, every idea as an action, and every action, be it transitory 
or permanent, is limited in its duration by the tense in which 
we put the verb." ^ Our scientific training enables us to rectify 
the errors which language might thus suggest to us ; but the 
influence of the word ought to be all-powerful when it has no 
check. Language thus superimposes upon the material world, 
such as it is revealed to our senses, a new world, composed wholly 
of spiritual beings which it has created out of nothing and which 
have been considered as the causes determining physical phe- 
nomena ever since. 

But its action does not stop there. When words were once 
forged to represent these personalities which the popular imagina- 
tion had placed behind things, a reaction affected these words 
themselves : they raised all sorts of questions, and it was to 
resolve these problems that myths were invented. It happened 
that one object received a plurality of names, corresponding to 
the plurality of aspects under which it was presented in ex- 
perience ; thus there are more than twenty words in the Vedas 
for the sky. Since these words were different, it was believed 
that they corresponded to so many distinct personalities. But 
at the same time, it was strongly felt that these same personalities 
had an air of relationship. To account for that, it was imagined 
that they formed a single family ; genealogies, a civil condition 
and a history were invented for them. In other cases, different 
things were designated by the same term : to explain these 

* The Scieyice of Thought, I, p. 327 ; Physic. Rel., pp. 125 fi. 

* Mélanges de mythologie et de linguistique, p. S. 

78 Elementary Forms of Religious Life 

homonyms, it was believed that the corresponding things were 
transformations of each other, and new fictions were invented 
to make these metamorphoses intelhgible. Or again, a word 
which had ceased to be understood, was the origin of fables 
designed to give it a meaning. The creative work of language 
continued then, making constructions ever more and more 
complex, and then mythology came to endow each god with a 
biography, ever more and more extended and complete, the 
result of all of which was that the divine personalities, at first 
confounded with things, finally distinguished and determined 

This is how the notion of the divine is said to have been con- 
structed. As for the religion of ancestors, it was only a reflection 
of this other. 1 The idea of the soul is said to have been first 
formed for reasons somewhat analogous to those given by Tylor, 
except that according to Max Miiller, they were designed to 
account for death, rather than for dreams. ^ Then, under the 
influence of diverse, partially accidental, circumstances,^ the 
souls of men, being once disengaged from the body, were drawn 
little by little within the circle of divine beings, and were thus 
finally deified themselves. But this new cult was the product of 
only a secondary formation. This is proven by the fact that 
deified men have generally been imperfect gods or demi-gods, 
whom the people have always been able to distinguish from the 
genuine deities.^ 


This doctrine rests, in part, upon a certain number of linguistic 
postulates which have been and still aF« very much questioned. 
Some have contested the reality of many of the similarities 
which Max Miiller claimed to have found between the names of 
,the gods in the various European languages. The interpretation 
which he gave them has been especially doubted : it has been 
asked if these names, far from being the mark of a very primitive 

^ Anthropological Religion, pp. 128-130. 

* This explanation is not as good as that of Tylor. According to Max Miiller, 
men could not admit that life stopped with death ; therefore they concluded 
that there were two beings within them, one of which survived the body. But it 
is hard to see what made them think that life continued after the body was 

* For the details, see Anthrop. Rel., pp. 351 £f. 

* Anthrop. Rel., p. 130. — This is what keeps Max Miiller from considering 
Christianity the climax of all this development. The religion of ancestors, he 
says, supposes that there is something divine in man. Now is that idea not the 
one at the basis of the teaching of Christ ? {ibid., pp. 378 fï.). It is useless to 
insist upon the strangeness of the conception which makes Christianity the latest 
of the cults of the dead. 

Leading Conceptions of the Elementary Religion 79 

religion, are not the slow product, either of direct borrowings or 
of natural intercourse with others. ^ Also, it is no longer admitted 
that the roots once existed in an isolated state as autonomous 
realities, nor that they allow us to reconstruct, even hypotheti- 
cally, the original language of the Indo-Europeans.^ Finally, 
recent researches would tend to show that the Vedic divinities 
did not all have the exclusively naturistic character attributed 
to them by Max Miiller and his school. ^ But we shall leave aside 
those questions, the discussion of which requires a special com- 
petence as a philologist, and address ourselves directly to the 
general principles of the system. It will be important here not 
to confound the naturistic theory with these controverted postu- 
lates ; for this is held by numbers of scholars who do not make 
language play the predominating rôle attributed to it by Max 

That men have an interest in knowing the world which sur- \ 
rounds them, and consequently that their reflection should have 
been applied to it at an early date, is something that everyone 
will readily admit. Co-operation with the things with which 
they were in immediate connection was so necessary for them 
that they could not fail to seek a knowledge of their nature. But 
if, as naturism pretends, it is of these reflections that religious 
thought was bom, it is impossible to explain how it was able to 
survive the first attempts made, and the persistence with which 
it has maintained itself becomes unintelligible. If we have need of 
knowing the nature of things, it is in order to act upon them in 
an appropriate manner. But the conception of the universe 
given us by religion, especially in its early forms, is too greatly 
mutilated to lead to temporarily useful practices. Things become 
nothing less than living and thinking beings, minds or person- 
alities like those which the religious imagination has made into 
the agents of cosmic phenomena. It is not by conceiving of 
them under this form or by treating them according to this 
conception that men could make them work for their ends. It is 
not by addressing prayers to them, by celebrating them in feasts 
and sacrifices, or by imposing upon themselves fasts and priva- 
tions, that men can deter them from working harm or oblige 
them to serve their own designs. Such processes could succeed 
only very exceptionally and, so to speak, miraculously. If, then, 
religion's reason for existence was to give us a conception of the 

^ See the discussion of the hypothesis in Gruppe, Griechishen Kulte und 
My then, pp. 79-184. 

* See Meillet, Introduction à l'étude comparative des langues indo-européennes, 
p. 119. 

' Uldenberg, Die Religion des Vedas, pp. 59 ff. ; Meillet, Le dieu Iranien 
Mythra, in Journal Asiatique, X, No. i, July-August, 1907, pp. 143 ff. 

8o Elementary Forms of Religious Life 

world which would guide us in our relations with it, it was in 
no condition to fulfil its function, and people would not have 
been slow to perceive it : failures, being infinitely more frequent 
than successes, would have quickly shown them that they were 
following a false route, and religion, shaken at each instant by 

L these repeated contradictions, would not have been able to 

It is undeniably true that errors have been able to perpetuate 
themselves in history ; but, except under a union of very ex- 
ceptional circumstances, they can never perpetuate themselves 
thus unless they were true practically, that is to say, unless, 
without giving us a theoretically exact idea of the things with 
which they deal, they express well enough the manner in which 
they affect us, either for good or for bad. Under these circum- 
stances, the actions which they determine have every chance of 
being, at least in a general way, the very ones which are proper, 
so it is easily explained how they have been able to survive the 
proofs of experience.^ But an error and especially a system of 
errors which leads to, and can lead to nothing but mistaken and 
useless practices, has no chance of living. Now what is there in 
common between the rites with which the believer tries to act 
upon nature and the processes by which science has taught 
us to make use of it, and which we now know are the only 
efficacious ones ? If that is what men demanded of religion, it 
is impossible to see how it could have maintained itself, unless 
clever tricks had prevented their seeing that it did not give 
them what they expected from it. It would be necessary to 
return again to the over simple explanations of the eighteenth 
century. 2 

Thus it is only in appearance that naturism escapes the 

1 In this category are a large number of the maxims of popular wisdom. 

* It is true that this argument does not touch those who see in religion a code 
(especially of hygiene) whose provisions, though placed under the sanction of 
imaginary beings, are nevertheless well founded. But we shall not delay to 
discuss a conception so insupportable, and which has, in fact, never been sustained 
in a systematic manner by persons somewhat informed upon the history of 
religions. It is difficult to see what good the terrible practices of the initiation 
bring to the health which they threaten ; what good the dietetic restrictions, 
which generally deal with perfectly clean animals, have hygienically ; how 
sacrifices, which take place far from a house, make it more solid, etc. Undoubtedly 
there are religious precepts which at the same time have a practical utility ; but 
they are lost in the mass of others, and even the services which they render are 
frequently not without some drawbacks. If there is a religiously enforced 
cleanliness, there is also a religious filthiness which is derived from these same 
principles. The rule which orders a corpse to be carried away from the camp 
because it is the seat of a dreaded spirit is undoubtedly useful. But the same 
belief requires the relatives to anoint themselves with the liquids which issue 
from a corpse in putrefaction, because they are supposed to have exceptional 
virtues. — From this point of view, magic has served a great deal more than 

Leading Conceptions of the Elementary Religion 8i 

objection which we recently raised against animism. It also 
makes religion a system of hallucinations, since it reduces it to 
an immense metaphor with no objective value. It is true that it 
gives religion a point of departure in reality, to wit, in the sensa- 
tions which the phenomena of nature provoke in us ; but by the 
bewitching action of language, this sensation is soon transformed 
into extravagant conceptions. Religious thought does not come 
in contact with reality, except to cover it at once with a thick 
veil which conceals its real forms : this veil is the tissue of fabulous 
beliefs which mythology brought forth. Thus the believer, like 
the delirious man, lives in a world peopled with beings and things 
which have only a verbal existence. Max Millier himself recog- 
nized this, for he regarded myths as the product of a disease of 
the intellect. At first, he attributed them to a disease of language, 
but since language and the intellect are inseparable for him, what 
is true of the one is true of the other. " When trying to explain 
the inmost nature of mythology," he says, " I called it a disease 
of Language rather than of Thought. . . . After I had fully ex- 
plained in my Science of Thought that language and thought are 
inseparable, and that a disease of language is therefore the same 
thing as a disease of thought, no doubt ought to have remained 
as to what I meant. To represent the supreme God as committing 
every kind of crime, as being deceived by men, as being angry 
with his wife and violent with his children, is surely a proof of a 
disease, of an unusual condition of thought, or, to speak more 
clearly, of real madness." ^ And this argument is not valid 
merely against Max Miiller and his theory, but against the very 
principle of naturism, in whatever way it may be applied. What- 
ever we may do.lf religi3n has as its principal object the expres- 
sion of the forces of nature, it is impossible to see in it anything 
more than a system of lying fictions, whose survival is incom- 

Max Miiller thought he escaped this objection, whose gravity 
he felt, by distinguishing radically between mythology and 
religion, and by putting the first outside the second. He claims 
the right of reserving the name of religion for only those beliefs 
which conform to the prescriptions of a sane moral system 
and a rational theology. The myths were parasitic growths 
which, under the influence of language, attached themselves 
upon these fundamental conceptions, and denatured them. 
Thus the belief in Zeus was religious in so far as the Greeks 
considered him the supreme God, father of humanity, protector 
of laws, avenger of crimes, etc. ; but all that which concerned 

^ Contributions to the Science of Mythology, 1, pp. 68 f . 

82 Elementary Forms of Religious Life 

the biography of Zeus, his marriages and his adventures, was 
only mythology.^ 

But this distinction is arbitrary. It is true that mythology 
has an aesthetic interest as well as one for the history of religions ; 
but it is one of the essential elements of the religious life, never- 
theless. If the myth were withdrawn from religion, it would be 
necessary to withdraw the rite also ; for the rites are generally 
addressed to definite personalities who have a name, a character, 
determined attributes and a history, and they vary according to 
the manner in which these personalities are conceived. The cult ren- 
dered to a divinity depends upon the character attributed to him ; 
and it is the myth which determines this character. Very frequently, 
the rite is nothing more than the myth put in action ; the Christian 
communion is inseparable from the myth of the Last Supper, from 
which it derives all its meaning. Then if all mythology is the 
result of a sort of verbal delirium, the question which we raised 
remains intact : the existence, and especially the persistence of 
the cult become inexplicable. It is hard to understand how men 
have continued to do certain things for centuries without any 
object. Moreover, it is not merely the peculiar traits of the divine 
personalities which are determined by mythology ; the very idea 
that there are gods or spiritual beings set above the various 
departments of nature, in no matter what manner they may be 
represented, is essentially mythical. ^ Now if all that which apper- 
tains to the notion of gods conceived as cosmic agents is blotted 
out of the religions of the past, what remains ? The idea of a 
divinity in itself, of a transcendental power upon which man 
depends and upon which he supports himself ? But that is 
only an abstract and philosophic conception which has been fully 
realized in no historical religion ; it is without interest for the 
science of religions.^ We must therefore avoid distinguishing 
between religious beliefs, keeping some because they seem to us 

* Lectures on the Science of Language, II, p. 456 ff. ; Physic. Rel., pp. 276 ff. — 
Also Bréal, Mélanges, p. 6, " To bring the necessary clarity into this question 
of the origin of mythology, it is necessary to distinguish carefully the gods, which 
are the immediate product of the human intelligence, from the fables, which are 
its indirect and involuntary product." 

* Max MUller recognized this. See Physic. Rel., p. 132, and Comparative 
Mythology, p. 58. " The gods are nomina and not numina, names without being 
and not beings without name." 

* It is true that Max Miiller held that for the Greeks, " Zeus was, and remained, 
in spite of all mythological obscurations, the name of the Supreme Deity " 
{Science of Language, II, p. 478). We shall not dispute this assertion, though 
it is historically contestable ; but in any case, this conception of Zeus could never 
have been more than a glimmer in the midst of all the other religious beliefs of 
the Greeks. 

Besides this, in a later work. Max Miiller went so far as to make even the 
notion of god in general the product of a wholly verbal process and thus of 
a mythological elaboration {Physic. Rel., p. 138). 


Leading Conceptions of the Elementary Religion 83 

to be true and sane and rejecting others because they shock and 
disconcert us. All myths, even those which we find the most 
unreasonable, have been believed. ^ Men have believed in 
them no less firmly than in their own sensations ; they have 
based their conduct upon them. In spite of appearances, it 
is therefore impossible that they should be without objective 

However, it will be said that in whatever manner religions may 
be explained, it is certain that they are mistaken in regard to 
the real nature of things : science has proved it. The modes of 
action which they counsel or prescribe to men can therefore 
rarely have useful effects : it is not by lustrations that the 
sick are cured nor by sacrifices and chants that the crops are 
made to grow. Thus the objection which we have made to 
naturism would seem to be applicable to all possible systems of 

Nevertheless, there is one which escapes it. Let us suppose 
that religion responds to quite another need than that of adapting 
ourselves to sensible objects : then it will not risk being weakened 
by the fact that it does not satisfy, or only badly satisfies, this 
need. If religious faith was not born to put man in harmony 
with the material world, the injuries which it has been able to 
do him in his struggle with the world do not touch it at its 
source, because it is fed from another. 

If it is not for these reasons that a man comes to believe, 
he should continue to believe even when these reasons are 
contradicted by the facts. It is even conceivable that faith 
should be strong enough, not only to support these contradictions, 
but also even to deny them and to keep the believer from seeing 
their importance ; this is what succeeds in rendering them 
inoffensive for religion. When the religious sentiment is active, 
it will not admit that rehgion can be in the wrong, and it readily 
suggests explanations which make it appear innocent ; if the 
rite does not produce the desired results, this failure is imputed 
either to some fault of execution, or to the intervention of another, 
contrary deity. But for that, it is necessary that these re- 
ligious ideas have their source in another sentiment than that 
betrayed by these deceptions of experience, or else whence could 
come their force of resistance ? 

Ï Undoubtedly outside the real myths there were always fables which were 
not believed, or at least were not believed in the same way and to the same 
degree, and hence had no religious character. The line of demarcation between 
fables and myths is certainly floating and hard to determine. But this is no 
reason for making all myths stories, any more than we should dream of making 
all stories myths. There is at least one characteristic which in a number of cases 
suffices to differentiate the religious myth : that is its relation to the cult. 

84 Elementary Forms of Religious Life 


But more than that, even if men had really had reasons for 
remaining obstinate, in spite of all their mistakes, in expressing 
cosmic phenomena in religious terms, it is also necessary that 
these be of a nature to suggest such an interpretation. Now 
when could they have gotten such a property ? Here again we 
find ourselves in the presence of one of those postulates which 
pass as evident only because they have not been criticized. It 
is stated as an axiom that in the natural play of physical forces 
there is all that is needed to arouse within us the idea of the 
sacred ; but when we closely examine the proofs of this propo- 
sition, which, by the way, are sufficiently brief, we find that they 
reduce to a prejudice. 

They talk about the marvel which men should feel as they 
discover the world. But really, that which characterizes the life 
of nature is a regularity which approaches monotony. Every 
morning the sun mounts in the horizon, every evening it sets ; 
every pionth the moon goes through the same cycle ; the river 
flows in an uninterrupted manner in its bed ; the same seasons 
periodically bring back the same sensations. To be sure, here 
and there an unexpected event sometimes happens : the sun is 
eclipsed, the moon is hidden behind clouds, the river overflows. 
But these momentary variations could only give birth to equally 
momentary impressions, the remembrance of which is gone 
after a little while ; they could not serve as a basis for these 
stable and permanent systems of ideas and practices which 
constitute religions. Normally, the course of nature is uniform, 
and uniformity could never produce strong emotions. Repre- 
senting the savage as filled with admiration before these marvels 
transports much more recent sentiments to the beginnings of 
history. He is much too accustomed to it to be greatly surprised 
by it. It requires culture and reflection to shake oft this yoke 
of habit and to discover how marvellous this regularity itself is. 
Besides, as we have already remarked,^ admiring an object is 
not enough to make it appear sacred to us, that is to say, to mark 
it with those characteristics which make all direct contact with 
it appear a sacrilege and a profanation. We misunderstand what 
the religious sentiment really is, if we confound it with every 
impression of admiration and surprise. 

But, they say, even if it is not admiration, there is a certain 
impression which men cannot help feeling in the presence of 
nature. He cannot come in contact with it, without realizing 

* See above, p. 28. 

Leading Conceptions of the Elementary Religion 85 

that it is greater than he. It overwhelms him by its immensity. 
This sensation of an infinite space which surrounds him, of an 
infinite time which has preceded and will follow the present 
moment, and of forces infinitely superior to those of which he is 
master, cannot fail, as it seems, to awaken within him the idea 
that outside of him there exists an infinite power upon which he 
depends. And this idea enters as an essential element into our 
conception of the divine. — . 

But let us bear in mind what the question is. We are trying to 1 
find out how men came to think that there are in reality two 
categories of things, radically heterogeneous and incomparable 
to each other. Now how could the spectacle of nature give rise 
to the idea of this duality ? Nature is always and everywhere 
of the same sort. It matters little that it extends to infinity : 
beyond the extreme limit to which my eyes can reach, it is not 
different from what it is here. The space which I imagine beyond 
the horizon is still space, identical with that which I see. The 
time which flows without end is made up of moments identical 
with those which I have passed through. Extension, like dura- 
tion, repeats itself indefinitely ; if the portions which I touch 
have of themselves no sacred character, where did the others 
get theirs ? The fact that I do not see them directly, is not i 
enough to transform them.^ A world of profane things may~y 
well be unlimited ; but it remains a profane world. Do they 
say that the physical forces with which we come in contact 
exceed our own ? Sacred forces are not to be distinguished from 
profane ones simply by their greater intensity, they are different ; 
they have special qualities which the others do not have. Quite / 
on the contrary, all the forces manifested in the universe are oV 
the same nature, those that are within us just as those that are 
outside of us. And especially, there is no reason which could 
have allowed giving a sort of pre-eminent dignity to some in 
relation to others. Then if rehgion really was bom because of 
the need of assigning causes to physical phenomena, the forces 
thus imagined would have been no more sacred than those con- 
ceived by the scientist to-day to account for the same facts. 2 

* More than that, in the language of Max Miiller, there is a veritable abuse of 
words. Sensuous experience, he says, implies, at least in certain cases, " beyond 
the known, something unknown, something which I claim the liberty to call infinite " 
(Natural Rel., p. 195 ; of. p. 218). The unknown is not necessarily the infinite, 
any more than the infinite is necessarily the unknown if it is in all points-the 
same, and consequently like the part which we know. It would be necessary to 
prove that the part of it which we perceive differs in nature from that which we 
do not perceive. 

* Max Miiller involuntarily recognizes this in certain passages. He confesses 
that he sees little difference between Agni, the god of fire, and the notion of 
ether, by which the modem physicist explains light and heat (Phys. Rel., pp. 
126 f.). Also, he connects the notion of divinity to that of agency (p. 138) or 

86 Elementary Forms of Religious Life 

This is as much as to say that there would have been no sacred 
j_beings and therefore no rehgion. 

But even supposing that this sensation of being " overwhelmed " 
were really able to suggest religious ideas, it could not have 
produced this effect upon the primitive, for he does not have it. 
He is in no way conscious that cosmic forces are so superior to 
his own. Since science has not yet taught him modesty, he 
attributes to himself an empire over things which he really does 
not have, but the illusion of which is enough to prevent his feeling 
dominated by them. As we have already pointed out, he thinks 
that he can command the elements, release the winds, compel 
the rain to fall, or stop the sun, by a gesture, etc.^ Religion itself 
contributes to giving him this security, for he believes that it 
arms him with extended powers over nature. His rites are, in 
part, means destined to aid him in imposing his will upon the 
world. Thus, far from being due to the sentiment which men 
should have of their littleness before the universe, religions 
are rather inspired by the contrary sentiment. Even the most 
elevated and idealistic have the effect of reassuring men in their 
struggle with things : they teach that faith is, of itself, able 
" to move mountains," that is to say, to dominate the forces of 
nature. How could they give rise to this confidence if they had 
had their origin in a sensation of feebleness and impotency ? 

Finally, if the objects of nature really became sacred because 
of their imposing forms or the forces which they manifest, then 
the sun, the moon, the sky, the mountains, the sea, the winds, in 
a word, the great cosmic powers, should have been the first to be 
raised to this dignity ; for there are no others more fitted to 
appeal to the senses and the imagination. But as a matter of 
fact, they were divinized but slowly. The first beings to which 
the cult is addressed — the proof will be found in the chapters 
which follow — are humble vegetables and animals, in relation to 
which men could at least claim an equality : they are ducks, 
rabbits, kangaroos, lizards, worms, frogs, etc. Their objective 
qualities surely were not the origin of the religious sentiments 
which they inspired. 

of a causality which is not natural and profane. The fact that religion repre- 
sents the causes thus imagined, under the form of personal agents, is not enough 
to explain how they got a sacred character. A personal agent can be profane, and 
also, many religious forces are essentially impersonal. 

1 We shall see below, in speaking of the ef&cacy of rites and faith, how these 
illusions are to be explained (Bk. Ill, ch. ii). 



History of the Question. — Method of Treating it 

HOWSOEVER opposed their conclusions may seem to be, 
the two systems which we have just studied agree upon 
one essential point : they state the problem in identical terms. 
Both undertake to construct the idea of the divine out of the 
sensations aroused in us by certain natural phenomena, either 
physical or biological. For the animists it is dreams, for the 
naturists, certain cosmic phenomena, which served as the point 
of departure for religious evolution. But for both, it is in the 
nature, either of man or of the universe, that we must look for 
the germ of the grand opposition which separates the profane 
from the sacred. 

But such an enterprise is impossible : it supposes a veritable 
creation ex nihilo. A fact of common experience cannot give 
us the idea of something whose characteristic is to be outside 
the world of common experience. A man, as he appears to 
himself in his dreams, is only a man. Natural forces, as our 
senses perceive them, are only natural forces, howsoever great 
their intensity may be. Hence comes the common criticism 
which we address to both doctrines. In order to explain how 
these pretended data of religious thought have been able to 
take a sacred character which has no objective foundation, it 
would be necessary to admit that a whole world of delusive 
representations has superimposed itself upon the other, de- 
natured it to the point of making it unrecognizable, and sub- 
stituted a pure hallucination for reality. Here, it is the illusions 
of the dream which brought about this transfiguration ; there, 
it is the brilliant and vain company of images evoked by the 
word. But in one case as in the other, it is necessary to regard 
religion as the product of a delirious imagination. 

Thus one positive conclusion is arrived at as the result of this 
critical examination. Since neither man nor nature have of 
themselves a sacred character, they must get it from another 


88 Elementary Forms of Religious Life 

source. Aside from the human individual and the physical 
world, there should be some other reality, in relation to which 
this variety of delirium which all religion is in a sense, has a 
significance and an objective value. In other words, beyond 
those which we have called animistic and naturistic, there 
should be another sort of cult, more fundamental and more 
primitive, of which the first are only derived forms or particular 

In fact, this cult does exist : it is the one to which ethnologists 
have given the name of totemism>> 


It was only at the end of the eighteenth century that the 
word totem appeared in ethnographical literature. It is found 
for the first time in the book of an Indian interpreter, J. Long, 
which was published in London in 1791.^ For nearly a half a 
century, totemism was known only as something exclusively 
American. 2 It was only in 1841 that Grey, in a passage which 
has remained celebrated, ^ pointed out the existence of wholly 
similar practices in Australia. From that time on, scholars 
began to realize that they were in the presence of a system of a 
certain generality. 

But they saw there only an essentially archaic institution, 
an ethnographical curiosity, having no great interest for the 
historian. MacLennan was the first who undertook to attach 
totemism to the general history of humanity. In a series of 
articles in the Fortnightly Review,* he set himself to show that 
totemism was not only a religion, but one from which were 
derived a multitude of beliefs and practices which are found in 
much more advanced religious systems. He even went so far 
as to make it the source of all the animal-worshipping and plant- 
worshipping cults which are found among ancient peoples. 
Certainly this extension of totemism was abusive. The cults 
of animals and plants depend upon numerous causes which 
cannot be reduced to one, without the error of too great sim- 
plicity. But this error, by its very exaggerations, had at least 
the advantage, that it put into evidence the historical importance 
of totemism. 

Students of American totemism had already known for a 

^ Voyages and Travels of an Indian Interpreter. 

' This idea was so common that even M. Réville continued to make America 
the classic land of totemism (Religions des peuples non civilisés, I, p. 242). 

* Journals 0/ Two Expeditions in North-West and Western Australia, II, p. 228. 

* The Worship of Animals and Plants. Totems and Totemism (1869, 1870). 

Totemism as an Elementary Religion 89 

long time that .this form of religion was most intimately united 
to a determined social organization, that its basis is the division 
of the social group into clans. ^ In 1877, in his Ancient Society,^ 
Lewis H. Morgan undertook to make a study of it, to determine 
its distinctive characteristics, and at the same time to point 
out its generality among the Indian tribes of North and Central 
America. At nearly the same moment, and even following the 
direct suggestion of Morgan, Fison and Howitt ^ established the 
existence of the same social system in Australia, as well as its 
relations with totemism. 

Under the influence of these directing ideas, observations 
could be made with better method. The researches which the 
American Bureau of Ethnology undertook, played an important 
part in the advance of these studies.^ By 1887, the documents 
were sufficiently numerous and significant to make Frazer 
consider it time to unite them and present them to us in a 
systematic form. Such is the object of his little book Totemism,^ 
where the system is studied both as a religion and as a legal 
institution. But this study was purely descriptive ; no effort 
was made to explain totemism * or to understand its funda- 
mental notions. 

Robertson Smith is the first who undertook this work of 
elaboration. He realized more clearly than any of his prede- 
cessors how rich this crude and confused religion is in germs for 
the future. It is true that MacLennan had already connected 
it with the great religions of antiquity ; but that was merely 
because he thought he had found here and there the cult of 
animals or plants. Now if we reduce totemism to a sort of 
animal or plant worship, we have seen only its most superficial 
aspect : we have even misunderstood its real nature. Going 

* This idea is found already very clearly expressed in a study by Gallatin 
entitled Synopsis of the Indian Tribes (Archœologia Americana, II, pp. 109 £f.), 
and in a notice by Morgan in the Cambrian Journal, 18O0, p. 149. 

* This work had been prepared for and preceded by two others by the same 
author: The League of the Iroquois (1851), and Systems of Consanguinity and 
Affinity of the Human Family (1871). 

' Kamilaroi and Kurnai, 1880. 

* In the very first volumes of the Annual Report of the Bureau of American 
Ethnology are found the study of Powell, Wyandot Government (I, p. 59), that of 
Gushing, Zufii Fetiches (II, p. 9), Smith, Myths of the Iroquois {ibid., p. 77), 
and the important work of Dorsey, Omaha Sociology (III, p. 211), which are 
also contributions to the study of totemism. 

' This first appeared, in an abridged form, in the Encyclopœdia Britannica 
(9th éd.). 

' In his Primitive Culture, Tylor had already attempted an explanation of 
totemism, to which we shall return presently, but which we shall not give here ; 
for by making totemism only a particular case of the ancestor-cult, he com- 
pletely misunderstood its importance. In this chapter we mention only those 
theories which have contributed to the progress of the study of totemism. 

go Elementary Forms of Religious Life 

beyond the mere letter of the totemic beHefs, Smith set himself 
to find the fundamental principles upon which they depend. 
In his book upon Kinship and Marriage in Early Arabia,^ he 
had already pointed out that totemism supposes a likeness in 
nature, either natural or acquired, of men and animals (or plants). 
In his The Religion of the Semites,'^ he makes this same idea the 
first origin of the entire sacrificial system : it is to totemism that 
humanity owes the principle of the communion meal. It is true 
that the theory of Smith can now be shown one-sided ; it is no 
longer adequate for the facts actually known ; but for all that, 
it contains an ingenious theory and has exercised a most fertile 
influence upon the science of religions. The Golden Bough ^ of 
Frazer is inspired by these same ideas, for totemism, which 
MacLennan had attached to the religions of classical antiquity, 
and Smith to the religions of the Semitic peoples, is here con- 
nected to the European folk-lore. The schools of MacLennan 
and Morgan are thus united to that of Mannhardt.^ 

During this time, the American tradition continued to develop 
with an independence which it has kept up until very recent 
times. Three groups of societies were the special object of the 
researches which were concerned with totemism. These are, 
first, certain tribes of the North-west, the Tlinkit, the Haida, the 
Kwakiutl, the Salish and the Tsimshian ; then, the great nation 
of the Sioux ; and finally, the Pueblo Indians in the south- 
western part of the United States. The first were studied princi- 
pally by Dall, Krause, Boas, Swanton, Hill Tout ; the second 
by Dorsey ; the last by Mindeleff, Mrs. Stevenson and Gushing.^ 
But however rich the harvest of facts thus gathered in all parts 
of the country may have been, the documents at our disposal 
were still fragmentary. Though the American religions contain 
numerous traces of totemism, they have passed the stage of real 
totemism. On the other hand, observations in Australia had 
brought little more than scattered beliefs and isolated rites, 
initiation rituals and interdictions relative to totemism. It was 
with facts taken from all these sources that Frazer attempted to 
draw a picture of totemism in its entirety. Whatever may be 
the incontestable merit of the reconstruction undertaken in 

Ï Published at Cambridge, 1885. 

* First edition, 1889. This is the arrangement of a course given at the 
University of Aberdeen in 1888. Cf. the article Sacrifice in the Encyclopœdia 
Britannica (9th edition). 

' London, 1890. A second edition in three volumes has since appeared (1900) 
and a third in five volumes is already in course of publication. 

* In this connection must be mentioned the interesting work of Sidney 
Hartland, The Legend of Perseus, 3 vols., 1894-1896. 

^ We here confine ourselves to giving the names of the authors ; their works 
will be indicated below, when we make use of them. 

Totemism as an Elementary Religion 91 

such circumstances, it could not help being incomplete and 
hypothetical. A totemic religion in complete action had not yet 
been observed. 

It is only in very recent years that this serious deficiency has 
been repaired. Two observers of remarkable ability, Baldwin 
Spencer and F. J. Gillen, discovered ^ in the interior of the 
Australian continent a considerable number of tribes whose 
basis and unity was founded in totemic beliefs. The results of 
their observations have been published in two works, which 
have given a new life to the study of totemism. The first of 
these, The Native Tribes of Central Australia,^ deals with the 
more central of these tribes, the Arunta, the Luritcha, and a 
little farther to the south, on the shores of Lake Eyre, the 
Urabunna. The second, which is entitled The Northern Tribes 
of Central Australia,'^ deals with the societies north of the 
Urabunna, occupying the territory between MacDonnell's Range 
and Carpenter Gulf. Among the principal of these we may 
mention the Unmatjera, the Kaitish, the Warramunga, the 
Worgaia, the Tjingilli, the Binbinga, the Walpari, the Gnanji 
and finally, on the very shores of the gulf, the Mara and the 

More recently, a German missionary, Carl Strehlow, who has 
also passed long years in these same Central Australian societies,^ 
has commenced to publish his own observations on two of these 
tribes, the Aranda and the Loritja (the Aruntaand Luritcha of 

^ If Spencer and Gillen have been the first to study these tribes in a scientific 
and thorough manner, they were not the first to talk about them. Howitt had 
already described the social organization of the Wuaramongo (Warramunga of 
Spencer and Gillen) in 1888 in his Further Notes on the Australian Classes in The 
Journal of the Anthropological Institute (hereafter, J. A. I.), pp. 44 f. The Arunta 
had already been briefly studied by Schulze (The Aborigines of the Upper and 
Middle Finke River, in Transactions of the Royal Society of South Australia, Vol. 
XIV, fasc. 2) : the organization of the Chingalee (the Tjingilli of Spencer and 
Gillen), the Wombya, etc., by Mathews (Wotnbya Organization of the Australian 
Aborigines, in American Anthropologist, New Series, Vol. II, p. 494 ; Divisions 
of some West Australian Tribes, ibid., p. 185; Proceedings Amer. Philos. Soc, 
XXXVII, pp. 151-152, and Journal Roy. Soc. of N.S. Wales, XXXII, p. 71 and 
XXXIII, p. III). The first results of the study made of the Arunta had also been 
published already in the Report on the Work of the Horn Scientific Expedition to 
Central Australia, Pt. IV (1896). The first part of this Report is by Stirling, the 
second by Gillen ; the entire publication was placed under the direction of 
Baldwin Spencer. 

* London, 1899. Hereafter, Native Tribes or Nat. Tr. 

^ London, 1904. Hereafter, Northern Tribes or Nor. Tr. 

* We write the Arunta, the Anula, the Tjingilli, etc., without adding the 
characteristic 5 of the plural. It does not seem very logical to add to these words, 
which are not European, a grammatical sign which would have no meaning 
except in our languages. Exceptions to this rule will be made when the name of 
the tribe has obviously been Europeanized (the Hurons for example). 

' Strehlow has been in Australia since 1892 ; at first he lived among the 
Dieri, and from them he went to the Arunta. 

92 Elementary Forms of Religious Life 

Spencer and Gillen).^ Having well mastered the language 
spoken by these peoples, ^ Strehlow has been able to bring us a 
large number of totemic myths and religious songs, which are 
given us, for the most part, in the original text. In spite of 
some differences of detail which are easily explained and whose 
importance has been greatly exaggerated,^ we shall see that 
the observations of Strehlow, though completing, making more 
precise and sometimes even rectifying those of Spencer and 
Gillen, confirm them in all that is essential. 

These discoveries have given rise to an abundant literature to 
which we shall have occasion to return. The works of Spencer 
and Gillen especially have exercised a considerable influence, 
not only because they were the oldest, but also because the facts 
were there presented in a systematic form, which was of a nature 
to give a direction to later studies,^ and to stimulate speculation. 
Their results were commented upon, discussed and interpreted 
in all possible manners. At this same time, Howitt, whose 
fragmentary studies were scattered in a number of different 
publications,^ undertook to do for the southern tribes what 
Spencer and Gillen had done for those of the centre. In his 
Native Tribes of South-East Australia,^ he gives us a view of the 
social organization of the peoples who occupy Southern Australia, 
New South Wales, and a good part of Queensland. The progress 
thus realized suggested to Frazer the idea of completing his 
Totemism by a sort of compendium ^ where would be brought 

* Die Aranda- und Loritja-Stàmme in Zenlral Australien. Four fascicules have 
been published up to the present. The last appeared at the moment when the 
present book was finished, so it could not be used. The two first have to do with 
the myths and legends, and the third with the cult. It is only just to add to the 
name of Strehlow that of von Leonhardi, who has had a great deal to do with 
this publication. Not only has he charged himself with editing the manuscripts 
of Strehlow, but by his judicious questions he has led the latter to be more 
precise on more than one point. It would be useful to consult an article 
which von Leonhardi gave the Globus, where numerous extracts from his corre- 
spondence with Strehlow will be found (Ueber einige religiose und totemistische 
V orstellungen der Arandaund Loritjain Zentral Australien, in Globus, XCI, p. 285). 
Cf. an article on the same subject by N. W. Thomas in Folk-lore, XVI, pp. 428 fl. 

^ Spencer and Gillen are not ignorant of it, but they are far from possessing 
it as thoroughly as Strehlow. 

" Notably by Klaatsch, Schlussbericht iiber meine Reise nach Australien, in 
Zeitschrift f. Ethnologie, 1907, pp. 635 fï. 

* The book of K. Langloh Parker, The Euahlayi Tribe, that of Eylmann, Die 
Eingeborenen der Kolonie Sudaustralien ; that of John Mathews. Two Repre- 
sentative Tribes of Queensland, and certain recent articles of Mathews all show 
the influence of Spencer and Gillen. 

' A list of these publications will be found in the preface to his Nat. Tr., 
pp. 8-9. 

'■ London, 1904. Hereafter we shall cite this work by the abbreviation 
Nat. Tr., but always mentioning the name of Howitt, to distinguish it from the 
first work of Spencer and Gillen, which we abbreviate in the same manner. 

" Totemism and Exogamy, 4 vols., London, 1910. The work begins with a 
re-edition of Totemism, reproduced without any essential changes. 

Totemism as an Elementary Religion 93 

together all the important documents which are concerned either 
with the totemic religion or the family and matrimonial organiza- 
tion which, rightly or wrongly, is believed to be connected with 
this religion. The purpose of this book is not to give us a general 
and systematic view of totemism, but rather to put the materials 
necessary for a construction of this sort at the disposition of 
scholars. 1 The facts are here arranged in a strictly ethno- 
graphical and geographical order : each continent, and within 
the continent, each tribe or ethnic group is studied separately. 
Though so extended a study, where so many diverse peoples 
are successively passed in review, could hardly be equally 
thorough in all its parts, still it is a useful hand-book to consult, 
and one which can aid greatly in facilitating researches. 


From this historical résumé it is clear that Australia is the 
most favourable field for the study of totemism, and therefore 
we shall make it the principal area of our observations. 

In his Totemism, Frazer sought especially to collect all the 
traces of totemism which could be found in history or ethno- 
graphy. He was thus led to include in his study societies the 
nature and degree of whose culture differs most widely : ancient 
Egypt, 2 Arabia and Greece,^ and the southern Slavs* are found 
there, side by side with the tribes of AustraUa and America. 
This manner of procedure is not at all surprising for a disciple 
of the anthropological school. For this school does not seek 
to locate religions in the social environments of which they are 
a part,^ and to differentiate them according to the different 
environments to which they are thus connected. But rather, 
as is indicated by the name which it has taken to itself, its purpose 
is to go beyond the national and historical differences to the 
universal and really human bases of the religious life. It is sup- 
posed that man has a religious nature of himself, in virtue of his 

^ It is true that at the end and at the beginning there are some general 
theories on totemism, which will be described and discussed below. But these 
theories are relatively independent of the collection of facts which accompanies 
them, for they had already been published in different articles in reviews, long 
before this work appeared. These articles are reproduced in the first volume 
(pp. 89-172). 

* Totemism, p. 12. ' Ibid., p. 15. * Ibid., p. 32. 

'" It should be noted that in this connection, the more recent work, Totemism 
and Exogamy, shows an important progress in the thought as well as the method 
of Frazer. Every time that he describes the religious or domestic institutions 
of a tribe, he sets himself to determine the geographic and social conditions in 
which this tribe is placed. Howsoever summary these analyses may be, they 
bear witness nevertheless to a rupture with the old methods of the anthropo- 
logical school. 

94 Elementary Forms of Religious Life 

own constitution, and independently of all social conditions, and 
they propose to study this.^ For researches of this sort, all 
peoples can be called upon equally well. It is true that they 
prefer the more primitive peoples, because this fundamental 
nature is more apt to be unaltered here ; but since it is found 
equally well among the most civilized peoples, it is but natural 
that they too should be called as witnesses. Consequently, all 
those who pass as being not too far removed from the origins, 
and who are confusedly lumped together under the rather im- 
precise rubric of savages, are put on the same plane and consulted 
indifferently. Since from this point of view, facts have an interest 
only in proportion to their generality, they consider themselves 
obliged to collect as large a number as possible of them ; the 
circle of comparisons could not become too large. 

Our method will not be such a one, for several reasons. 

In the first place, for the sociologist as for the historian, social 
facts vary with the social system of which they form a part ; 
they cannot be understood when detached from it. This is why 
two facts which come from two different societies cannot be 
profitably compared merely because they seem to resemble each 
other ; it is necessary that these societies themselves resemble 
each other, that is to say, that they be only varieties of the same 
species. The comparative method would be impossible, if social 
types did not exist, and it cannot be usefully applied except 
within a single type. What errors have not been committed for 
having neglected this precept ! It is thus that facts have been 
unduly connected with each other which, in spite of exterior 
resemblances, really have neither the same sense nor the same 
importance : the primitive democracy and that of to-day, the 
collectivism of inferior societies and actual socialistic tendencies, 
the monogamy which is frequent in Australian tribes and that 
sanctioned by our laws, etc. Even in the work of Frazer such 
confusions are found. It frequently happens that he assimilates 
simple rites of wild-animal-worship to practices that are really 
totemic, though the distance, sometimes very great, which 
separates the two social systems would exclude all idea of assimi- 
lation. Then if we do not wish to fall into these same errors, 
instead of scattering our researches over all the societies possible, 
we must concentrate them upon one clearly determined type. 

It is even necessary that this concentration be as close as 
possible. One cannot usefully compare facts with which he is 

^ Undoubtedly we also consider that the principal object of the science of 
religions is to find out what the religious nature of man really consists in. How- 
ever, as we do not regard it as a part of his constitutional make-up, but rather as 
the product of social causes, we consider it impossible to find it, if we leave aside 
his social environment. 

Totemism as an Elementary Religion 95 

not perfectly well acquainted. But when he undertakes to 
include all sorts of societies and civilizations, one cannot know 
any of them with the necessary thoroughness ; when he assembles 
facts from every country in order to compare them, he is obliged 
to take them hastily, without having either the means or the 
time to carefully criticize them. Tumultuous and summary 
comparisons result, which discredit the comparative method 
with many intelligent persons. It can give serious results only 
when it is applied to so limited a number of societies that each of 
them' can be studied with sufficient precision. The essential 
thing is to choose those where investigations have the greatest 
chance to be fruitful. 

Also, the value of the facts is much more important than their 
number. In our eyes, the question whether totemism has been 
more or less universal or not, is quite secondary. ^ If it interests 
us, it does so before all because in studying it we hope to discover 
relations of a nature to make us understand better what religion 
is. Now to establish these relations it is neither necessary nor 
always useful to heap up numerous experiences upon each other ; 
it is much more irnportant to have a few that are well studied 
and really significant. One single fact may make a law appear, 
where a multitude of imprecise and vague observations would 
only produce confusion. In every science, the scholar would be 
overwhelmed by the facts which present themselves to him, if he 
did not make a choice among them. It is necessary that he 
distinguish those which promise to be the most instructive, that 
he concentrate his attention upon these, and that he temporarily 
leave the others to one side. 

That is why, with one reservation which will be indicated 
below, we propose to limit our research to Australian societies. 
They fulfil all the conditions which were just enumerated. They 
are perfectly homogeneous, for though it is possible to distinguish 
varieties among them, they all belong to one common type. 
This homogeneity is even so great that the forms of social organi- 
zation are not only the same, but that they are even designated 
by identical or equivalent names in a multitude of tribes, some- 
times very distant from each other. ^ Also, Australian totemism 
is the variety for which our documents are the most complete. 
Finally, that which we propose to study in this work is the 
most primitive and simple religion which it is possible to find. 
It is therefore natural that to discover it, we address ourselves 

* We cannot repeat too frequently that the importance which we attach to 
totemism is absolutely independent of whether it was ever universal or not. 

* This is the case with the phratries and matrimonial classes ; on this point, 
see Spencer and Gillen, Northern Tribes, ch. iii; Howitt, Native Tribes, pp. 109 
and 137-142 ; Thomas, Kinship and Marriage in Australia, ch. vi and vii. 

96 Elementary Forms of Religious Life 

to societies as slightly evolved as possible, for it is evidently 
there that we have the greatest chance of finding it and studying 
it well. Now there are no societies which present this charac- 
teristic to a higher degree than the Australian ones. Not only is 
their civilization most rudimentary — the house and even the 
hut are still unknown — but also their organization is the most 
primitive and simple which is actually known ; it is that which 
we have elsewhere called organization on a basis of clans. '^ In the 
next chapter, we shall have occasion to restate its essential 

However, though making Australia the principal field of 
our research, we think it best not to leave completely aside the 
societies where totemism was first discovered, that is to say, the 
Indian tribes of North America. 

This extension of the field of comparison has nothing about 
it which is not legitimate. Undoubtedly these people are more 
advanced than those of Australia. Their civilization has become 
much more advanced : men there live in houses or under tents, 
and there are even fortified villages. The size of the society is 
much greater, and centralization, which is completely lacking 
in Australia, is beginning to appear there ; we find vast con- 
federations, such as that of the Iroquois, under one central 
authority. Sometimes a complicated system of differentiated 
classes arranged in a hierarchy is found. However, the essential 
lines of the social structure remain the same as those in Australia ; 
it is always the organization on a basis of clans. Thus we are not 
in the presence of two different types, but of two varieties of a 
single type, which are still very close to each other. They repre- 
sent two successive moments of a single evolution, so their homo- 
geneousness is still great enough to permit comparisons. 

Also, these comparisons may have their utility. Just because 
their civilization is more advanced than that of the Australians, 
certain phases of the social organization which is common to 
both can be studied more easily among the first than among 
the second. As long as men are still making their first steps in 
the art of expressing their thought, it is not easy for the observer 
to perceive that which moves them ; for there is nothing to 
translate clearly that which passes in these obscure minds which 
have only a confused and ephemeral knowledge of themselves. 
For example, religious symbols then consist only in formless 
combinations of lines and colours, whose sense it is not easy 
to divine, as we shall see. There are many gestures and move- 
ments by which interior states express themselves ; but being 

^ Division du Travail social, 3rd éd., p. 150. 

Totemism as an Elementary Religion 97 

essentially ephemeral, they readily elude observation. That is 
why totemism was discovered earUer in America than in Australia ; 
it was much more visible there, though it held relatively less place 
,in the totality of the religious life. Also, wherever behefs and 
institutions do not take a somewhat definite material form, they 
are more liable to change under the influence of the slightest 
circumstances, or to become wholly effaced from the memory. 
Thus the Australian clans frequently have something floating 
and Protean about them, while the corresponding organization 
in America has a greater stability and more clearly defined 
contours. Thus, though American totemism is further removed 
from its origins than that of AustraUa, still there are important 
characteristics of which it has better kept the memory. 

In the second place, in order to understand an institution, it 
is frequently well to follow it into the advanced stages of its 
evolution ; ^ for sometimes it is only when it is fully developed 
that its real signification appears with the greatest clearness. In 
this way also, American totemism, since it has a long history 
behind it, could serve to clarify certain aspects of Australian 
totemism. 2 At the same time, it will put us in a better condition 
to see how totemism is bound up with the forms which follow, 
and to mark its place in the general historical development of 

So in the discussions which follow, we shall not forbid ourselves 
the use of certain facts borrowed from the Indian societies of 
North America. But we are not going to study American 
totemism here ; ^ such a study must be made directly and by 
itself, and cannot be mixed with the one which we are under- 
taking ; it raises other problems and implies a wholly different 
set of special investigations. We shall have recourse to American 
facts merely in a supplementary way, and only when they seem 
to be able to make us understand Australian facts to advantage. 
It is these latter which constitute the real and immediate object 
of our researches.* 

» It is to be understood that this is not always the case. It frequently happens, 
as we have already said, that the simpler forms aid to a better understanding of 
the more complex. On this point, there is no rule of method which is applicable 
to every possible case. 

* Thus the individual totemism of America will aid us in understanding the 
function and importance of that in Australia. As the latter is very rudimentary, 
it would probably have passed unobserved. 

' Besides, there is not one unique type of totemism in America, but several 
different species which must be distinguished. 

* We shall leave this field only very exceptionally, and when a particularly 
instructive comparison seems to us to impose itself. 




The Totem as Name and as Emblem 

OWING to its nature, our study will include two parts. 
Since every religion is made up of intellectual conceptions 
and ritual practices, we must deal successively with the beliefs 
and rites which compose the totemic religion. These two elements 
of the rehgious life are too closely connected with each other 
to allow of any radical separation. -In principle, the cult is 
derived from the beUefs, yet it reacts upon them ; the myth 
is frequently modelled after the rite in order to account for it, 
especially when its sense is no longer apparent. On the other 
hand, there are behefs which are clearly manifested only through 
the rites which express them. So these two parts of our analysis 
cannot fail to overlap. However, these two orders of facts are 
so different that it is indispensable to study them separately. 
And since it is impossible to understand anything about a 
religion while unacquainted with the ideas upon which it rests, 
we must seek to become acquainted with these latter first 
of all. 

But it is not our intention to retrace all the speculations into 
which the religious thought, even of the Australians alone, has 
run. The things we wish to reach are the elementary notions 
at the basis of the religion, but there is no need of following them 
through all the development, sometimes very confused, which 
the mythological imagination of these peoples has given them. 
We shall make use of myths when they enable us to understand 
these fundamental ideas better, but we shall not make mythology 
itself the subject of our studies. In so far as this is a work of art, 
it does not fall within the jurisdiction of the simple science of 
religions. Also, the intellectual evolution from which it results 
is of too great a complexity to be studied indirectly and from a 
foreign point of view. It constitutes a very difficult problem 
which must be treated by itself, for itself and with a method 
peculiar to itself. 

102 Elementary Forms of Religious Life 

Among the beliefs upon which totemism rests, the most im- 
portant are naturally those concerning the totem ; it is with 
these that we must begin. 

"~ I 

~j^ At the basis ofjiearly-alhthe-Australian tribes- we find a group 
which holds a preponderating place in the collective life : this is 
the clan. Two essential traits characterize it. 

In the first place, the individuals who compose it consider 
themselves united by a bond of kinship, but one which is of a 
very special nature. This relationship does not come from the 
fact that they have definite blood connections with one another ; 
they are relatives from the mere fact that they have the same 
name. They are not fathers and mothers, sons or daughters, 
uncles or nephews of one another in the sense which we now give 
these words ; yet they think of themselves as forming a single 
family, which is large or small according to the dimensions of 
the clan, merely because they are collectively designated by the 
same word. When we say that they regard themselves as a 
single family, we do so because they recognize duties towards 
each other which are identical with those which have always 
been incumbent upon kindred : such duties as aid, vengeance, 
mourning, the obligation not to marry among themselves, etc. 

By this first characteristic, the clan does not differ from the 
Roman gens or the Greek yeVoç ; for this relationship also came 
merely from the fact that all the members of the gens had the 
same name,^ the nomen gentilicium. And in one s^nse, the gens 
is a clan ; but it is a variety which should not be confounded 
with the Australian clan.^ This latter is distinguished by the 
fact that its name is also the name of a determined species of 
material things with which it believes that it has very particular 
relations, the nature of which we shall presently describe ; they 
are especially relations of kinship. The species of things which 
serves to designate the clan collectively is called its totem. The 
totem of the clan is also that of each of its members. 

Each clan has its totem, which belongs to it alone ; two different 
clans of the same tribe cannot have the same. In fact, one is a 
member of a clan merely because he has a certain name. All 
who bear this name are members of it for that very reason ; in 
whatever manner they may be spread over the tribal territory, 

^ This is the definition given by Cicero : Gentiles sunt qui inter se eodem 
nomine sunt {Top. 6). (Those are of the same gens who have the same name 
among themselves.) 

* It may be said in a general way that the clan is a family group, where 
kinship results solely from a common name ; it is in this sense that the gens is a 
clan. But the totemic clan is a particular sort of the class thus constituted. 

Totemic Beliefs 103 

they all have the same relations of kinship with one another.^ 
Consequently, two groups having the same totem can only be 
two sections of the same clan. Undoubtedly, it frequently 
happens that all of a clan does not reside in the same locality, 
but has representatives in several different places. However, 
this lack of a geographical basis does not cause its unity to be 
the less keenly felt. 

In regard to the word totem, we may say that it is the one 
employed by the Ojibway, an Algonquin tribe, to designate the 
sort of thing whose name the clan bears. ^ Although this expression 
is not at all Australian,' and is found only in one single society in 
America, ethnographers have definitely adopted it, and use it 
to denote, in a general way, the system which we are describing. 
Schoolcraft was the first to extend the meaning of the word thus 
and to speak of a " totemic system." * This extension, of which 
there are examples enough in ethnography, is not without in- 
conveniences. It is not normal for an institution of this import- 
ance to bear a chance name, taken from a strictly local dialect, 
and bringing to mind none of the distinctive characteristics of 
the thing it designates. But to-day this way of employing the 
word is so universally accepted that it would be an excess of 
purism to rise against this usage. ^ 

In a very large proportion of the cases, the objects which 
serve as totems belong either to the animal or the vegetable 
kingdom, but especially to the former. Inanimate things are 
much more rarely employed. Out of more than 500 totemic 
names collected by Howitt among the tribes of south-eastern 
Australia, there are scarcely forty which are not the names of 
plants or animals ; these are the clouds, rain, hail, frost, the 

^ In a certain sense, these bonds of solidarity extend even beyond the frontiers 
of the tribe. When individuals of different tribes have the same totem, they 
have peculiar duties towards each other. This fact is expressly stated for certain 
tribes of North America (see Frazer, Totemism and Exogamy, III, pp. 57, 81, 299, 
356-357). The texts relative to Australia are less explicit. However, it is 
probable that the prohibition of marriage between members of a single totem is 

* Morgan, Ancient Society, p. 165. 

' In Australia the words employed differ with the tribes. In the regions 
observed by Grey, they said Kobong ; the Dieri say Mur du (Howitt, Nat Tr., 
p. 91) ; the Narrinyeri, Ngaitye (Talpin, in Curr, II, p. 244) ; the Warramunga, 
Mungdi or Mungdii {Nor. Tr., p. 754), etc. 

* Indian Tribes of the United States, IV, p. 86. 

* This fortune of the word is the more regrettable since we do not even know 
exactly how it is written. Some write totatn, others toodaim, or dodaim, or 
ododam (see Frazer, Totemism, p. i). Nor is the meaning of the word determined 
exactly. According to the report of the first observer of the Ojibway, J. Long, 
the word totam designated the protecting genius, the individual totem, of which 
we shall speak below (Bk. II, ch. iv) and not the totem of the clan. But the 
accounts of other explorers say exactly the contrary (on this point, see Frazer, 
Totemism and Exogamy, III, pp. 49-52). 

104 Elementary Forms of Religious Life 

moon, the sun, the wind, the autumn, the summer, the winter, 
certain stars, thunder, fire, smoke, water or the sea. It is notice- 
able how small a place is given to celestial bodies and, more 
generally, to the great cosmic phenomena, which were destined 
to so great a fortune in later religious development. Among all 
the clans of which Howitt speaks, there were only two which 
had the moon as totem, ^ two the sun,^ three a star,^ three the 
thunder,* two the lightning. ^ The rain is a single exception ; it, 
on the contrary, is very frequent.® 

These are the totems which can be spoken of as normal. But 
totemism has its abnormalities as well. It sometimes happens 
that the totem is not a whole object, but the part of an object. 
This fact appears rather rarely in Australia ;' Howitt cites only 
one example.^ However, it may well be fhat this is found with 
a certain frequency in the tribes where the totemic groups are 
excessively subdivided ; it might be said that the totems had to 
break themselves up in order to be able to furnish names to these 
numerous divisions. This is what seems to have taken place 
among the Arunta and the Loritja. Strehlow has collected 442 
totems in these two societies, of which many are not an animal 
species, but some particular organ of the animal of the species, 
such as the tail or stomach of an opossum, the fat of the kangaroo, 

We have seen that normally the totem is not an individual, 
but a species or a variety : it is not such and such a kangaroo or 
crow, but the kangaroo or crow in general. Sometimes, however, 
it is a particular object. First of all, this is necessarily the case 
when the thing serving as totem is unique in its class, as the sun, 
the moon, such or such a constellation, etc. It also happens that 
clans take their names from certain geographical irregularities or 
depressions of the land, from a certain ant-hill, etc. It is true 

* The Wotjobaluk (p. 121) and the Buandik (p. 123). * The same. 

* The Wolgal (p. 102), the Wotjobaluk and the Buandik. 

* The Muruburra (p. 117), the Wotjobaluk and the Buandik. 

* The Buandik and the Kaiabara (p. 116). It is to be remarked that all the 
examples come from only five tribes. 

' Thus, out of 204 kinds of totems, collected by Spencer and Gillen out of a 
large number of tribes, 188 are animals or plants. The inanimate objects are 
the boomerang, cold weather, darkness, fire, lightning, the moon, red ochre, 
resin, salt water, the evening star, a stone, the sun, water, the whirlwind, the 
wind and hail-stones {Nor. Tr., p. 773. Cf. Frazer, Totemism and Exogamy, 1, 
pp. 253-254). 

' Frazer {Totemism, pp. 10 and 13) cites a rather large number of cases and 
puts them in a special group which he calls split-totems, but these are taken from 
tribes where totemism is greatly altered, such as in Samoa or the tribes of Bengal. 

" Howitt, Nat. Tr., p. 107. 

* See the tables collected by Strehlow, op. cit.. II, pp. 61-72 (cf. Ill, pp. xiii- 
xvii). It is remarkable that these fragmentary totems are taken exclusively 
from animal totems. 

Totemic Beliefs 105 

that we have only a small number of examples of this in Australia ; 
but Strehlow does mention some.^ But the very causes which 
have given rise to these abnormal totems show that they are of a 
relatively recent origin. In fact, what has made certain geo- 
graphical features of the land become totems is that a mythical 
ancestor is supposed to have stopped there or to have performed 
some act of his legendary life there. ^ But at the same time, these 
ancestors are represented in the myths as themselves belonging 
to clans which had perfectly regular totems, that is to say, ones 
taken from the animal or vegetable kingdoms. Therefore, the 
totemic names thus commemorating the acts and performances 
of these heroes cannot be primitive ; they belong to a form of 
totemism that is already derived and deviated. It is even 
permissible to ask if the meteorological totems have not 
a similar origin ; for the sun, the moon and the stars are 
frequently identified with the ancestors of the mythological 
epoch. 3 

Sometimes, but no less exceptionally, it is an ancestor or a group 
of ancestors which serves as totem directly. In this case, the clan 
takes its name, not from a thing or a species of real things, but 
from a purely mythical being. Spencer and Gillen had already 
mentioned two or three totems of this sort. Among the Warra- 
munga and among the Tjingilli there are clans which bear the name 
of an ancestor named Thaballa who seems to be gaiety incarnate. * 
Another Warramunga clan bears the name of a huge fabulous 
serpent named Wollunqua, from which the clan considers itself 
descended.^ We owe other similar facts to Strehlow.* In any 
case, it is easy enough to see what probably took place. Under 
the influence of diverse causes and by the very development of 
mythological thought, the collective and impersonal totem 
became effaced before certain mythical personages who advanced 
to the first rank and became totems themselves. 

* strehlow, II, pp. 52 and 72. 

' For example, one of these totems is a cave where an ancestor of the Wild 
Cat totem rested ; another is a subterranean gallery which an ancestor of the 
Mouse clan dug, etc. {ibid., p. 72). 

=» Nat. Tr.,Y>^. $61 fi. Strehlow, II, p. 71, note 2. Howitt, iVo/. Ty., pp. 426 fif.; 
On Australian Medicine Men, J.A.I., XVI, p. 53 ; Further Notes on the Australian 
Class Systems, J. A. I.. XVIII, pp. 63 ff. 

* Thaballa means " laughing boy," according to the translation of Spencer 
and Gillen. The members of the clan which bear this name think they hear 
him laughing in the rocks which are his residence {Nor. Jr., pp. 207, 215, 226 
note). According to a myth given on p. 422, there was an initial group of 
mythical Thaballa (cf. p. 208). The clan of the Kati. " full-grown men," as 
Spencer and Gillen say, seems to be of the same sort {Nor. Tr., p. 207). 

* Nor. Tr., pp. 226 fï. 

* Strehlow, II, pp. 71 f. He mentions a totem of the Loritja and Arunta 
which is very close to the serpent Wollunqua : it is the totem of a mythical 

io6 Elementary Forms of Religious Life 

Howsoever interesting these different irregularities may be, 
they contain nothing which forces us to modify our definition 
of a totem. They are not, as has sometimes been beheved,^ 
different varieties of totems which are more or less irreducible 
into each other or into the normal totem, such as we have defined 
it. They are merely secondary and sometimes even aberrant 
forms of a single notion which is much more general, and there is 
every ground for believing it the more primitive. 

The manner in which the name is acquired is more important 
for the organization and recruiting of the clan than for religion ; 
it belongs to the sociology of the family rather than to religious 
sociology. 2 So we shall confine ourselves to indicating summarily 
the most essential principles which regulate the matter. 

In the different tribes, three different systems are in use. 

In a great number, or it might even be said, in the greater 
number of the societies, the child takes the totem of its mother, 
by right of birth : this is what happens among the Dieri and the 
Urabunna of the centre of Southern Australia ; the Wotjobaluk 
and the Gournditch-Mara of Victoria ; the Kamilaroi, the Wirad- 
juri, the Wonghibon and the Euahlayi of New South Wales ; and 
the Wakelbura, the Pitta-Pitta and the Kumandaburi of Queens- 
land, to mention only the most important names. In this case, 
owing to a law of exogamy, the mother is necessarily of a different 
totem from her husband, and on the other hand, as she lives in 
his community, the members of a single totem are necessarily 
dispersed in different localities according to the chances of their 
marriages. As a result, the totemic group lacks a territorial 

Elsewhere the totem is transmitted in the paternal line. In 
this case, if the child remains with his father, the local group is 
largely made up of people belonging to a single totem ; only the 
married women there represent foreign totems. In other words, 
each locality has its particular totem. Up until recent times, 
this scheme of organization was found in Australia only among 
the tribes where totemism was in decadence, such as the Narrin- 
yeri, where the totem has almost no religious character at all 

^ This is the case with Klaatsch, in the article already cited (see above, 
p. 92, n. 3). 

' As we indicated in the preceding chapter, totemism is at the same time of 
interest for the question of religion and that of the family, for the clan is a 
family. In the lower societies, these two problems are very closely connected. 
But both are so complex that it is indispensable to treat them separately. Also, 
the primitive family organization cannot be understood before the primitive 
religious beliefs are known ; for the latter serve as the basis of the former. This 
is why it is necessary to study totemism as a religion before studying the totemic 
clan as a family group. 

Totemic Beliefs 107 

any more.^ It was therefore possible to believe that there was 
a close connection between the totemic system and descent in 
the uterine line. But Spencer and Gillen have observed, in the 
northern part of central Australia, a whole group of tribes where 
the totemic religion is still practised but where the transmission 
of the totem is in the paternal line : these are the Warramunga, 
the Quanji, the Umbia, the Binbinga, the Mara and the Anula.^ 
Finally, a third combination is the one observed among the 
Arunta and Loritja. Here the totem of the child is not necessarily 
either that of the mother or that of the father ; it is that of a 
mythical ancestor who came, by processes which the observers 
recount in different ways,^ and mysteriously fecundated the 
mother at the moment of conception. A special process makes 
it possible to learn which ancestor it was and to which totemic 
group he belonged.* But since it was only chance which de- 
termined that this ancestor happened to be near the mother, 
rather than another, the totem of the child is thus found to 
depend finally upon fortuitous circumstances.^ 

Outside of and above the totems of clans there are totems of 
phratries which, though not differing from the former in nature, 
must none the less be distinguished from them. 

A phratry is a group of clans which are united to each other 
by particular bonds of fraternity. Ordinarily the Australian 
tribe is divided into two phratries between which the different 
clans are distributed. Of course there are some tribes where this 
organization has disappeared, but everything leads us to believe 
that it was once general. In any case, there are no tribes in 
Australia where the number of phratries is greater than two. 

Now in nearly all the cases where the phratries have a name 

^ See Taplin, The Narrinveri Tribe, in Curr, II, pp. 244 f. ; Howitt. Nat. Tr.. 
p. 131. 

* Nor. Tr., pp. 163, 169, 170, 172. It is to be noted that in all these tribes, 
except the Mara and the Anula, the transmission of the totem in the paternal 
line is only a general rule, which has exceptions. 

* According to Spencer and Gillen (Nat. Tr., pp. 123 ff.), the soul of the 
ancestor becomes reincarnate in the body of the mother and becomes the soul of 
the child ; according to Strehlow (II, pp. 51 ff.), the conception, though being the 
work of the ancestor, does not imply any reincarnation ; but in neither interpreta- 
tion does the totem of the child necessarily depend upon that of the parents. 

* Nat. Tr., p. 133 ; Strehlow, II, p. 53. 

* It is in large part the locality where the mother believes that she conceived 
which determines the totem of the child. Each totem, as we shall see, has its 
centre and the ancestors preferably frequent the places serving as centres for 
their respective totems. The totem of the child is therefore that which belongs 
to the place where the mother believes that she conceived. As this should 
generally be in the vicinity of the place which serves as totemic centre for her 
husband, the child should generally follow the totem of his father. It is un- 
doubtedly this which explains why the greater part of the inhabitants of a given 
locality belong to the same totem (Nat. Tr., p. 9). 

io8 Elementary Forms of Religious Life 

whose meaning has been estabhshed, this name is that of an 
animal ; it would therefore seem that it is a totem. This has 
been well demonstrated in a recent work by A. Lang.^ Thus, 
among the Gournditch (Victoria), the phratries are called 
Krokitch and Kaputch ; the former of the words designates the 
white cockatoo and the latter the black cockatoo. 2 The same 
expressions are found again among the Buandik and the Wotjoba- 
luk.2 Among the Wurunjerri, the names employed are Bunjil 
and Waang, which designate the eagle-hawk and the crow.* 
The words Mukwara and Kilpara are used for the same purpose 
in a large number of tribes of New South Wales ; ^ they designate 
the same birds. ^ It is also the eagle-hawk and the crow which 
have given their names to the two phratries of the Ngarigo and 
the Wolgal.' Among the Kuinmurbura, it is the white cockatoo 
and the crow.^ Many other examples might be cited. Thus we 
are led to regard the phratry as an ancient clan which has been 
dismembered ; the actual clans are the product of this dismember- 
ment, and the solidarity which unites them is a souvenir of their 
primitive unity.' It is true that in certain tribes, the phratries 
no longer have special names, as it seems ; in others where these 
names exist, their meaning is no longer known, even to the 
members. But there is nothing surprising in this. The phratries 
are certainly a primitive institution, for they are everywhere in a 
state of regression ; their descendants the clans have passed to 
the first rank. So it is but natural that the names which they 
bore should have been effaced from memory little by little, 
when they were no longer understood ; for they must belong 
to a very archaic language no longer in use. This is proved by 
the fact that in many cases where we know the animal whose 
name the phratry bears, the word designating this animal in the 
current language is very different from the one employed here.^" 

1 The Secret of the Totem, pp. 159 ff. Cf. Fison and Howitt, Kamilaroi and 
Kurnai, pp. 40 f. ; John Mathews, Eaglehawk and Crow ; Thomas, Kinship and 
Marriage in Australia, pp. 52 ff. 

'•' Howitt, Nat. Tr., p. 124. 

3 Howitt, pp. 121, 123, 124 ; Curr, III, p. 461. 

* Howitt, p. 126. ^ Howitt, pp. 98 ff. 

* Curr, n, p. 165 ; Brough Smyth, I, p. 423 ; Howitt, op. cit., p. 429. 
' Howitt, pp. loi, 102. 

* J. Mathews, Two Representative Tribes of Queensland, p. 139. 

* Still other reasons could be given in support of this hypothesis, but it would 
be necessary to bring in considerations relative to the organization of the family, 
and we wish to keep these two studies separate. Also this question is only of 
secondary interest to our subject. 

1" For example, Mukwara, which is the name of a phratry among the Barkinji, 
the Paruinji and the Milpulko, designates the eagle-hawk, according to Brough 
Smyth ; now one of the clans of this phratry has the eagle-hawk as totem. But 
here the animal is designated by the word Bilyara. Many cases of the same 
thing are cited by Lang, op. cit., p. 162. 

Totemic Beliefs 109 

Between the totem of the phratry and the totems of the clans 
there exists a sort of relation of subordination. In fact, in 
principle each clan belongs to one and only one phratry ; it is 
very exceptional that it has representatives in the other phratry. 
This is not met with at all except among certain central tribes, 
notably the Arunta ; ^ also even where, owing to disturbing 
influences, overlappings of this sort have taken place, the great 
part of the clan is included entirely within one or the other of the 
two groups of the tribe ; only a small minority is to be found in 
the other one.^ As a rule then, the two phratries do not overlap 
each other ; consequently, the list of totems which an individual 
may have is predetermined by the phratry to which he belongs. 
In other words, the phratry is like a species of which the clans 
are varieties. We shall presently see that this comparison is not 
purely metaphorical. 

In addition to the phratries and clans, another secondary group 
is frequently met with in Australian societies, which is not with- 
out a certain individuality : these are the matrimonial classes. 

By this name they designate certain subdivisions of the 
phratry, whose number varies with the tribe : there are sometimes 
two and sometimes four per phratry.^ Their recruiting and 
operation are regulated by the two following principles. In the 
first place, each generation in a phratry belongs to different 
clans from the immediately preceding one. Thus, when there 
are only two classes per phratry, they necessarily alternate 
with each -other every generation. The children make up the 
class of which their parents are not members ; but grandchildren 
are of the same class as their grandparents. Thus, among the 
Kamilaroi, the Kupathin phratry has two classes, Ippai and 
Kumbo ; the Dilby phratry, two others which are called Murri 
and Kubbi. As descent is in the uterine line, the child is in the 
phratry of its mother ; if she is a Kupathin, the child will be 
one also. But if she is of the Ippai class, he will be a Kumbo ; 
if the child is a girl, her children will again be in the Ippai class. 

1 Spencer and Gillen, Nat. Tr., p. 115. According to Howitt (op. cit., pp. 121 
and 454). among the Wotjobaluk, the clan of the pelican is found in the two 
phratries equally. This fact seems doubtful to us. It is ver>' possible that the 
two clans may have two varieties of pelicans as totems. Information given by 
Mathews on the same tribe seems to point to this (Aboriginal Tribes of N.S. Wales 
and Victoria, in Journal and Proceedings of the Royal Society of N.S. Wales. 1904, 
pp. 287 f.). 

* In connection with this question, see our memoir on Le Totémisme, in the 
Année Sociologique, Vol. V, pp. 82 If. 

^ On the question of Australian matrimonial classes in general, see our 
memoir on La Prohibition de l'inceste, in the Année Soc, I, pp. 9 it., and especially 
for the tribes with eight classes, L'Organisation matrimoniale des sociétés Austra- 
liennes, in Année Soc, VIII, pp. 118-147. 

no Elementary Forms of Religious Life 

Likewise, the children of the women of the Muni class will be 
in the Kubbi class, and the children of the Kubbi women will be 
Murri again. When there are four classes per phratry, instead of 
two, the system is naturally more complex, but the principle is 
the same. The four classes form two couples of two classes each, 
and these two classes alternate with each other every generation 
in the manner just indicated. Secondly, the members of one 
class can in principle ^ marry into only one of the classes of the 
other phratry. The Ippai must marry into the Kubbi class and 
the Murri into the Kumbo class. It is because this organization 
profoundly affects matrimonial relations that we give the group 
the name of matrimonial class. 

Now it may be asked whether these classes do not sometimes 
have totems like the phratries and clans. 

This question is raised by the fact that in certain tribes of 
Queensland, each matrimonial class has dietetic restrictions that 
are peculiar to it. The individuals who compose it must abstain 
from eating the flesh of certain animals which the others may 
consume freely. ^ Are these animals not totems ? 

But dietetic restrictions are not the characteristic marks of 
totemism. The totem is a name first of all, and then, as we shall 
see, an emblem. Now in the societies of which we just spoke, 
there are no matrimonial classes which bear the name of an 
animal or plant, or which have an emblem. ^ Of course it is 
possible that these restrictions are indirectly derived from 
totemism. It might be supposed that the animals which these 
interdictions protect were once the totems of clans which have 
since disappeared, while the matrimonial classes remained. It 
is certain that they have a force of endurance which the clans 
do not have. Then these interdictions, deprived of their original 

^ This principle is not maintained everywhere with an equal strictness. In 
the central tribes of eight classes notably, beside the class with which marriage 
is regularly permitted, there is another with which a sort of secondary concu- 
binage is allowed (Spencer and Gillen, Nor. Tr., p. 106). It is the same with 
certain tribes of four classes. Each class has a choice between the two classes 
of the other phratry. This is the case with the Kabi (see Mathews, in Curr, III, 

* See Roth, Ethnological Studies among the North-West-Central Queensland 
Aborigines, pp. 56 ff. ; Palmer, Notes on some Australian Tribes, J.A.I., XIII 
(1884), pp. 302 tf. 

^ Nevertheless, some tribes are cited where the matrimonial classes bear the 
names of animals or plants : this is the case with the Kabi (Mathev.^, Two Repre- 
sentative Tribes, p. 150), the tribes observed by Mrs. Bates {The Marriage Laws 
and Customs of the West Australian Aborigines, in Victorian Geographical Journal, 
XXIII-XXIV, p. 47), and perhaps in two tribes observed by Palmer. But 
these facts are very rare and their significance badly established. Also, it is not 
surprising that the classes, as well as the sexual groups, should sometimes adopt 
the names of animals. This exceptional extension of the totemic denominations 
in no way modifies our conception of totemism. 

Totemic Beliefs m 

field, may have spread themselves out over the entire class, 
since there were no other groups to which they could be attached. 
But it is clear that if this regulation was born of totemism, it 
represents only an enfeebled and denatured form of it.^ 

All that has been said of the totem in Australian societies is 
equally applicable to the Indian tribes of North America. The 
only difference is that among these latter, the totemic organization 
has a strictness of outline and a stability which are not found in 
Australia. The Australian clans are not only very numerous, 
but in a single tribe their number is almost unlimited. Observers 
cite some of themi as examples, but without ever succeeding in 
giving us a complete list. This is because the list is never definitely 
terminated. The same process of dismemberment which broke 
up the original phratries and give birth to clans properly so-called 
still continues within these latter ; as a result of this progressive 
crumbling, a clan frequently has only a very small effective 
force. 2 In America, on the contrary, the totemic system has 
better defined forms. Although the tribes there are considerably 
larger on the average, the clans are less numerous. A single 
tribe rarely has more than a dozen of them,^ and frequently less ; 
each of them is therefore a much more important group. But 
above all, their number is fixed ; they know their exact number, 
and they it tell to us.'* 

* Perhaps the same explanation is applicable to certain other tribes of the 
South-East and the East where, if we are to believe the informers of Howitt, 
totems specially attached to each matrimonial class are to be found. This is the 
case among the Wiradjuri, the Wakelbura and the Bunta-Murra on the BuUoo 
River (Howitt, Nat. Tr., pp. 2io, 221, 226). However, the evidence collected is 
suspect, according to his own admission. In fact, it appears from the lists which 
he has drawn up, that many totems are found equally in the two classes of the 
same phratry. 

The explanation which we propose, after Frazer (Totemism and Exogamy, 
pp. 531 ff.), raises one difficulty. In principle, each clan and consequently each 
totem, is represented equally in the two classes of a single phratry, since one of 
the classes is that of the children and the other that of the parents from whom 
the former get their totems. So when the clans disappeared, the totemic inter- 
dictions which survived should have remained in both matrimonial classes, 
while in the actual cases cited, each class has its own. Whence comes this 
differentiation ? The example of the Kaiabara (a tribe of southern Queensland) 
allows us to see how it may have come about. In this tribe, the children have 
the totem of their mother, but it is particularized by some distinctive mark. If 
the mother has the black eagle-hawk as totem, the child has the white eagle- 
hawk (Howitt, Nat. Tr., p. 229). This appears to be the beginning of a tendency 
for the totems to differentiate themselves according to the matrimonial classes. 

* A tribe of only a few hundred members frequently has fifty or sixty clans, 
or even many more. On this point, see Durkheim and Mauss, De quelques formes 
primitives de classification, in the Année Sociologique, Vol. VI, p. 28, n. i. 

* Except among the Pueblo Indians of the South-West, where they are more 
numerous. See Hodge, Pueblo Indian Clans, in American Anthropologist, ist 
series, Vol. IX, pp. 345 ff. It may always be asked whether the groups which 
have these totems are clans or sub-clans. 

* See the tables arranged by Morgan, Ancient Society, pp. 153-185. 

112 Elementary Forms of Religious Life 

This difference is due to the superiority of their social economy. 
From the moment when these tribes were observed for the first 
time, the social groups were strongly attached to the soil, and 
consequently better able to resist the decentralizing forces which 
assailed them. At the same time, the society had too keen a 
sentiment of its unity to remain unconscious of itself and of the 
parts out of which it was composed. The example of America 
thus enables us to explain even better the organization at the 
base of the clans. We would take a mistaken view, if we judged 
this only on the present conditions in Australia. In fact, it is in 
a state of change and dissolution there, which is not at all normal ; 
it is much rather the product of a degeneration which we see, 
due both to the natural decay of time and the disorganizing effect 
of the whites. To be sure, it is hardly probable that the Australian 
clans ever had the dimensions and solid structure of the American 
ones. But there must have been a time when the distance between 
them was less considerable than it is to-day, for the American 
societies would never have succeeded in making so solid a 
structure if the clans had always been of so fluid and inconsistent 
a nature. 

This greater stability has even enabled the archaic system of 
phratries to maintain itself in America with a clearness and a 
relief no longer to be found in Australia. We have just seen that 
in the latter continent the phratry is everywhere in a state of 
decadence ; very frequently it is nothing more than an anony- 
mous group ; when it has a name, this is either no longer under- 
stood, or in any case, it cannot mean a great deal to the native, 
since it is borrowed from a foreign language, or from one no 
longer spoken. Thus we have been able to infer the existence of 
totems for phratries only from a few survivals, which, for the 
most part, are so slightly marked that they have escaped the 
attention of many observers. In certain parts of America, on 
the contrary, this institution has retained its primitive importance. 
The tribes of the North-west coast, the Tlinkit and the Haida 
especially, have now attained a relatively advanced civilization ; 
yet they are divided into two phratries which are subdivided into 
a certain number of clans : the phratries of the Crow and the 
Wolf among the Tlinkit, ^ of the Eagle and the Crow among the 
Haida. 2 And this division is not merely nominal ; it corresponds 
to an ever-existing state of tribal customs and is deeply marked 
with the tribal life. The moral distance separating the clans is 

*i Krause, Die Tlinkit-Indianer , p. 112; Swan ton, Social Condition, Beliefs 
and Linguistic Relationship of the Tlingit Indians, in XXVIth Rep., p. 308. 
* Swan ton, Contnbutions to the Ethnology of the Haida, p. 62. 

Totemic Beliefs 113 

very slight in comparison with that separating the phratries. ^ 
The name of each is not a word whose sense is forgotten or only 
vaguely known ; it is a totem in the full sense of the term ; they 
have all its essential attributes, such as will be described below. ^ 
Consequently, upon this point also, American tribes must not be 
neglected, for we can study the totems of phratries directly there, 
while Australia offers only obscure vestiges of them. 


But the totem is not merely a name ; it is an emblem, a 
veritable coat-of-arms whose analogies with the arms of heraldry 
have often been remarked. In speaking of the Australians, Grey 
says, " each family adopt an animal or vegetable as their crest 
and sign," ^ and what Grey calls a family is incontestably a 
clan. Also Fison and Howitt say, " the Australian divisions 
show that the totem is, in the iirst place, the badge of a group." * 
Schoolcraft says the same thing about the totems of the Indians 
of North America. " The totem is in fact a design which cor- 
responds to the heraldic emblems of civilized nations, and 
each person is authorized to bear it as a proof of the identity 
of the family to which it belongs. This is proved by the real 
etymology of the word, which is derived from dodaim, which 
means village or the residence of a family group." ^ Thus when 
the Indians entered into relations with the Europeans and con- 
tracts were formed between them, it was with its totem that 
each clan sealed the treaties thus concluded.* 

The nobles of the feudal period carved, engraved and designed 
in every way their coats-of-arms upon the walls of their castles, 
their arms, and every sort of object that belonged to them ; the 
blacks of Australia and the Indians of North America do the 

^ " The distinction between the two clans is absolute in every respect," says 
Swanton, p. 68 ; he gives the name clan to what we call phratries. The two 
phratries, he says elsewhere, are like two foreign nations in their relations to 
each other. 

* Among the Haida at least, the totem of the real clans is altered more than 
that of the phratries. In fact, usage permits a clan to sell or give away the right 
of bearing its totem, as a result of which each clan has a number of totems, some 
of which it has in common with other clans (see Swanton, pp. 107 and 268). 
Since Swanton calls the phratries clans, he is obliged to give the name of family 
to the real clans, and of householu to the regular families. But the real sense of 
his terminology is not to be doubted. 

* Journals of two Expeditions %n N.W. and W. Australia, II, p. 228. 

* Kamilaroi and Kurnai, p. 165. 

^ Indian Tribes, I, p. 420 ; cf. I, p. 52. This etymology is very doubtful. 
Cf. Handbook of American Indians North of Mexico [Smithsonian Inst. Bur. of 
EthnoL, Pt. II, S.V., Totem, p. 787). 

* Schoolcraft, Indian Tribes, III, 184 ; Garrick Mallery, Picture Writing of 
the American Indians, in Tenth Report, 1893, p. 377. 

114 elementary Forms of Religious Lije 

same thing with their totems. The Indians who accompanied 
Samuel Hearne painted their totems on their shields before going 
into battle. 1 According to Charlevoix, in time of war, certain 
tribes of Indians had veritable ensigns, made of bits of bark 
fastened to the end of a pole, upon which the totems were repre- 
sented. ^ Among the Tlinkit, when a conflict breaks out between 
two clans, the champions of the two hostile groups wear helmets 
over their heads, upon which are painted their respective totems.^ 
Among the Iroquois, they put the skin of the animal which 
serves as totem upon each wigwam, as a mark of the clan.'* 
According to another observer, the animal was stuffed and set 
up before the door.^ Among the Wyandot, each clan has its 
own ornaments and its distinctive paintings.® Among the 
Omaha, and among the Sioux generally, the totem is painted 
on the tent.' 

Wherever the society has become sedentary, where the tent is 
replaced by the house, and where the plastic arts are more 
fully developed, the totem is engraved upon the woodwork 
and upon the walls. This is what happens, for example, among 
the Haida, the Tsimshian, the Salish and the Tlinkit. " A 
very particular ornament of the house, among the Tlinkit," 
says Krause, " is the totemic coat-of-arms." Animal forms, 
sometimes combined with human forms, are engraved upon 
the posts at the sides of the door of entry, which are as high 
as 15 yards ; they are generally painted with very bright colours.^ 
However, these totemic decorations are not very numerous 
in the Tlinkit village ; they are found almost solely before the 
houses of the chiefs and rich men. They are much more frequent 
in the neighbouring tribe of the Haida ; here there are always 
several for each house.' With its many sculptured posts arising 
on every hand, sometimes to a great height, a Haida village 
gives the impression of a sacred city, all bristling with belfries 
or little minarets.^" Among the Salish, the totem is frequently 
represented upon the interior walls of the house. ^^ Elsewhere, it 

* Hearne, Journey to the Northern Ocean, p. 148 (quoted from Frazer, Totemism, 

P- 30)- 

* Charlevoix, Histoire et description de la Nouvelle France, V, p. 329. 

* Krause, Tlinkil-Ihdianer, p. 248. 

* Erminnie A. Smith, Myths of the Iroquois, in Sec. Rep. of the Bur. of Ethnol., 
p. 78. 

* Dodge, Our Wild Indians, p. 225. 

« Powell, Wyandot Government, in First Rep. of the Bur of Ethnol., 1881, p. 64. 

"> Dorsey, Omaha Sociology, in Third Rep., pp. 229, 240, 248. 

« Krause, op. cit., pp. 130 f. * Krause, p. 308. 

^o See a photograph of a Haida village in Swanton, op. cit., PI. IX. Cf. Tylor, 
Totem Post of the Haida Village of M asset, J. A. I., New Series I, p. 133. 

11 Hill Tout, Report on the Ethnology of the Statlumh of British Columbia, J. A ./., 
XXXV, p. 155- 

Totemic Beliefs 115 

is found upon the canoes, the utensils of every sort and the 
funeral piles. ^ 

The preceding examples are taken exclusively from the Indians 
of North America. This is because sculpture, engravings and 
permanent figurations are not possible except where the technique 
of the plastic arts has reached a degree of perfection to which 
the Australian tribes have not yet attained. Consequently the 
totemic representations of the sort which we just mentioned are 
rarer and less apparent in Australia than in America. However, 
cases of them are cited. Among the Warramunga, at the end of 
the burial ceremonies, the bones of the dead man are interred, 
after they have been dried and reduced to powder ; beside the 
place where they are deposited, a figure representing the 
totem is traced upon the ground. ^ Among the Mara and the 
Anula, the body is placed in a piece of hollow wood decorated 
with designs characteristic of the totem. ^ In New South Wales, 
Oxley found engravings upon the trees near the tomb where a 
native was buried ^ to which Brough Smyth attributes a totemic 
character. The natives of the Upper Darling carve totemic 
images upon their shields. ^ According to Collins, nearly all the 
utensils are covered with ornaments which probably have the 
same significance ; figures of the same sort are found upon the 
rocks.® These totemic designs may even be more frequent than 
it seems, for, owing to reasons which will be discussed below, it 
is not always easy to see what their real meaning is. 

These different facts give us an idea of the considerable place 
held by the totem in the social life of the primitives. However, 
up to the present, it has appeared to us as something relatively 
outside of the man, for it is only upon external things that we 
have seen it represented. But totemic images are not placed 
only upon the walls of their houses, the sides of their canoes, 
their arms, their utensils and their tombs ; they are also found 
on the bodies of the men. They do not put their coat-of-arms 
merely upon the things which they possess, but they put it upon 
their persons ; they imprint it upon their flesh, it becomes a 

^ Krause, op. cit., p. 230 ; Swanton, Haida, pp. 129, 135 ff. ; Schoolcraft, 
op. cit., I, pp. 52-53, 337, 356. In the latter case the totem is represented up- 
side down, in sign of mourning. Similar usages are found among the Creek 
(C. Swan, in Schoolcraft, V, p. 265) and the Delaware (Heckewelder, An Account 
of the History, Manners and Customs of the Indian Nations who once inhabited 
Pennsylvania, pp. 246—247). 

* Spencer and Gillen, Nor. Tr., pp. 168, 537, 540. 
' Ibid., p. 174. 

* Brough Smyth, The Aborigines of Victoria, I, p. 99 n. 

'■• Brough Smyth, I, p. 284. Strehlow cites a fact of the same sort among the 
Arunta (III, p. 68). 

« An Account of the English Colony in N.S. Wales, II, p. 381. 

ii6 Elementary Forms of Religious Life 

part of them, and this world of representations is even by far 
the more important one. 

In fact, it is a very general rule that the members of each 
clan seek to give themselves the external aspect of their totem. 
At certain religious festivals among the Tlinkit, the person who 
is to direct the ceremonies wears a garment which represents, 
either wholly or in part, the body of the animal whose name he 
bears. ^ These same usages are also found in all the North-West 
of America. 2 Ihey are found again among the Minnitaree, when 
they go into combat, ^ and among the Indians of the Pueblos.^ 
Elsewhere, when the totem is a bird, men wear the feathers of this 
bird on their heads. ^ Among the Iowa, each clan has a special 
fashion of cutting the hair. In the Eagle clan, two large tufts 
are arranged on the front of the head, while there is another 
one behind ; in the Buffalo clan, they are arranged in the form 
of horns.® Among the Omaha, analogous arrangements are 
found : each clan has its own head-dress. In the Turtle clan, for 
example, the hair is all shaved off, except six bunches, two on 
each side of the head, one in front, and one behind, in such a 
way as to imitate the legs, the head and the tail of the animal.' 

But it is more frequently upon the body itself that the totemic 
mark is stamped : for this is a way of representation within the 
capacity of even the least advanced societies. It has sometimes 
been asked whether the common rite of knocking out a young 
man's two upper teeth at the age of puberty does not have the 
object of reproducing the form of the totem. The fact is not 
established, but it is worth mentioning that the natives themselves 
sometimes explain the custom thus. For example, among the 
Arunta, the extraction of teeth is practised only in the clans of 
the rain and of water ; now according to tradition, the object 
of this operation is to make their faces look like certain black 
clouds with light borders which are believed to announce the 
speedy arrival of rain, and which are therefore considered things 
of the same family.^ This is a proof that the native himself is 
conscious that the object of these deformations is to give him, 
at least conventionally, the aspect of his totem. Among these 

^ Krause, p. 237. 

^ Swanton, Social Condition, Beliefs and Linguistic Relationship of the Tlingit 
Indians, in XXVIth Rep., pp. 435 If. ; Boas, The Social Organization and Secret 
Societies of the Kwakiutl Indians, p. 358. 

^ Frazer, Totemism, p. 26. 

* Bourke, The Snake Dance of the Moquis of Arizona, p. 229 ; J. W. Fewkes, 
The Group of Tusayan Ceremonials called Katcinas, in XVth Rep., 1897, pp. 

' Millier, Geschichte der Amerikanischen Urreligionen. p. 327. 

^ Schoolcraft, op. cit.. Ill, p. 269. 

' Dorsey, Omaha Sociol., Third Rep., pp. 229, 238, 240, 245. 

* Spencer and Gillen, Nat. Tr., p. 451. 

Totemic Beliefs 117 

same Arunta, in the course of the rites of sub-incision, certain 
gashes are cut upon the sisters and the future wife of the novice ; 
scars result from these, whose form is also represented upon a 
certain sacred object of which we shall speak presently and which 
is called the churinga ; as we shall see, the lines thus drawn 
upon the churinga are emblematic of the totem. ^ Among the 
Kaitish, the euro is believed to be closely connected with the 
rain ; 2 the men of the rain clan wear little ear-rings made of 
euro teeth. ^ Among the Yerkla, during the initiation the young 
man is given a certain number of slashes which leave scars ; the 
number and form of these varies with the totems.^ An informer 
of Fison mentions the same fact in the tribes observed by him.^ 
According to Howitt, a relationship of the same sort exists 
among the Dieri between certain arrangements of scars and the 
water totem. ^ Among the Indians of the North- West, it is a 
very general custom for them to tattoo themselves with the 

But even if the tattooings which are made by mutilations or 
scars do not always have a totemic significance,^ it is different 
with simple designs drawni upon the body : they are generally 
representations of the totem. It is true that the native does not 
carry them every day. When he is occupied with purely economic 
occupations, or when the small family groups scatter to hunt or 
fish, he does not bother with all this paraphernalia, which is 
quite complicated. But when the clans unite to live a common 
life and to assist at the religious ceremonies together, then he 
must adorn himself. As we shall see, each of the ceremonies 
concerns a particular totem, and in theory the rites which are 
connected with a totem can be performed only by the men of 
that totem. Now those who perform,^ who take the part of 

^ Spencer and Gillen, Nat. Tr., p. 257. 

* The meaning of these relations will be seen below (Bk. II, ch. iv). 
' Spencer and Gillen, Nor. Tr., p. 296. 

* Howitt, Nat. Tr., pp. 744-746 ; cf. p. 129. 

* Kamilaroi and Kurnai, p. 66 n. It is true that other informers contest 
this fact. 

* Howitt, Nat. Tr., p. 744. 

' Swanton, Contributions to the Ethnology of the Haida, pp. 41 ff., PI. XX and 
XXI ; Boas, The Social Organization of the Kwakiutl, p. 318 ; Swanton, Tlingit, 
PI. XVI ff. — In one place, outside the two ethnographic regions which we are 
specially studying, these tattooings are put on the animals which belong to the 
clan. The Bechuana of South Africa are divided into a certain number of clans ; 
there are the people of the crocodile, the buffalo, the monkey, etc. Now the 
crocodile people, for example, make an incision in the ears of their cattle whose 
form is like the jaws of this animal (Casalis, Les Basoutos, p. 221). According to 
Robertson Smith, the same custom existed among the ancient Arabs {Kinship 
and Marriage in Early Arabia, pp. 212-214). 

However, according to Spencer and Gillen, there are some which have no 
religious sense (see Nat. Tr., pp. 41 f. ; Nor. Tr., pp. 45, 54-56). 

* Among the Arunta, this rule has exceptions which will be explained below. 

ii8 Elementary Forms of Religious Life 

officiants, and sometimes even those who assist as spectators, 
always have designs representing the totem on their bodies.^ 
One of the principal rites of initiation, by which a young man 
enters into the religious life of the tribe, consists in painting the 
totemic symbol on his body.^ It is true that among the Arunta 
the design thus traced does not always and necessarily represent 
the totem of the initiated ; ^ but these are exceptions, due, 
undoubtedly, to the disturbed state of the totemic organization 
of this tribe. ^ Also, even among the Arunta, at the most solemn 
moment of the initiation, which is its crown and consecration, 
when the neophyte is allowed to enter the sanctuary where all 
the sacred objects belonging to the clan are preserved, an 
emblematic painting is placed upon him ; this time, it is the 

1 spencer and Gillen, Nat. Tr., p. 162 ; Nor. Tr., pp. 179, 259, 292, 295 f. ; 
Schulze, he. cit., p. 221. The thing thus represented is not always the totem 
itself, but one of those things which, being associated to this totem, are regarded 
as being in the same family of things. 

* This is the case, for example, among the Warramunga, the Walpari, the 
Wulmala, the Tjingilli, the Umbaia and the Unmatjera {Nor. Tr., 339, 348). 
Among the Warramunga, at the moment when the design is executed, the per- 
formers address the initiated with the following words : " That mark belongs to 
your place ; do not look out along another place." " This means," say Spencer 
and Gillen, " that the young man must not interfere with ceremonies belonging 
to other totems than his own : it also indicates the very close association which 
is supposed to exist between a man and his totem and any spot especially con- 
nected with the totem " (Nor. Tr., p. 584 and n.). Among the Warramunga, the 
totem is transmitted from father to child, so each locality has its own. 

' Spencer and Gillen, Nat. Tr., pp. 215, 241, 376. 

* It will be remembered (see above, p. 107) that in this tribe, the child may 
have a different totem than his father, his mother, or his relatives in general. 
Now the relatives on both sides are the performers designated for the ceremonies 
of initiation. Consequently, since in principle a man can have the quality of 
performer or officiant only for the ceremonies of his own totem, it follows that in 
certain cases the rites by which the young man is initiated must be in connection 
with a totem that is not his own. That is why the paintings made on the body 
of the novice do not necessarily represent his own totem : cases of this sort will 
be found in Spencer and Gillen, Nat. Tr., p. 229. That there is an anomaly here 
is well shown by the fact that the circumcision falls to the totem which pre- 
dominates in the local group of the initiate, that is to say, to the one which 
would be the totem of the initiate himself, if the totemic organization were not 
disturbed, if among the Arunta it were what it is among the Warramunga (see 
Spencer and Gillen, ibid., p. 219). 

The same disturbance has had another consequence. In a general way, its 
effect is to extend a little the bonds attaching each totem to a special group, 
since each totem may have members in all the local groups possible, and even in 
the two phratries. The idea that these ceremonies of a totem might be cele- 
brated by an individual of another totem — an idea which is contrary to the very 
principles of totemism, as we shall see better after a while — has thus been 
accepted witnout too much resistance. It has been admitted that a man to 
whom a spirit revealed the formula for a ceremony had the right of presiding 
over it, even when he was not of the totem in question himself (Nat. Tr., p. 519). 
But that this is an exception to the rule and the product of a sort of toleration is 
proved by the fact that the beneficiary of the formula does not have the free 
disposition of it ; if he transmits it — and these transmissions are frequent — it can 
be only to a member of the totem which the rite concerns (Nat. Tr., ibid.). 

Totemic Beliefs 119 

totem of the young man which is thus represented. ^ The bonds 
which unite the individual to his totem are even so strong that 
in the tribes on the North-west coast of North America, the 
emblem of the clan is painted not only upon the living but also 
upon the dead : before a corpse is interred, they put the totemic 
mark upon it.^ 


These totemic decorations enable us to see that the totem is 
not merely a name and an emblem. It is in the course of the 
religious ceremonies that they are employed ; they are a part 
of the liturgy ; so while the totem is a collective label, it also 
has a religious character. In fact, it is in connection with it, 
that things are classified as sacred or profane. It is the very 
type of sacred thing. 

The tribes of Central Australia, especially the Arunta, the 
Loritja, the Kaitish, the Unmatjera, and the Ilpirra,^ make con- 
stant use of certain instruments in their rites which are called the 
churinga by the Arunta, according to Spencer and Gillen, or the 
tjunmga, according to Strehlow.^ They are pieces of wood or 
bits of poHshed stone, of a great variety of forms, but generally 
oval or oblong.^ Each totemic group has a more or less important 
collection of these. Upon each of these is engraved a design 
representing the totem of this same groups A certain number of 
the churinga have a hole at one end, through which goes a thread 
made of human hair or that of an opossum. Those which, are 
made of wood and are pierced in this way serve for exactly the 
same purposes as those instruments of the cult to which English 
ethnographers have given the name of " bull-roarers." By 
means of the thread by which they are suspended, they are 
whirled rapidly in the air in such a way as to produce a sort of 
humming identical with that made by the toys of this name 
still used by our children ; this deafening noise has a ritual 

* Nat. Tr., p. 140. In this case, the novice keeps the decoration with which 
he has thus been adorned until it disappears of itself by the effect of time. 

* Boas. General Report on the Indians of British Columbia in British Associa- 
tion for the Advancement of Science, Fifth Rep. of the Committee on the N.W. Tribes 
of the Dominion of Canada, p. 41. 

* There are also some among the Warramunga, but in smaller numbers than 
among the Arunta ; they do not figure in the totemic ceremonies, though they 
do have a place in the m5rths {Nor. Tr., p. 163). 

* Other names are used by other tribes. We give a generic sense to the 
Arunta term because it is in this tribe that the churinga have the most important 
place and have been studied the best. 

* Strehlow, II, p. 81. 

* There are a few which have no apparent design (see Spencer and Gillen, 
Nat. Tr., p. 144). 

120 Elementary Forms of Religious Life 

significance and accompanies all ceremonies of any importance. 
These sorts of churinga are real bull-roarers. But there are 
others which are not made of wood and are not pierced ; con- 
sequently they cannot be employed in this way. Nevertheless, 
they inspire the same religious sentiments. 

In fact, every churinga, for whatever purpose it may be 
employed, is counted among the eminently sacred things ; there 
are none which surpass it in religious dignity. This is indicated 
even by the word which is used to designate them. It is not 
only a substantive but also an adjective meaning sacred. Also, 
among the several names which each Arunta has, there is one so 
sacred that it must not be revealed to a stranger ; it is pronounced 
but rarely, and then in a low voice and a sort of mysterious 
murmur. Now this name is called the aritna churinga (aritna 
means name) . ^ In general, the word churinga is used to designate 
all ritual acts ; for example, ilia churinga signifies the cult of the 
emu.^ Churinga, when used substantively, therefore designates 
the thing whose essential characteristic is sacredness. Profane 
persons, that is to say, women and young men not yet initiated 
into the religious life, may not touch or even see the churinga ; 
they are only allowed to look at it from a distance, and even 
this is only on rare occasions.' 

The churinga are piously kept in a special place, which the 
Arunta call the ertnatulunga^ This is a cave or a sort of cavern 
hidden in a deserted place. The entrance is carefully closed by 
means of stones so cleverly placed that a stranger going past 
it could not suspect that the religious treasury of the clan was 
so near to him. The sacred character of the churinga is so great 
that it communicates itself to the locality where they are stored : 
the women and the uninitiated cannot approach it. It is only 
after their initiation is completely finished that the young men 
have access to it : there are some who are not esteemed worthy 

* Nat. Tr., pp. 139 and 648 ; Strehlow, II, p. 75. 

* Strehlow, who writes tjurunga, gives a slightly different translation to the 
word. " This word," he says, " means that which is secret and personal {der 
eigene geheime). Tju is an old word which means hidden or secret, and runga 
means that which is my own." But Kempe, who has more authority than 
Strehlow in this matter, translates tju by great, powerful, sacred (Kempe, 
Vocabulary of the Tribes inhabiting Macdonell Ranges, s.v. Tju, in Transactions 
of the R. Society of Victoria, Vol. XIII). At bottom, the translation of Strehlow 
is not so different from the other as might appear at first glance, for what is 
secret is hidden from the knowledge of the profane, that is, it is sacred. As for 
the meaning given to runga, it appears to us very doubtful. The ceremonies of 
the emu belong to all the members of that clan ; all may participate in them ; 
therefore they are not personal to any one of them. 

* Nat. Tr., pp. 130-132 ; Strehlow, II, p. 78. A woman who has seen a 
churinga or a man who has shown one to her are both put to death. 

* Strehlow calls this place, defined in exactly the same terms as by Spencer 
and Gillen, arknaiiaua instead of ertnatulunga (Strehlow, II, p. 78). 

Totemic Beliefs 121 

of this favour except after years of trial. ^ The religious nature 
radiates to a distance and communicates itself to all the sur- 
roundings :, everything near by participates in this same nature 
and is therefore withdrawn from profane touch. Is one man 
pursued by another ? If he succeeds in reaching the ertnatulunga, 
he is saved ; he cannot be seized there. ^ Even a wounded animal 
which takes refuge there must be respected.^ Quarrels are for- 
bidden there. It is a place of peace, as is said in the Germanic 
societies ; it is a sanctuary of the totemic group, it is a veritable 
place of asylum. 

But the virtues of the churinga are not manifested merely 
by the way in which it keeps the profane at a distance. If it is 
thus isolated, it is because it is something of a high religious 
value whose loss would injure the group and the individuals 
severely. It has aU sorts of marvellous properties : by contact 
it heals wounds, especially those resulting from circumcision ; * 
it has the same power over sickness ; ^ it is useful for making 
the beard grow ; ^ it confers important powers over the totemic 
species, whose normal reproduction it ensures ; ' it gives men 
force, courage and perseverance, while, on the other hand, it 
depresses and weakens their enemies. This latter belief is so 
firmly rooted that when two combatants stand pitted against 
one another, if one sees that the other has brought churinga 
against him, he loses confidence and his defeat is certain.^ Thus 
there is no ritual instrument which has a more important place 
in the religious ceremonies. ^ By means of various sorts of anoint- 
ings, their powers are communicated either to the officiants or 
to the assistants ; to bring this about, they are rubbed over the 
members and stomach of the faithful after being covered with 
grease \^^ or sometimes they are covered with a down which 
flies away and scatters itself in every direction when they are 

^ Nor. Tr., p. 270 ; Nat. Tr., p. 140. 

* Nat. Tr., p. 135. 

' Strehlow, II, p. 78. However, Strehlow says that if a murderer takes 
refuge near an ertnatulunga, he is unpityingly pursued there and put to death. 
We find some difficulty in conciliating this fact with the privilege enjoyed by 
animals, and ask ourselves if the rigour with which a criminal is treated is not 
something recent and should not be attributed to a weakening of the taboo 
which originally protected the ertnatulunga. 

« Nat. Tr.. p. 248. 

* Ibid., pp. 545 f. Strehlow, II, p. 79. For example, the dust detached by 
rubbing a churinga with a stone, when dissolved in water, forms a potion which 
restores health to sick persons. 

« Nat. Tr., pp. 545 f. Strehlow (II, p. 79) contests this fact. 

' For example, the churinga of the yam totem, if placed in the soil, make the 
yams grow (Nor. Tr., p. 275), It has the same power over animals (Strehlow, 
II, pp. 76,78; III, pp. 3.7)- 

* Nat. Tr., p. 135 ; Strehlow, II, p. 79. 

" Nor. Tr., p. 278. " Ibid., p. 180. 

122 Elementary Forms of Religious Life 

whirled ; this ' a way of disseminating the virtues which are in 

But they are not useful merely to individuals ; the fate of the 
clan as a whole is bound up with theirs. Their loss is a disaster ; 
it is the greatest misfortune which can happen to the group.- 
Sometimes they leave the ertnatulunga, for example when they 
are loaned to other groups.^ Then follows a veritable public 
mourning. , For two weeks, the people of the totem weep and 
lament, covering their bodies with white clay just as they do 
when they have lost a relative.* And the churinga are not left 
at the free disposition of everybody ; the ertnatulunga where 
they are kept is placed under the control of the chief of the 
group. It is true that each individual has special rights to some 
of them ; ^ yet, though he is their proprietor in a sense, he 
cannot make use of them except with the consent and under 
the direction of the chief. It is a collective treasury ; it is the 
sacred ark of the clan.^ The devotion of which they are the 
object shows the high price that is attached to them. The 
respect with which they are handled is shown by the solemnity 
of the movements.' They are taken care of, they are greased, 
rubbed, polished, and when they are moved from one locality to 
another, it is in the midst of ceremonies which bear witness to 
the fact that this displacement is regarded as an act of the highest 

Now in themselves, the churinga are objects of wood and 
stone like all others ; they are distinguished from profane 
things of the same sort by only one particularity : this is that 
the totemic mark is drawn or engraved upon them. So it is 
this mark and this alone which gives them their sacred character. 
It is true that according to Spencer and Gillen, the churinga 
serve as the residence of an ancestor's soul and that it is the 
presence of this soul which confers these properties.^ While 

1 Nor. Tr., pp. 272 f. "^ Nat. Tr., p. 135. 

' One group borrows the churinga of another with the idea that these latter 
will communicate some of the virtues which are in them and that their presence 
will quicken the vitality, of the individuals and of the group {Nat. Tr., pp. 158 fi.). 

* Ibid., p. 136. 

* Each individual is united by a particular bond to a special churinga which 
assures him his life, and also to those which he has received as a heritage from 
his parents. 

* Nat. Tr., p. 154 ; Nor. Tr., p. 193. The churinga are so thoroughly collec- 
tive that they take the place of the " message-sticks " with which the messengers 
of other tribes are provided, when they are sent to summon foreign groups to 
a ceremony (Nat. Tr., pp. 141 f.). 

' Ibid., p. 326. It should be remarked that the bull-roarers are used in the 
same way (Mathews, Aboriginal Tribes of N.S. Wales and Victoria, in Jour, of 
Roy. Soc. of N.S. Wales. XXXVIII, pp. 307 f.). 

8 Nat. Tr., pp. 161,. 259 fl. » Ibid., p. 138. 

Totemic Beliefs 123 

declaring this interpretation inexact, Strehlow, in his turn, 
proposes another which does not differ materially from the other : 
he claims that the churinga are considered the image of the 
ancestor's body, or the body itself. ^ So, in any case, it would be 
sentiments inspired by the ancestor which fix themselves upon 
the material object, and convert it into a sort of fetish. But in the 
first place, both conceptions, — which, by the way, scarcely differ 
except in the letter of the myth, — have obviously been made up 
afterwards, to account for the sacred character of the churinga. 
In the constitution of these pieces of wood and bits of stone, 
and in their external appearance, there is nothing which pre- 
destines them to be considered the seat of an ancestral soul, or 
the image of his body. So if men have imagined this myth, it 
was in order to explain the religious respect which these things 
inspired in them, and the respect was not determined by the 
myth. This explanation, like so many mythological explanations, 
resolves the question only by repeating it in shghtly different 
terms ; for saying that the churinga is sacred and saying that it 
has such and such a relation with a sacred being, is merely to 
proclaim the same fact in two different ways ; it is not accounting 
for them. Moreover, according to the avowal of Spencer and 
Gillen, there are some churinga among the Arunta which are made 
by the old men of the group, to the knowledge of and before the 
eyes of all ; ^ these obviously do not come from the great ancestors. 
However, except for certain differences of degree, they have the 
same power as the others and are preserved in the same manner. 
Finally, there are whole tribes where the churinga is never 
associated with a spirit. ^ Its religious nature comes to it, then, 
from some other source, and whence could it come, if not from 
the totemic stamp which it bears ? It is to this image, therefore, 
that the demonstrations of the rite are really addressed ; it is 
this which sanctifies the object upon which it is carved. 

Among the Arunta and the neighbouring tribes, there are two 
other liturgical instruments closely connected with the totem 

' Strehlow, I, Vorwort. in fine ; II, pp. 76, 77 and 82. For the Arunta, it is 
the body of the ancestor itself ; for the Loritja, it is only an image. 

2 when a child has just been born, the mother shows the father the spot where 
she believes that the soul of the ancestor entered her. The father, accompanied 
by a few relatives, goes to this spot and looks for the churinga which the ancestor 
is believed to have left at the moment that he reincarnated himself. If it is found 
there, some old man of the group undoubtedly put it there (this is the hypothesis 
of Spencer and Gillen). If they do not find it, a new churinga is made in a deter- 
mined manner {Nat. Tr., p. 132. Cf. Strehlow, II, p. 80). 

^ This is the case among the Warramunga, the Urabunna, the Worgaia, the 
Umbaia, the ïjingilli and the Guangi (Nor. Tr., pp. 258, 275 f.). Then, say 
Spencer and Gillen, " they were regarded as of especial value because of their 
association with a totem " {ibid., p. 276). There are examples of the same fact 
among the Arunta {Nat. Tr., 15O). 

124 Elementary Forms of Religious Life 

and the churinga itself, which ordinarily enters into their com- 
position : they are the nurtunja and the waninga. 

The nurtunja,^ which is found among the northern Arunta 
and their immediate neighbours, ^ is made up principally of a 
vertical support which is either a single lance, or several lances 
united into a bundle, or of a simple pole.^ Bunches of grass are 
fastened all around it by means of belts or little cords made of 
hair. Above this, down is placed, arranged either in circles or in 
parallel lines which run from the top to the bottom of the support. 
The top is decorated with the plumes of an eagle-hawk. This is 
only the most general and typical form ; in particular cases, it 
has all sorts of variations.*- 

The waninga, which is found only among the southern Arunta, 
the Urabunna and the Loritja, has no one unique model either. 
Reduced to its most essential elements, it too consists in a vertical 
support, formed by a long stick or by a lance several yards high, 
with sometimes one and sometimes two cross-pieces.^ In the 
former case, it has the appearance of a cross. Cords made either 
of human hair or opossum or bandicoot fur diagonally cross the 
space included between the arms of the cross and the extremities 
of the central axis ; as they are quite close to each other, they 
form a network in the form of a lozenge. When there are two, 
cross-bars, these cords go from one to the other and from these 
to the top and bottom of the support. They are sometimes 
covered with a layer of down, thick enough to conceal the founda- 
tion. Thus the waninga has the appearance of a veritable flag.* 

Now the nurtunja and the waninga, which figure in a multitude 
of important rites, are the object of a religious respect quite like 
that inspired by the churinga. The process of their manufacture 
and erection is conducted with the greatest solemnity. Fixed in 
the earth, or carried by an officiant, they mark the central point 
of the ceremony : it is about them that the dances take place 
and the rites are performed. In the course of the initiation, the 

* Strehlow writes tnatanja (I, pp. 4-5). 

■ The Kaitish, the Ilpirra, the Unmatjera ; but it is rare among the latter. 
' The pole is sometimes replaced by very long churinga, placed end to ead. 

* Sometimes another smaller one is hung from the top of the nurtunja. In 
other cases, the nurtunja is in the form of a cross or a T. More rarely, the central 
support is lacking [Nat. Tr., pp. 298-300, 360-364, 627). 

■'' Sometimes there are even three of these cross-bars. 

* Nat. Tr., pp. 231-234, 306-310, 627. In addition to the nurtunja and the 
waninga, Spencer and Gillen distinguish a third sort of sacred post or flag, called 
the kanana {Nat. Tr., pp. 364, 370, 629), whose functions they admit they have 
been unable to determine. They merely note that it " is regarded as something 
common to the members of all the totems." According to Strehlow (II, p. 23, 
n. 2) the kanana of which Spencer and Gillen speak, is merely the nurtunja of 
the Wild Cat totem. As this animal is the object of a tribal cult, the veneration 
of which it is the object might easily be common to all the clans. 

Totemic Beliefs 125 

novice is led to the foot of a nurtunja erected for the occasion. 
Someone says to him, " There is the nurtunja of your father ; 
many young men have already been made by it." After that, 
the initiate must kiss the nurtunja. ^ By this kiss, he enters 
into relations with the religious principle which resides there ; 
it is a veritable communion which should give the young man 
the force required to support the terrible operation of sub-incision. ^ 
The nurtunja also plays a considerable rôle in the mythology of 
these societies. The myths relate that in the fabulous times of 
the great ancestors, the territory of the tribe was overrun in 
every direction by companies composed exclusively of individuals 
of the same totem. ^ Each of these troops had a nurtunja with 
it. When it stopped to camp, before scattering to hunt, the 
members fixed their nurtunja in the ground, from the top of 
which their churinga was suspended.^ That is equivalent to 
saying that they confided the most precious things they had to 
it. It was at the same time a sort of standard which served as a 
rallying-centre for the group. One cannot fail to be struck by 
the analogies between the nurtunja and the sacred post of the 
Omaha. ^ 

Now its sacred character can come from only one cause : that 
is that it represents the totem materially. The vertical lines or 
rings of down which cover it, and even the cords of different 
colours which fasten the arms of the waninga to the central 
axis, are not arranged arbitrarily, according to the taste of the 
makers ; they must conform to a type strictly determined by 
tradition which, in the minds of the natives, represents the 
totem. ^ Here we cannot ask, as we did in the case of the churinga, 
whether the veneration accorded to this instrument of the cult 
is not merely the reflex of that inspired by the ancestors ; for it 
is a rule that each nurtunja and each waninga last only during 
the ceremony where they are used. They are made all over 
again every time that it is necessary, and when the rite is once 
accomplished, they are stripped of their ornaments and the 
elements out of which they are made are scattered.' They are 
nothing more than images — and temporary images at that — 

^ Nor. Tr., p. 342 ; Nat. Tr., p. 309. 

* Nat. Tr., p. 255. ^ Ibid., ch. x and xi. * Ibid., pp. 138, 144. 

* See Dorsey, Siouan Cults, Xlth Rep., p. 413 ; Omaha Sociology, Third Rep., 
p. 234. It is true that there is only one sacred post for the tribe, while there is 
a nurtunja for each clan. But the principle is the same. 

° Nat. Tr., pp. 232, 308, 313, 334, etc. ; Nor. Tr., 1S2, 186, etc. 

' Nat. Tr., p. 346. It is true that some say that the nurtunja represents the 
lance of the ancestor who was at the head of each clan in Alcheringa times. But 
it is only a symbolic representation of it ; it is not a sort of relic, like the churinga, 
which is beheved to come from the ancestor himself. Here the secondary 
character of the explanation is very noticeable. 

126 Elementary Forms of Religious Life 

of the totem, and consequently it is on this ground, and on this 
ground alone, that they play a religious rôle. 

So the churinga, the nurtunja and the waninga owe their 
religious nature solely to the fact that they bear the totemic 
emblem. It is the emblem that is sacred. It keeps this character, 
no matter where it may be represented. Sometimes it is painted 
upon rocks ; these paintings are called churinga ilkinia, sacred 
drawings.^ The decorations with which the officiants and 
assistants at the religious ceremonies adorn themselves have 
the same name : women and children may not see them.^ In the 
course of certain rites, the totem is drawn upon the ground. 
The way in which this is done bears witness to the sentiments 
inspired by this design, and the high value attributed to it ; it 
is traced upon a place that has been previously sprinkled, and 
saturated with human blood, ^ and we shall presently see that 
the blood is in itself a sacred liquid, serving for pious uses only. 
When the design has been made, the faithful remain seated on 
the ground before it, in an attitude of the purest devotion.^ If 
we give the word a sense corresponding to the mentality of the 
primitive, we may say that they adore it. This enables us to 
understand how the totemic blazon has remained something very 
precious for the Indians of North America : it is always sur- 
rounded with a sort of religious halo. 

But if we are seeking to understand how it comes that these 
totemic representations are so sacred, it is not without interest 
to see what they consist in. 

Among the Indians of North America, they are painted, 
engraved or carved images which attempt to reproduce as faith- 
fully as possible the external aspect of the totemic animal. The 
means employed are those which we use to-day in similar circum- 
stances, except that they are generally cruder. But it is not the 
same in Australia, and it is in the Australian societies that 
we must seek the origin of these representations. Although the 
Australian may show himself sufficiently capable of imitating 
the forms of things in a rudimentary way,^ sacred representations 
generally seem to show no ambitions in this line : they consist 
essentially in geometrical designs drawn upon the churinga, the 
nurtunga, rocks, the ground, or the human body. They are 
either straight or curved lines, painted in different ways,^ and 

^ Nat. Tr., pp. 614 ff., esp. p. 617 ; Nor. Tr., p. 749. 

'^ Nat. Tr., p. 624. * Ibid., p. 179. * Ibid., p. 181. 

^ See the examples given in Spencer and Gillen, Nat. Tr., Fig. 131. Here are 
designs, many of which evidently have the object of representing animals, plants, 
the heads of men, etc., though of course all are very conventional. 

" Nat. Tr., p. O17 ; Nor. Tr., p. 716 ft. 

Totemic Beliefs 127 

the whole having only a conventional meaning. The connection 
between the figure and the thing represented is so remote and 
indirect that it cannot be seen, except when it is pointed out. 
Only the members of the clan can say what meaning is attached 
to such and such combinations of lines. ^ Men and women are 
generally represented by semicircles, and animals by whole 
circles or spirals, ^ the tracks of men or animals by lines of points, 
etc. The meaning of the figures thus obtained is so arbitrary 
that a single design may have two different meanings for the 
men of two different totems, representing one animal here, and 
another animal or plant there. This is perhaps still more apparent 
with the nurtunja and waninga. Each of them represents a 
different totem. But the few and simple elements which enter 
into their composition do not allow a great variety of combina- 
tions. The result is that two nurtunja may have exactly the 
same appearance, and yet express two things as different as a 
gum tree and an emu.^ When a nurtunja is made, it is given a 
meaning which it keeps during the whole ceremony, but which, 
in the last resort, is fixed by convention. 

These facts prove that if the Australian is so strongly inclined 
to represent his totem, it is in order not to have a portrait of it 
before his eyes which would constantly renew the sensation of 
it ; it is merely because he feels the need of representing the 
idea which he forms of it by means of material and external 
signs, no matter what these signs may be. We are not yet ready 
to attempt to understand what has thus caused the primitive 
to write his idea of his totem upon his person and upon different 
objects, but it is important to state at once the nature of the 
need which has given rise to these numerous representations.* 

^ Nat. Tr., p. 145 ; Strehlow, II, p. 80. 

* Nat. Tr., p. 151. ' Ibid., p. 346. 

* It cannot be doubted that these designs and paintings also have an aesthetic 
character ; here is the first form of art. Since they are also, and even above all, a 
written language, it follows that the origins of design and those of writing are one. 
It even becomes clear that men commenced designing, not so much to fix upon 
wood or stone beautiful forms which charm the senses, as to translate his thought 
into matter (cf. Schoolcraft, Indian Tribes, I, p. 405 ; Dorsey, Sionan Cults, 
pp. 394 £f.). 


TOTEMic BELIEFS — Continued 

The Toiemic Animal and Man 

BUT totemic images are not the only sacred things. There 
are real things which are also the object of rites, because 
of the relations which they have with the totem : before all 
others, are the beings of the totemic species and the members 
of the clan. 

First of all, since the designs which represent the totem arouse 
religious sentiments, it is natural that the things whose aspect 
these designs reproduce should have this same property, at least 
to a certain degree. 

For the most part, these are animals or plants. The profane 
function of vegetables and even of animals is ordinarily to serve 
as food ; then the sacred character of the totemic animal or plant 
is shown by the fact that it is forbidden to eat them. It is true 
that since they are sacred things, they can enter into the com- 
position of certain mystical repasts, and we shall see, in fact, 
that they sometimes serve as veritable sacraments ; yet normally 
they cannot be used for everyday consumption. Whoever 
oversteps this rule, exposes himself to grave dangers. It is not 
that the group always intervenes to punish this infraction 
artificially ; it is believed that the sacrilege produces death 
' automatically. A redoubtable principle is held to reside in the 
totemic plant or animal, which cannot enter into the profane 
^organism without disorganizing it or destroying it.^ In certain 
tribes at least, only the old men are free from this prohibition ; ^ 
we shall see the reason for this later. 

However, if this prohibition is formal in a large number of 

1 See the cases in Taplin, The Narrinyeri, p. 63 ; Howitt, Nat. Tr., pp. 146, 
769 ; Fison and Howitt, Kamilaroi and Kurnai, p. 169 ; Roth, Superstition, 
Magic and Medicine, § 150 ; Wyatt, Adelaide and Encounter Bay Tribe, in Woods, 
p. 168 ; Meyer, ibid., p. 186. 

* This is the case with the Warramunga [Nor. Tr., p. 168). 


Totemic Beliefs 129 

tribes * — with certain exceptions which will be mentioned later 
— it is incontestable that it tends to weaken as the old totemic 
organization is disturbed. But the restrictions which remain 
even then prove that these attenuations are not admitted without 
difficulty. For example, when it is permitted to eat the plant 
or animal that serves as totem, it is not possible to do so freely ; 
only a little bit may be taken at a time. To go beyond this 
amount is a ritual fault that has grave consequences. ^ Elsewhere, 
the prohibition remains intact for the parts that are regarded 
as the most precious, that is to say, as the most sacred ; for 
example, the eggs or the fat.^ In still other parts, consumption 
is not allowed except when the animal in question has not yet 
reached full maturity.* In this case, they undoubtedly think 
that its sacred character is not yet complete. So the barrier 
which isolates and protects the totemic being yields but slowly 
and with active resistance, which bears witness to what it must 
have been at first. 

It is true that according to Spencer and Gillen these restrictions 
are not the remnants of what was once a rigorous prohibition 
now losing hold, but the beginnings of an interdiction which is 
only commencing to establish itself. These writers hold ^ that 
at first there was a complete liberty of consumption and that 
the limitations which were presently brought are mlatively 
recent. They think they find the proof of their theory in the two 
following facts. In the first place, as we just said, there are 
solemn occasions when the members of the clan or their chief 
not only may, but must eat the totemic animal or plant. More- 
over, the myths relate that the great ancestors, the founders 
of the clans, ate their totems regularly : now, it is said, these 
stories cannot be understood except as an echo of a time when 
the present prohibitions did not exist. 

But the fact that in the course of certain solemn ceremonies 
a consumption of the totem, and a moderate one at that, is 
ritually required in no way implies that it was once an ordinary 
article of food. Quite on the contrary, the food that one eats at 
a mystical repast is essentially sacred, and consequently for- 
bidden to the profane. As for the myths, a somewhat summary 
critical method is employed, if they are so readily given the 

*■ For example, among the Warramunga, the Urabunna, the Wonghibon, the 
Yuin, the Wotjobaluk, the Buandik, Ngeumba, etc. 

* Among the Kaitish, if a man of the clan eats too much of his totem, the 
members of the other phratry have recourse to a magic operation which is expected 
to kill him (Nor. Tr., p. 284 ; cf. Nat. Tr., p. 204 ; Langloh Parker, The Euulitavi 
Tribe, p. 20). 

* Nat. Tr., p. 202, n. ; Strehlow, II, p. 58. 

* Nor. Tr., p. 173. * Nat. Tr., pp. 207 fi. 


130 Elementary Forms of Religious Life 

value of historical documents. In general, their object is to 
interpret existing rites rather than to commemorate past events ; 
they are an explanation of the present much more than a history. 
In this case, the traditions according to which the ancestors 
of the fabulous epoch ate their totem are in perfect accord with 
the beliefs and rites which are always in force. The old men and 
those who have attained a high religious dignity are freed from 
the restrictions under which ordinary men are placed : ^ they 
can eat the sacred thing because they are sacred themselves ; 
this rule is in no way peculiar to totemism, but it is found in all 
the most diverse religions. Now the ancestral heroes were nearly 
gods. It is therefore still more natural that they should eat the 
sacred food ; "^ but that is no reason why the same privilege 
should be awarded to the simple profane.^ 

However, it is neither certain nor even probable that the 
prohibition was ever absolute. It seems to have always been 
suspended in case of necessity, as, for example, when a man 
is famished and has nothing else with which to nourish himself.* 
A stronger reason for this is found when the totem is a form of 
nourishment which a man cannot do without. Thus there are a 
great many tribes where water is a totem ; a strict prohibition 
is manifestly impossible in this case. However, even here, the 
privilege granted is submitted to certain restrictions which 
greatly limit its use and which show clearly that it goes against 
a recognized principle. Among the Kaitish and the Warramunga, 
a man of this totem is not allowed to drink water freely ; he may 
not take it up himself ; he may receive it only from the hands of 
a third party who must belong to the phratry of which he is not 
a member. ^ The complexity of this procedure and the embarrass- 
ment which results from it are still another proof that access to 
the sacred thing is not free. This same iTile is applied in certain 
central tribes every time that the totem is eaten, whether from 

^ See above, p. 128. 

* It should also be borne in mind that in these myths the ancestors are never 
represented as nourishing themselves regularly with their totem. Consumption 
of this sort is, on the contrary, the exception. Their ordinary food, according to 
Strehlow, was the same as that of the corresponding animal (see Strehlow, I, p. 4). 

' Also, this whole theory rests upon an entirely arbitrary hypothesis : 
Spencer and Gillen, as well as Frazer, admit that the tribes of central Australia, 
and especially the Arunta, represent the most archaic and consequently the 
purest form of totemism. We shall presently say why this conjecture seems to 
us to be contrary to all probability. It is even probable that these authors would 
not have accepted their thesis so readily if they had not refused to regard 
totemism as a religion and if they had not consequently misunderstood the 
sacred character of the totem. 

* Taplin, The Narrinyeri, p. 64 ; Howitt, Nat. Tr., pp. 145 and 147 ; Spencer 
and Gillen, Nat. Tr., p. 202 ; Grey, loc. cit. ; Curr, III, p. 462. 

" Nor. Tr., pp. 160, 167. It is not enough that the intermediary be of another 
totem : as we shall see, every totem of a phratry is forbidden in a certain measure 
for the members of the phratry who are of a different totem. 

Totemic Beliefs 131 

necessity or any other cause. It should also be added that 
when this formality is not possible, that is, when a man is alone 
or with members of his own phratry only, he may, on necessity, 
do without an intermediary. It is clear that the prohibition is 
susceptible of various moderations. 

Nevertheless, it rests upon ideas so strongly ingrained in the 
mind that it frequently survives its original cause for being. 
We have seen that in all probability, the different clans of a 
phratry are only subdivisions of one original clan which has 
been dismembered. So there was a time when all the clans, 
being welded together, had the same totem ; consequently, 
wherever the souvenir of this common origin is not completely 
effaced, each clan continues to feel itself united to the others and 
to consider that their totems are not completely foreign to it. 
For this reason an individual may not eat freely of the totems 
held by the different clans of the phratry of which he is a member ; 
he may touch them only if the forbidden plant or animal is given 
him by a member of the other phratry. ^ 

Another survival of the same sort is the one concerning the 
maternal totem. There are strong reasons for believing that at 
first, the totem was transmitted in the uterine line. Therefore, 
wherever descent in the paternal line has been introduced, this 
probably took place only after a long period, during which the 
opposite principle was applied and the child had the totem of his 
mother along with all the restrictions attached to it. Now in 
certain tribes where the child inherits the paternal totem to-day, 
some of the interdictions which originally protected the totem of 
his mother still survive : he cannot eat it freely. ^ In the present 
state of affairs, however, there is no longer anything corresponding 
to this prohibition. 

* Nov. Tr., p. 167. We can now explain more easily how it happens that 
when an interdiction is not observed, it is the other phratry which revenges this 
sacrilege (see above, p. 129, n. 2). It is because it has an interest in seeing that 
the rule is observed. In fact, they believe that when the rule is broken, the 
totemic species may not reproduce abundantly. Now the members of the other 
phratry consume it regularly : therefore it is they who are affected. That is 
why they revenge themselves. 

* This is the case among the Loritja (Strehlow, II, pp. 60, 61), the Worgaia, 
the Warramunga, the Walpari, the Mara, the Anula and the Binbinga {Nor. Tr., 
pp. 166, 167, 171, 173). It may be eaten by a Warramunga or a Walpari, but 
only when ofiered by a member of the other phratry. Spencer and Gillen remark 
(p. 167, n.), that in this regard the paternal and the maternal totems appear to be 
under different rules. It is true that in both cases the offer must come from the 
other phratry. But when it is a question of the paternal totem, or the totem 
properly so-called, this phratry is the one to which the totem does not belong ; 
for the maternal totem, the contrary is the case. Probably the principle was 
first established for the former, then mechanically extended to the other, though 
the situation was different. When the rule had once become established that the 
prohibition protecting the totem could be neglected only on the invitation of the 
other phratry, it was apphed also to the maternal totem. 

132 Elementary Forms of Religious Life 

To this prohibition of eating is frequently added that of 
killing the totem, or picking it, when it is a plant. ^ However, 
here also there are exceptions and tolerations. These are especially 
in the case of necessity, when the totem is a dangerous animal, ^ 
for example, or when the man has nothing to eat. There are even 
tribes where men are forbidden to hunt the animals whose 
names they bear, on their own accounts, but where they may kill 
them for others,^ But the way in which this act is generally 
accomplished clearly indicates that it is something illicit. One 
excuses himself as though for a fault, and bears witness to the 
chagrin which he suffers and the repugnance which he feels,* 
while precautions are taken that the animal may suffer as little 
as possible.^ 

In addition to these fundamental interdictions, certain cases 
of a prohibition of contact between a man and his totem are 
cited. Thus among the Omaha, in the clan of the Elk, no one 
may touch any part of the body of a male elk ; in the sub-clan of 
the Buffalo, no one is allowed to touch the head of this animal.* 
Among the Bechuana, no man dares to clothe himself in the skin 
of his totem.' But these cases are rare ; and it is natural that 
they should be exceptional, for normally a man must wear the 
image of his totem or something which brings it to mind. The 
tattooings and the totemic costumes would not be possible if all 
contact were forbidden. It has also been remarked that this 
prohibition has not been found in Australia, but only in those 
societies where totemism has advanced far from its original 
form ; it is therefore probably of late origin and due perhaps to 
the influence of ideas that are really not totemic at all.® 

^ For example, among the Warramunga [Nor. Tr., p. i66), the Wotjobaluk. 
the Buandik, the Kurnai (Hewitt, pp. 146 f.) and the Narrinyeri (Taplin, The 
Narrinyeri, p. 63). 

* Even this is not always the case. An Arunta of the Mosquito totem must 
not kill this insect, even when it bothers him : he must confine liimself to driving 
it away (Strehlow, II, p. 58 ; cf. Taplin, p. 63). 

^ Among the Kaitish and the Unmatjera (Nor. Tr., p. 160). It even happens 
that in certain cases an old man gives a young one of a different totem one of his 
churinga, so that he may kill the donor's totem more easily {ibid., p. 272). 

* Howitt, Nat. Tr., p. 146 ; Grey, op. cit., II, p. 228 ; Casalis, Basoutos, 
p. 221. Among these latter, " one must be purified after committing such a 
sacrilege." '' Strehlow, II, pp. 58, 59, 61. 

* Dorsey, Omaha Sociology, Ilird Rep., pp. 225, 231. ' Casalis, ibid. 

^ Even among the Omaha, it is not certain that the interdictions of contact, 
certain examples of which we have just cited, are really of a totemic nature, for 
many of them have no direct connection with the animal that serves as totem of 
the clan. Thus in the sub-clan of the Eagle, the characteristic interdiction is 
against touching the head of a buffalo (Dorsey, op. cit., p. 239) ; in another sub- 
clan with the same totem, they must not touch verdigris, charcoal, etc. {ibid., 
p. 245). 

We do not mention other interdictions mentioned by Frazer, such as those of 
naming or looking at the animal or plant, for it is still less certain that they 
are of totemic origin, except perhaps for certain facts observed among the 

Totemic Beliefs 133 

If we now compare these various interdictions with those~l 
whose object is the totemic emblem, contrarily to all that could 
be foreseen, it appears that these latter are more numerous, 
stricter, and more severely enforced than the former. The figures 
of all sorts which represent the totem are surrounded with a 
respect sensibly superior to that inspired by the very being 
whose form these figures reproduce. The churinga, the nurtunja 
and the waninga can never be handled by the women or the 
uninitiated, who are even allowed to catch glimpses of it only 
very exceptionally, and from a respectful distance. On the other 
hand, the plant or animal whose name the clan bears may be 
seen and touched by everybody. The churinga are preserved 
in a sort of temple, upon whose threshold all noises from the 
profane life must cease ; it is the domain of sacred things. On 
the contrary, the totemic animals and plants live in the profane J 
world and are mixed up with the common everyday life. Since | 
the number and importance of the interdictions which isolate 
a sacred thing, and keep it apart, correspond to the degree of 
sacredness with which it is invested, we arrive at the remarkable 
conclusion that the images of totemic beings are more sacred than 
the beings themselves. Also, in the ceremonies of the cult, it is I 
the churinga and the nurtunja which have the most important 
place ; the animal appears there only very exceptionally. In a 
certain rite, of which we shall have occasion to speak, ^ it serves 
as the substance for a religious repast, but it plays no active 
rôle. The Arunta dance around the nurtunja, and assemble 
before the image of their totem to adore it, but a similar demon- 
stration is never made before the totemic being itself. If this 
latter were the primarily sacred object, it would be with it, 
the sacred animal or plant, that the young initiate would com- 
municate when he is introduced into the religious life ; but we 
have seen that on the contrary, the most solemn moment of the 
initiation is the one when the novice enters into the sanctuary 
of the churinga. It is with them and the nurtunja that he com- 
municates. The representations of the totem are therefore more 
actively powerful than the totem itself. 

Bechuana [Totemism, pp. 12-13). Frazer admits too readily — and in this regard, 
he has imitators — that the prohibitions against eating or touching an animal 
depend upon totemic beliefs. However, there is one case in Australia, where the 
sight of the animal seems to be forbidden. According to Strehlow (II, p. 59), 
among the Arunta and the Loritja, a man who has the moon as totem must not 
look at it very long, or he would be likely to die at the hand of an enemy. But 
we believe that this is a unique case. We must not forget, also, that astronomical 
totems were probably not primitive in Australia, so this prohibition may be the 
product of a complex elaboration. This hypothesis is confirmed by the fact that 
among the Euahlayi, looking at the moon is forbidden to all mothers and children, 
no matter what their totems may be (L. Parker, The Euahlayi, p. 53). 
1 See Bk. Ill, ch. ii. § 2. 

134 Elementary Forms of Religious Life 


We must now determine the place of man in the scheme of 
religious things. 

By the force of a whole group of acquired habits and of language 
itself, we are inclined to consider the common man, the simple 
believer, as an essentially profane being. It may well happen 
that this conception is not literally true for any religion ; ^ in 
any case, it is not applicable to totemism. Every member of 
the clan is invested with a sacred character which is not materially 
inferior to that which we just observed in the animal. This 
personal sacredness is due to the fact that the man believes that 
while he is a man in the usual sense of the word, he is also an 
animal or plant of the totemic species. 

In fact, he bears its name ; this identity of name is therefore 
supposed to imply an identity of nature. The first is not merely 
considered as an outward sign of the second ; it supposes it 
logically. This is because the name, for a primitive, is not merely 
a word or a combination of sounds ; it is a part of the being, and 
even something essential to it. A member of the Kangaroo clan 
calls himself a kangaroo ; he is therefore, in one sense, an animal 
of this species. " The totem of any man," say Spencer and 
Gillen, " is regarded as the same thing as himself ; a native once 
said to us when we were discussing the matter with him, ' That 
one,' pointing to his photograph which we had taken, ' is the 
same thing as me : so is a kangaroo ' (his totem)." ^ So each 
individual has a double nature : two beings coexist within him, 
a man and an animal. 

In order to give a semblance of intelligibility to this duality, 
so strange for us, the primitive has invented myths which, it is 
true, explain nothing and only shift the difficulty, but which, 
by shifting it, seem at least to lessen the logical scandal. With 
slight variations of detail, all are constructed on the same plan : 
their object is to establish genealogical connections between the 
man and the totemic animal, making the one a relative of the 
other. By this common origin, which, by the way, is represented 
in various manners, they believe that they account for their 
common nature. The Narrinyeri, for example, have imagined 
that certain of the first men had the power of transforming 

* Perhaps there is no religion which makes man an exclusively profane being. 
For the Christian, the soul which each of us has within him and which constitutes 
the very essence of our being, has something sacred about it. We shall see that 
this conception of the soul is as old as religious thought itself. The place of man 
in the hierarchy of sacred things is more or less elevated. 

* Nat. Tr., p. 202. 

Totemic Beliefs 135 

themselves into beasts. ^ Other AustraUan societies place at the 
beginning of humanity either strange animals from which the 
men were descended in some unknown way,^ or mixed beings, 
half-way between the two kingdoms,^ or else unformed creatures, 
hardly representable, deprived of all determined organs, and 
even of all definite members, and the different parts of whose 
bodies were hardly outlined.^ Mythical powers, sometimes 
conceived under the form of animals, then intervened and made 
men out of these ambiguous and innumerable beings which 
Spencer and Gillen say represent " stages in the transformation 
of animals and plants into human beings." ^ These transforma- 
tions are represented to us under the form of violent and, as it 
were, surgical operations. It is under the blows of an axe or, if 
the operator is a bird, blows of the beak, that the human indi- 
vidual was carved out of this shapeless mass, his members 
separated from each other, his mouth opened and his nostrils 
pierced.® Analogous legends are found in America, except that 
owing to the more highly developed mentality of these peoples, 
the representations which they employ do not contain confusions 
so troublesome for the mind. Sometimes it is a legendary 
personage who, by an act of his power, metamorphosed the 
animal who gives its name to the clan into a man.' Sometimes 
the myth attempts to explain how, by a series of nearly natural 
events and a sort of spontaneous evolution, the animal trans- 
formed himself little by little, and finally took a human form.^ 

* Taplin, The Narrinyeri, pp. 59—61. 

* Among certain clans of the Warramunga, for example {Nor. Tr., p. 162). 

' Among the Urabunna {Nor. Tr., p. 147). Even when they tell us that the 
first beings were men, these are really only semi-human, and have an animal 
nature at the same time. This is the case with certain Unmatjera {ibid., pp. 153- 
154). Here we find ways of thought whose confusion disconcerts us, but which 
must be accepted as they are. We would denature them if we tried to introduce 
a clarity that is foreign to them (cf. Nat. Tr., p. 119). 

* Among the Arunta {Nat. Tr., pp. 388 ff.) ; and among certain Unmatjera 
{Nor. Tr., p. 153). 

'^ Nat. Tr., p. 389. Cf. Strehlow, I, pp. 2-7. 

* Nat.. Tr.. p. 389 ; Strehlow, I, pp. 2 ff . Undoubtedly there is an echo of the 
initiation rites in this mythical theme. The initiation also has the object of 
making the young man into a complete man, and on the other hand, it also 
implies actual surgical operations (circumcision, sub-incision, the extraction of 
teeth, etc.). The processes which served to form the first men would naturally 
be conceived on the same model. 

' This the case with the nine clans of the Moqui (Schoolcraft, Indian Tribes, 
IV, p. 86), the Grain clan among the Ojibway (Morgan, yincient Society, p. 180), 
and the Nootka clans (Boas, VI th Rep. on the N.IV. Tribes of Canada, p. 43), etc. 

* It is thus that the Turtle clan of the Iroquois took form. A group of turtles 
had been forced to leave the lake where they dwelt and seek another home. 
One of them, which was larger than the others, stood this exercise very badly 
owing to the heat. It made such violent efforts that it got out of its shell. The 
process of transformation, being once commenced, went on by itself and the 
turtle finally became a man who was the ancestor of the clan (Erminnie A. Smith, 

136 Elementary Forms of Religious Life 

It is true that there are societies (the Haida, Thnkit, Tsim- 
shian) where it is no longer admitted that man was born of an 
animal or plant ; but the idea of an affinity between the animals 
of the totemic species and the members of the clan has survived 
there nevertheless, and expresses itself in myths which, though 
differing from the preceding, still retain all that is essential in 
them. Here is one of the fundamental themes. The ancestor 
who gives his name to the clan is here represented as a human 
being, but who, in the course of various wanderings, has been 
led to live for a while among the fabulous animals of the very 
species which gave the clan its name. As the result of this inti- 
mate and prolonged connection, he became so like his new 
companions that when he returned to^ men, they no longer 
recognized him. He was therefore given the name of the animal 
which he resembled. It is from his stay in this mythical land 
that he brought back the totemic emblem, together with the 
powers and virtues believed to be attached to it.^ Thus in this 
case, as in the others, men are believed to participate in the 
nature of the animal, though this participation may be conceived 
in slightly different forms. ^ 

So man also has something sacred about him. Though diffused 

The Myths of the Iroquois, Ilnd Report, p. 77). The Crab clan of the Choctaw 
was formed in a similar manner. Some men surprised a certain number of crabs 
that lived in the neighbourhood, took them home with them, taught them to talk 
and to walk, and finally adopted them into their society (Catlin, North American 
Indians, II, p. 128). 

^ For example, here is a legend of the Tsimshian. In the course of a hunt, an 
Indian met a black bear which took him to its home, and taught him to catch 
salmon and build canoes. The man stayed with the bear for two years, and then 
returned to his native village. But the people were afraid of him, because he 
was just like a bear. He could not talk or eat anything except raw food. Then 
he was rubbed with magic herbs and gradually regained his original form. After 
that, whenever he was in trouble, he called upon his bear friends, who came to 
aid him. He built a house and painted a bear on the foundation. His sister 
made a blanket for the dance, upon which a bear was designed. That is why the 
descendants of this sister had the bear as their emblem (Boas, Kwakiutl, p. 323. 
Cf. Vth Rep. on the N .W. Tribes of Canada, pp. 23, 29 ff. ; Hill Tout, Report on the 
Ethnolof^y of the Statlumh of British Columbia, in J. A. I., 1905, XXXV, p. 150). 

Thus wc sec the inconveniences in making this mystical relationship between 
the man and the animal the distinctive characteristic of totemism, as M. Van 
Gennep proposes (Totémisme et méthode comparative, in Revue de l'histoire des 
religions. Vol. LVIII, July, 1908, p. 55). This relationship is a mythical repre- 
sentation of otherwise profound facts ; but it may be omitted without causing 
the disappearance of the essential traits of totemism. Undoubtedly there are 
always close bonds between the people of the clan and the totemic animal, but 
these are not necessarily bonds of blood-relationship, though they are frequently 
conceived in this form. 

* There are also some Tlinkit myths in which the relationship of descent 
between the man and the animal is still more carefully stated. It is said that the 
clan is descended from a mixed union, if we may so speak, that is to say, one 
where either the husband or the wife was an animal of the species whose name the 
clan bears (sec Swanton, Social Condition, Beliefs, etc., of the Tlinkit Indians, 
XXVIth Rep., pp. 415-418). 

Totemic Beliefs ^yj 

into the whole organism, this characteristic is especially apparent 
in certain privileged places. There are organs and tissues that 
are specially marked out : these are particularly the blood and 
the hair. 

In the first place, human blood is so holy a thing that in the 
tribes of Central Australia, it frequently serves to consecrate 
the most respected instruments of the cult. For example, in 
certain cases, the nurtunja is regularly anointed from top to 
bottom with the blood of a man.^ It is upon ground all saturated 
with blood that the men of the Emu, among the Arunta, trace 
their sacred images.'^ We shall presently see that streams of 
blood are poured upon the rocks which represent the totemic 
animals and plants.'- There is no religious ceremony where blood 
does not have some part to play.* During the initiation, the 
adults open their veins and sprinkle the novice with their blood ; 
and this blood is so sacred a thing that women may not be 
present while it is flowing ; the sight of it is forbidden them, 
just as the sight of a churinga is.^ The blood lost by a young 
initiate during the very violent operations he must undergo 
has very particular virtues : it is used in various ceremonies.* 
That which flows during the sub-incision is piously kept by the 
Arunta and buried in a place upon which they put a piece of 
wood warning passers-by of the sacredness of the spot ; no 
woman should approach it.' The religious nature of blood also 
explains the equal importance, religiously, of the red ochre, 
which is very frequently employed in ceremonies ; they rub the 
churinga with it and use it in ritual decorations.^ This is due to 
the fact that because of its colour, it is regarded as something 
kindred to blood. Many deposits of red ochre which are found 
in the Arunta territory are even supposed to be the coagulated 
blood which certain heroines of the mythical period shed on to 
the soil.^ 

Hair has similar properties. The natives of the centre wear 
belts made of human hair, whose religious functions we have 
already pointed out : they are also used to wrap up certain 

1 Nal. Tr., p. 284. « Ibid., p. 179. 

» See Bk. III. ch. ii. Cf. Nat. Tr., pp. 184, 201. 

* Ibid., pp. 204, 262, 284. 

* Among the Dieri and the Parnkalla. See Howitt, Nat. Tr., pp. 658, 661, 
668, 669-671. 

* Among the Warramunga, the blood from the circumcision is drunk by the 
mother {Nor. Tr., p. 352). Among the Binbinga, the blood on the knife which 
was used in the sub-incision must be licked off by the initiate {ibid., p. 368). In 
general, the blood coming from the genital organs is regarded as especially sacred 
{Nat. Tr.. p. 464 ; Nor. Tr., p. 598). 

' Nat. Tr., p. 268. » Ibid., pp. 144, 568. 

' Ibid., pp. 442, 464. This myth is quite common in Australia. 

138 Elementary Forms of Religious Life 

instruments of the cult.^ Does one man loan another one of his 
churinga ? As a sign of acknowledgment, the second makes 
a present of hair to the first ; these two sorts of things are there- 
fore thought to be of the same order and of equivalent value. ^ 
So the operation of cutting the hair is a ritual act, accompanied 
by definite ceremonies : the individual operated upon must 
squat on the ground, with his face turned in, the direction of the 
place where the fabulous ancestors from which the clan of his 
mother is believed to be descended, are thought to have camped.^ 
For the same reason, as soon as a man is dead, they cut his hair 
off and put it away in some distant place, for neither women 
nor the non-initiated have the right of seeing it : it is here, far 
from profane eyes, that the belts are made.^ 

Other organic tissues might be mentioned which have similar 
properties, in varying degrees : such are the whiskers, the fore- 
skin, the fat of the liver, etc.^ But it is useless to multiply 
examples. Those already given are enough to prove that there 
is something in man which holds profane things at a distance 
and which possesses a religious power ; in other words, the 
human organism conceals within its depths a sacred principle, 
which visibly comes to the surface in certain determined cases. 
This principle does not differ materially from that which causes 
the religious character of the totem. In fact, we have just seen 
that the different substances in which it incarnates itself especially 
enter into the ritual composition of the objects of the cult 
(nurtunja, totemic designs), or else are used in the anointings 
whose object is to renew the virtues either of the churinga or of 
the sacred rocks ; they are things of the same species. 

Sometimes the religious dignity which is inherent in each 
member of the clan on this account is not equal for all. Men 
possess it to a higher degree than women ; in relation to them, 
women are like profane beings.^ Thus, every time that there is 

1 Nat. Tr., p. 627. * Ibid., p. 466. 

' Ibid. It is believed that if all these formalities are not rigorously observed, 
grave calamities will fall upon the individual. 

* Nat. Tr., p. 538 ; Nor. Tr., p. 604. 

- After the foreskin has been detached by circumcision, it is sometimes hidden, 
just like the blood ; it has special virtues ; for example, it assures the fecundity 
of certain animal and vegetable species (Nor. Tr., pp. 353 f.). The whiskers are 
mixed with the hair, and treated as such {ibid., pp. 604, 544). They also play 
a part in the myths [ibid., p. 158). As for the fat, its sacred character is shown 
by the use made of it in certain funeral rites. 

* This is not saying that the woman is absolutely profane. In the myths, 
at least among the Arunta, she plays a religious rôle much more important 
than she does in reality {Nat. Tr., pp. 195 f.). Even now she takes part in certain 
initiation rites. Finally, her blood has religious virtues (see Nat. Tr., p. 464; 
of. La prohibition de l'inceste et ses origines, Année Sociol., I, pp. 41 ff.). 

It is upon this complex situation of the woman that the exogamic restrictions 
depend. We do not speak of them here because they concern the problem of 
domestic and matrimonial organization more directly than the present one. 

Totemic Beliefs 139 

an assembly, either of the totemic group or of the tribe, the 
men have a separate camp, distinct from that of the women, 
and into which these latter may not enter : they are separated 
off.^ But there are also differences in the way in which men 
are marked with a religious character. The young men not yet 
initiated are wholly deprived of it, since they are not admitted 
to the ceremonies. It is among the old men that it reaches its 
greatest intensity. They are so very sacred that certain things 
forbidden to ordinary people are permissible for them : they may 
eat the totemic animal more freely and, as we have seen, there 
are even some tribes where they are freed from all dietetic re- 

So we must be careful not to consider totemism a sort of 
"nriimal worship. The attitude of a man towards the animals or 
plants whose name he bears is not at all that of a believer to- 
wards his god, for he belongs to the sacred world himself. Their 
relations are rather those of two beings who are on the same 
level and of equal value. . The most that can be said is that 
in certain cases, at least, the animal seems to occupy a slightly 
more elevated place in the hierarchy of sacred things. It is 
because of this that it is sometimes called the father or the 
grandfather of the men of the clan, which seems to show that 
they feel themselves in a state of moral dependence in regard 
to it. 2 But in other, and perhaps even more frequent cases, it 
happens that the expressions used denote rather a sentiment 
of equality. The totemic animal is called the friend or the 
elder brother of its human fellows.' Finally, the bonds which 
exist between them and it are much more like those which unite 
the members of a single family ; the animals and the men are 
made of the same flesh, as the Buandik say.* On account of this 
kinship, men regard the animals of the totemic species as kindly 
associates upon whose aid they think they can rely. They 
call them to their aid ^ and they come, to direct their blows in 
the hunt and to give warning of whatever dangers there may be.* 

* Nat. Tr., p. 460. 

* Among the Wakclbura, according to Hewitt, p. 146 ; among the Bechuana. 
according to Casalis, Basoutos, p. 221. 

* Among the Buandik and Kurnai (Howitt, ibid.) ; among the Arunta 
(Strehlow, II, p. 58). 

* Howitt, ibid. 

* In the TuUy River district, says Roth {Superstition, Magic and Medicine, in 
North Queensland Ethnography, No. 5, § 74), as an individual goes to sleep or gets 
up in the morning, he pronounces in a rather low voice the name of the animal 
after which he is named himself. The purpose of this practice is to make the 
man clever or lucky in the hunt, or be forewarned of the dangers to which he may 
be exposed from this animal. For example, a man who has a species of serpent 
as his totem is protected from bites if this invocation has been made regularly. 

" Taplin, Narrinyeri, p. 64 ; Howitt, Nat. Tr., p. 147 ; Roth, loc. cit. 

140 Elementary Forms of Religious Life 

In return for this, men treat them with regard and are never 
cruel to them ; ^ but these attentions in no way resemble a cult. 
Men sometimes even appear to have a mysterious sort of 
property-right over their totems. The prohibition against 
killing and eating them is applied only to members of the clan, 
of course ; it could not be extended to other persons without 
making life practically impossible. If, in a tribe like the Arunta, 
where there is such a host of different totems, it were forbidden 
to eat, not only the animal or plant whose name one bears, but 
also all the animals and all the plants which serve as totems 
to other clans, the sources of food would be reduced to nothing. 
Yet there are tribes where the consumption of the totemic plant 
or animal is not allowed without restrictions, even to foreigners. 
Among the Wakelbura, it must not take place in the presence of 
men of this totem. 2 In other places, their permission must be 
"given. For example, among the Kaitish and the Unmatjera, 
whenever a man of the Emu totem happens to be in a place 
occupied by a grass-seed clan, and gathers some of these seed, 
before eating them he must go to the chief and say to him, 
" I have gathered these seeds in your country." To this the 
chief replies, " All right ; you may eat them." But if the Emu 
man ate them before demanding permission, it is believed that 
he would fall sick and run the risk of dying. ^ There are even 
cases where the chief of the group must take a little of the food 
and eat it himself : it is a sort of payment which must be made.* 
For the same reason, the churinga gives the hunter à certain 
power over the corresponding animal : by rubbing his body 
with a Euro churinga, for example, a man acquires a greater 
chance of catching euros. ^ This is the proof that the fact of 
participating in the nature of a totemic being confers a sort of 
eminent right over this latter. Finally, there is one tribe in 
northern Queensland, the Karingbool, where the men of the 
totem are the only ones who have a right to kill the animal or, 
if the totem is a tree, to peel off its bark. Their aid is indispensable 
to all others who want to use the flesh of this animal or the wood 
of this tree for their own personal ends.® So they appear as 
proprietors, though it is quite evidently over a special sort of 
property, of which we find it hard to form an idea. 

* Strehlow. II, p. 58. « Howitt, p. 148. 
' Nor. Tr/, pp. 159-160. * Ibid. 

* Ibid., p. 225 ; Nat. Tr., pp. 202, 203. 

* A. L. P. Cameron, On Two Queensland Tribes, in Science of Man, Australasian 
Anthropological Journal, 1904, VII, 28, col. i. 



The Cosmological Syste?n of Totemism and the Idea of Class 

WE are beginning to see that totemism is a much more 
complex reUgion than it first appeared to be. We have 
already distinguished three classes of things which it recognizes 
as sacred, in varying degrees : the totemic emblem, the animal 
or plant whose appearance this emblem reproduces, and the 
members of the clan. However, this list is not yet complete. 
In fact, a religion is not merely a collection of fragmentary 
beliefs in regard to special objects like those we have just been 
discussing. To a greater or less extent, all known religions have 
been systems of ideas which tend to embrace the universality of 
things, and to give us a complete representation of the world. 
If totemism is to be considered as a religion comparable to the 
others, it too should offer us a conception of the universe. As 
a matter of fact, it does satisfy this condition. 


The fact that this aspect of totemism has generally been 
neglected is due to the too narrow notion of the clan which has 
been prevalent. Ordinarily it is regarded as a mere group of 
hiunan beings. Being a simple subdivision of the tribe, it seems 
that like this, it is made up of nothing but men. But in reason- 
ing thus, we substitute our European ideas for those which 
the primitive has of man and of society. For the Australian, 
things themselves, everything which is in the universe, are a 
part of the tribe ; they are constituent elements of it and, so 
to speak, regular members of it ; just like men, they have 
a determined place in the general scheme of organization of the 
society. " The South Australian savage," says Fison, " looks 
upon the universe as the Great Tribe, to one of whose divisions 
he himself belongs ; and all things, animate and inanimate, 
which belong to his class are parts of the body corporate whereof 
he himself is a part."^ As a consequence of this principle, 
whenever the tribe is divided into two phratries, all known 
things are distributed between them. " All nature," says 
Palmer, in speaking of the Bellinger River tribe, " is also divided 
into class [phratry] names. . . . The sun and moon and stars 

* Kamilaroi and Kurnai, p. 170. 


Elementary Forms of Religious Life 

are said ... to belong to classes [phratries] just as the blacks 
themselves. "1 The Port Mackay tribe in Queensland has two 
phratries with the names Yungaroo and Wootaroo, as do the 
neighbouring tribes. Now as Bridgmann says, " all things, 
animate and inanimate, are divided by these tribes into two 
classes, named Yungaroo and Wootaroo." ^ Nor does the classifi- 
cation stop here. The men of each phratry are distributed 
among a certain number of clans ; likewise, the things attributed 
to each phratry are in their turn distributed among the clans 
of which the phratry is composed. A certain tree, for example, 
will be assigned to the Kangaroo clan, and to it alone ; then, 
just like the human members of the clan, it will have the Kangaroo 
as totem ; another will belong to the Snake clan ; clouds will 
be placed under one totem, the sun under another, etc. All"! 
known things will thus be arranged in a sort of tableau or syste- 
matic classification embracing the whole of nature. j 

We have given a certain number of these classifications else- 
where ;^ at present we shall confine ourselves to repeating a few 
of these as examples. One of the best known of these is the one 
found in the Mount Gambler tribe. This tribe includes two 
phratries, named respectively the Kumite and the Kroki ; each 
of these, in its turn, is subdivided into five clans. Now " every- 
thing in nature belongs to one or another of these ten clans " ;"* 
Fison and Howitt say that they are all " included " within it. 
In fact, they are classified under these ten totems just like 
species in their respective classes. This is well shown by the 
following table based on information gathered by Curr and by 
Fison and Howitt.^ 




Pelican . 



Black cockatoo 

A non-poisonous snake . 


An edible root 

A white crestless cockatoo 

Things classed in each clan. 
f Smoke, honeysuckle, certain 
trees, etc. 

Blackwood-trees, dogs, fire, 
frost, etc. 

Rain, thunder, lightning, clouds, 
hail, winter, etc. 

The stars, the moon, etc. 

Fish, seal, eel, the stringybark- 
tree, etc. 

Duck, crayfish, owls, etc. 

Bustard, quail, a small kanga- 
roo, etc. 

Kangaroo, the summer, the sun, 
wind, the autumn, etc. 
Details are lacking for the fourth and fifth Kroki clans. 

* Notes on some Australian Tribes, J. A. I., XIII, p. 300. 

* In Curr, Australian Race, 111, p. 45 ; Brough Smyth, The Aborigines oj 
Victoria, I, p. 91 ; Fison and Howitt, Kamilaroi and Kurnai, p. 168. 

* Durkheim and Mauss, De quelques formes primitives de classification, in 
Année Sociol., VI, pp. i ff. * Curr, III, p. 461. 

^ Curr and Fison were both informed by the same person, D. S. Stewart. 

Totemic Beliefs 143 

The list of things attached to each clan is quite incomplete ; 
Curr himself warns us that he has limited himself to enumerating 
some of them. But through the work of Mathews and of Howitt^ 
we have more extended information to-day on the classification 
adopted by the Wotjobaluk tribe, which enables us to under- 
stand better how a system of this kind is able to include the whole 
universe, as known to the natives. The Wotjobaluk also are 
divided into two phratries called Gurogity and Gumaty (Kro- 
kitch and Gamutch according to Howitt^) ; not to prolong this 
enumeration, we shall content ourselves with indicating, after 
Mathews, the things classed in some of the clans of the Gurogity 

In the clan of the Yam are classified the plain-turkey, the 
native cat, the mopoke, the dyim-dyim owl, the malice hen, the 
rosella parrot, the peewee. 

In the MusseP clan are the grey emu, the porcupine, the cur- 
lew, the white cockatoo, the wood-duck, the mallee lizard, the 
stinking turtle, the flying squirrel, the ring-tail opossum, the 
bronze- wing pigeon, the wijuggla. 

In the Sun clan are the bandicoot, the moon, the kangaroo-rat, 
the black and white magpies, the opossum, the ngiirt hawk, the 
gum-tree grub, the wattle-tree grub, the planet Venus. 

In the clan of the Warm Wind* are the grey-headed eagle- 
hawk, the carpet snake, the smoker parrot, the shell parrot, the 
munakan hawk, the dikkomur snake, the ring-neck parrot, the 
mirudai snake, the shingle-back lizard. 

If we remember that there are many other clans (Howitt 
names twelve and Mathews fourteen and adds that his list is 
incomplete^), we will understand how all the things in which the 
native takes an interest find a natural place in these classifications. 

Similar arrangements have been observed in the most diverse 

* Mathews, Aboriginal Tribes of N.S. Wales and Victoria, in Journal and 
Proceedings of the Royal Society of N.S. Wales, XXXVIII, pp. 287 f. ; Howitt. 
Nat. Tr., p. 121. 

* The feminine form of the names given by Mathews is Gurogigurk and 
Gamatykurk. These are the forms which Howitt reproduces, with a slightly 
different orthography. The names are also equivalent to those used by the 
Mount Gambler tribe (Kumite and Kroki). 

' The native name of this clan is Dyàlup, which Mathews does not translate. 
This word appears to be identical with Jallup, by which Howitt designates a 
sub-clan of the same tribe, and which he translates " mussel." That is why we 
think we can hazard this translation. 

* This is the translation of Howitt ; Mathews renders the word (Wartwurt, 
" heat of the midday sun." 

° The tables of Mathews and Howitt disagree on many important points. It 
even seems that clans attributed by Howitt to the Kroki phratry are given to 
the Gamutch phratry by Mathews, and inversely. This proves the great diffi- 
culties that these observations present. But these differences are without 
interest for our present question. 

144 Elementary Forms of Religious Life 

parts of the Australian continent ; in South Austraha, in Vic- 
toria, and in New South Wales (among the Euahlayi^) ; very 
clear traces of it are found in the central tribes. '-^ In Queensland, 
where the clans seem to have disappeared and where the matri- 
monial classes are the only subdivisions of the phratry, things 
are divided up among these classes. Thus, the Wakelbura are 
divided into two phratries, Mallera and Wutaru ; the classes 
of the first are called Kurgilla and Banbe, those of the second, 
Wungo and Obu. Now to the Banbe belong the opossum, the 
kangaroo, the dog, honey of little bees, etc. ; to the Wungo are 
attributed the emu, the bandicoot, the black duck, the black 
snake, the brown snake ; to the Obu, the carpet snake, the 
honey of stinging bees, etc. ; to the Kurgilla, the porcupine, the 
turkey of the plains, water, rain, fire, thunder, etc.^ 

This same organization is found among the Indians of North 
America. The Zuhi have a system of classification which, in its 
essential lines, is in all points comparable to the one we have 
just described. That of the Omaha rests on the same principles 
as that of the Wotjobaluk.* An echo of these same ideas sur- 
vives even into the more advanced societies. Among the Haida, 
all the gods and mythical beings who are placed in charge of 
the different phenomena of nature are classified in one or the 
other of the two phratries which make up the tribe just like men ; 
some are Eagles, the others, Crows. ^ Now the gods of things are 
only another aspect of the things which they govern.® This 
mythological classification is therefore merely another form of 
the preceding one. So we may rest assured that this way of 
conceiving the world is independent of all ethnic or geographic 
particularities ; and at the same time it is clearly seen to be 
closely united to the whole system of totemic beliefs. 


f" In the paper to which we have already made allusion several 
times, we have shown what light these facts throw upon the 
way in which the idea of kind or class was formed in humanity. 
In fact, these systematic classifications are the first we meet with 

^ Mrs. Langloh Parker, The Euahlayi Tribe, pp. 12 ff. 

* The facts will be found below. 

' Carr, III, p. 27. Cf. Howitt, Nat. Tr., p. 112. We are merely mentioning 
the most characteristic facts. For details, one may refer to the memoir already 
mentioned on Les classifications primitives. 

* Ibid., pp. 34 fï. ^ Swanton, The Haida, pp. 13-14, 17, 22. 

" This is especially clear among the Haida. Swanton says that with them 
every animal has two aspects. First, it is an ordinary animal to be hunted and 
eaten ; but it is also a supernatural being in the animal's form, upon which men 
depend. The mythical beings corresponding to cosmic phenomena have the same 
ambiguity (Swanton, ibid., 16, 14, 25). 

Totemic Beliefs 145 

in history, and we have just seen that they are modelled upon 
the social organization, or rather that they have taken the forms 
of society as their framework. It is the phratries which have 
served as classes, and the clans as species. It is because men 
were organized that they have been able to organize things, for 
in classifying these latter, they limited themselves to giving them 
places in the groups they formed themselves. And if these 
different classes of things are not merely put next to each other, 
but are arranged according to a unified plan, it is because the 
social groups with which they commingle themselves are unified 
and, through their union, form an organic whole, the tribe. 
The unity of these first logical systems merely reproduces the 
unity of the society. Thus we have an occasion for verifying the 
proposition which we laid down at the commencement of this 
work, and for assuring ourselves that the fundamental notions of 
the intellect, the essential categories of thought, may be the 
product of social factors. The above-mentioned facts show 
clearly that this is the case with the very notion of category itself, j 

However, it is not our intention to deny that the individual! 
intellect has of itself the power of perceiving resemblances 
between the different objects of which it is conscious. Quite on 
the contrary, it is clear that even the most primitive and simple 
classifications presuppose this faculty. The Australian does not 
place things in the same clan or in different clans at random. 
For him as for us, similar images attract one another, while 
opposed ones repel one another, and it is on the basis of these 
feehngs of afiinity or of repulsion that he classifies the corre- 
sponding things in one place or another. ^ 

There are also cases where we are able to perceive the reasons 
which inspired this. The two phratries were very probably the 
original and fundamental bases for these classifications, which 
were consequently bifurcate at first. Now, when a classification 
is reduced to two classes, these are almost necessarily conceived 
as antitheses ; they are used primarily as a means of clearly 
separating things between which there is a very marked contrast. 
Some are set at the right, the others at the left. As a matter 
of fact this is the character of the Australian classifications. If 
the white cockatoo is in one phratry, the black one is in the 
other ; if the sun is on one side, the moon and the stars of night 
are on the opposite side.^ Very frequently the beings which 
serve as the totems of the two phratries have contrary colours. 2 

» See above, p. 142. This is the case among the Goumditch-mara (Howitt, 
Nat. Tr.. p. 124). in the tribes studied by Cameron near the Dead Lake, and 
among the Wotjobaluk {ibid., pp. 125, 250). 

^ J. Mathews, Two Representative Tribes, p. 139; Thomas, Kinship and 
Marriage, pp. 53 f. 

14b Elementary Forms of Religious Life 

These oppositions are even met with outside of Austraha. Where 
one of the phratries is disposed to peace, the other is disposed to 
war ;^ if one has water as its totem, the other has earth. ^ This 
is undoubtedly the explanation of why the two phratries have 
frequently been thought of as naturally antagonistic to one 
another. They say that there is a sort of rivalry or even a con- 
stitutional hostility between them.^ This opposition of things 
has extended itself to persons ; the logical contrast has begotten 
a sort of social conflict.* 

It is also to be observed that within each phratry, those things 
have been placed in a single clan which seem to have the greatest 
affinity with that serving as totem. For example, the moon has 
been placed with the black cockatoo, but the sun, together with 
the atmosphere and the wind, with the white cockatoo. Or again, 
to a totemic animal has been united all that serves him as food,^ 
as well as the animals with which he has the closest connection.® 
Of course, we cannot always understand the obscure psychology 
which has caused many of these connections and distinctions, 
but the preceding examples are enough to show that a certain 
intuition of the resemblances and differences presented by things 

* Among the Osage, for example (see Dorsey, Siouan Sociology, in XVth Rep., 

PP- 233 ff- 

* At Mabuiag, an island in Torres' Strait (Haddon, Head Hunters, p. 132), 
the same opposition is found between the two phratries of the Arunta : one 
includes the men of a water totem, the other those of earth (Strehlow, I, p. 6). 

* Among the Iroquois there is a sort of tournament between the two phratries 
(Morgan, Ancient Society, p. 94). Among the Haida, says Swanton, the members 
of the two phratries of the Eagle and the Crow " are frequently considered 
as avowed enemies. Husband and wife (who must be of different phratries) do 
not hesitate to betray each other " (The Haida, p. 62). In Australia this hostility 
is carried into the myths. The two animals serving the phratries as totems are 
frequently represented as in a perpetual war against each other (see J. Mathews, 
Eaglehawk and Crow, a study of Australian Aborigines, pp. 14 fi.). In games, 
each phratry is the natural rival of the other (Howitt, Nat. Tr., p. 770). 

* So Thomas has wrongly urged against our theory of the origin of the 
phratries its inability to explain their opposition (Kinship and Marriage, p. 69). 
We do not believe that it is necessary to connect this opposition to that of the 
profane and the sacred (see Hertz, La prééminence de la main droite, in the 
Revue Philosophique, Dec, 1909, p. 559). The things of one phratry are not 
profane for the other ; both are a part of the same religious system (see below, 

P- 155)- 

' For example, the clan of the Tea-tree includes the grasses, and consequently 
herbivorous animals (see Kamilaroi and Kurnai, p. 169). This is undoubtedly 
the explanation of a particularity of the totemic emblems of North America 
pointed out by Boas. " Among the Tlinkit," he says, " and all the other tribes 
of the coast, the emblem of a group includes the animals serving as food to the 
one whose name the group bears " (Fifth Rep. of the Committee, etc., British 
Association for the Advancement of Science, p. 25). 

* Thus, among the Arunta, frogs are connected with the totem of the gum- 
tree, because they are frequently found in the cavities of this tree ; water is 
related to the water-hen ; with the kangaroo is associated a sort of parrot 
frequently seen flying about this animal (Spencer and Gillen, Na^ Tr.,pp. 146-147, 

Totemic Beliefs 147 

has played an important part in the genesis of these classifica- 
tions. ^ 

But the feeling of resemblances is one thing and the idea of ' 
class is another. The class is the external framework of which 
objects perceived to be similar form, in part, the contents. Now 
the contents cannot furnish the frame into which they fit. They 
are made up of vague and fluctuating images, due to the super- 
imposition and partial fusion of a determined number of individual 
images, which are found to have common elements ; the frame- 
work, on the contrary, is a definite form, with fixed outlines, but 
which may be applied to an undetermined number of things, per- 
ceived or not, actual or possible. In fact, every class has possi- 
bilities of extension which go far beyond the circle of objects 
which we know, either from direct experience or from resemblance. 
This is why every school of thinkers has refused, and not with 
good reason, to identify the idea of class with that of a generic 
image. The generic image is only the indistinctly-bounded 
residual representation left in us by similar representations, when 
they are present in consciousness simultaneously ; the class is 
a logical symbol by means of which we think distinctly of these 
similarities and of other analogous ones. Moreover, the best 
proof of the distance separating these two notions is that an 
animal is able to form generic images though ignorant of the 
art of thinking in classes and species. -' 

The idea of class is an instrument of thought which has ob- ~\ 
viously been constructed by men. But in constructing it, we 
have at least had need of a model ; for how could this idea ever 
have been born, if there had been nothing either in us or around 
us which was capable of suggesting it to us ? To reply that it 
was given to us a priori is not to reply at all ; this lazy man's 
solution is, as has been said, the death of analysis. But it is hard 
to see where we could have found this indispensable model except 
in the spectacle of the collective life. In fact, a class is not an 
ideal, but a clearly defined group of things between which in- 
ternal relationships exist, similar to those of kindred. Now the 
only groups of this sort known from experience are those formed 
by men in associating themselves. Material things may be able 
to form collections of units, or heaps, or mechanical assemblages 
with no internal unity, but not groups in the sense we have 
given the word. A heap of sand or a pile of rock is in no way 
comparable to that variety of definite and organized society 
which forms a class. In all probability, we would never have 
thought of uniting the beings of the universe into homogeneous 
groups, called classes, if we had not had the example of human 
societies before our eyes, if we had not even commenced by making 

148 Elementary Forms of Religious Life 

things themselves members of men's society, and also if human 

^ groups and logical groups had not been confused at first. ^ 

r-- It is also to be borne in mind that a classification is a system 

whose parts are arranged according to a hierarchy. There are 

dominating members and others which are subordinate to the 

first ; species and their distinctive properties depend upon 

classes and the attributes which characterize them ; again, the 

different species of a single class are conceived as all placed on 

the same level in regard to each other. Does someone prefer 

to regard them from the point of view of the understanding ? 

Then he represents things to himself in an inverse order : he 

puts at the top the species that are the most particularized and 

the richest in reality, while the types that are most general and 

the poorest in qualities are at the bottom. Nevertheless, all are 

represented in a hierarchic form. And we must be careful not 

to believe that the expression has only a metaphorical sense here : 

there are really relations of subordination and co-ordination, the 

establishment of which is the object of all classification, and men 

would never have thought of arranging their knowledge in this 

way if they had not known beforehand what a hierarchy was. 

/ But neither the spectacle of physical nature nor the mechanism 

i of mental associations could furnish them with this knowledge. 

f — The hierarchy is exclusively a social affair. It is only in society 

that there are superiors, inferiors and equals. Consequently, 

even if the facts were not enough to prove it, the mere analysis 

of these ideas would reveal their origin. We have taken them 

from society, and projected them into our conceptions of the 

world. It is society that has furnished the outlines which logical 

! thought has filled in. 


But these priipitive classifications have a no less direct in- 
terest for the ori^ns of religious thought. 

They imply that all the things thus classed in a single clan or 
a single phratry are closely related both to each other and to the 
thing serving as the totem of this clan or phratry. When an Austra- 
lian of the Port Mackay tribe says that the sun, snakes, etc., are of 
the Yungaroo phratry, he does not mean merely to apply a com- 
mon, but none the less a purely conventional, nomenclature to 

^ One of the signs of this primitive lack of distinction is that territorial bases 
are sometimes assigned to the classes just as to the social divisions with which 
they were at first confounded. Thus, among the Wotjobaluk in Australia and 
the Zufii in America, things are ideally distributed among' the different regions of 
space, just as the clans are. Now this regional distribution of things and that of 
the clans coincide (see De quelques formes primitives de classification, pp. 34 fif.). 
Classifications keep something of this special character even among relatively 
advanced peoples, as for example, in China [ihid., pp. 55 fif.). 

Totemic Beliefs 14g 

these different things ; the word has an objective signification for 
him. He believes that " alUgators really are Yungaroo and that 
kangaroos are Wootaroo. The sun is Yungaroo, the moon Woot- 
aroo, and so on for the constellations, trees, plants, etc."^ An 
internal bond attaches them to the group in which they are placed ; 
they are regular members of it. It is said that they belong to the 
group, 2 just exactly as the individual men make a part of it ; 
consequently, the same sort of a relation unites them to these 
latter. Men regard the things in their clan as their relatives or 
associates ; they call them their friends and think that they are 
made out of the same flesh as themselves. ^ Therefore, between 
the two there are elective affinities and quite special relations of 
agreement. Things and people have a common name, and in 
a certain way they naturally understand each other and har- 
monize with one another. For example, when a Wakelbura of 
the Mallera phratry is buried, the scaffold upon which the body 
is exposed " must be made of the wood of some tree belonging 
to the Mallera phratry."* The same is true for the branches that 
cover the corpse. If the deceased is of the Banbe class, a Banbe 
tree must be used. In this same tribe, a magician can use in his 
art only those things which belong to his own phratry ;^ since 
the others are strangers to him, he does not know how to make 
them obey him. Thus a bond of mystic sympathy unites each 
individual to those beings, whether living or not, which are asso- 
ciated with him ; the result of this is a belief in the possibility 
of deducing what he will do or what he has done from what they 
are doing. Among these same Wakelbura, when a man dreams 
that he has killed an animal belonging to a certain social division, 
he expects to meet a man of this same division the next day.* 
Inversely, the things attributed to a clan or phratry cannot be 
used against the members of this clan or phratry. Among the 
Wotjobaluk, each phratry has its own special trees. Now in 
hunting an animal of the Gurogity phratry, only arms whose 
wood is taken from trees of the other phratry may be used, and 
vice versa ; otherwise the hunter is sure to miss his aim. ' The 
native is convinced that the arrow would turn of itself and refuse, 
so to speak, to hit a kindred and friendly animal. 

^ Bridgmann, in Brough Smyth, The Aborigines of Victoria, I, p. 91. 
^ Fison and Howitt, Kamilaroi and Kurnai, p. 168 ; Howitt, Further Notes on 
the Australian Class Systems, J. A. I., XVIII, p. 60. 

' Curr, III, p. 461. This is about the Mount Gambier tribe. 

* Howitt, On some Australian Beliefs, J. A. I., XIII, p. igi, n. i. 

* Howitt, Notes on Australian Message Sticks, J. A. I., XVIII p. 326 • Further 
Notes. J.A.I., XVIII, p. 61, n. 3. 

* Curr, III, p. 28. 

' Mathews, Ethnological Notes on the Aboriginal Tribes of N.S. Wales and 
Victoria, in Journ. and Proceed, of the Royal Soc. of N.S. Wales, XXXVIII, p. 294. 

150 Elementary Forms of Religious Life 

' Thus the men of the clan and the things which are classified 
in it form by their union a sohd system, all of whose parts are 
united and vibrate sympathetically. This organization, which at 
first may have appeared to us as purely logical, is at the same 
time moral. A single principle animates it and makes its unity : 

^this is the totem. Just as a man who belongs to the Crow clan 
has within him something of this animal, so the rain, since it is 
of the same clan and belongs to the same totem, is also necessarily 
considered as being " the same thing as a crow " ; for the same 
reason, the moon is a black cockatoo, the sun a white cockatoo, 
every black-nut tree a pelican, etc. All the beings arranged in 
a single clan, whether men, animals, plants or inanimate objects, 
are merely forms of the totemic being. This is the meaning of the 
formula which we have just cited and this is what makes the two 
really of the same species : all are really of the same flesh in the 
sense that all partake of the nature of the totemic animal. Also, 
the qualifiers given them are those given to the totem. ^ The 
Wotjobaluk give the name Mir both to the totem and to the 
things classed with it.^ It is true that among the Arunta, where 
visible traces of classification still exist, as we shall see, difterent 
words designate the totem and the other beings placed with it ; 
however, the name given to these latter bears witness to the close 
relations which unite them to the totemic animal. It is said that 
they are its intimates, its associates, lis friends ; it is believed that 
they are inseparable from it.^ So there is a feeling that these are 
very closely related things. 

But we also know that the totemic animal is a sacred being. 
All the things that are classified in the clan of which it is the 
emblem have this same character, because in one sense, they are 
animals of the same species, just as the man is. They, too, are 
sacred, and the classifications which locate them in relation to 
the other things of the universe, by that very act give them a 
place in the religious world. For this reason, the animals or 
plants among these may not be eaten freely by the human 
members of the clan. Thus in the Mount Gambler tribe, the men 
whose totem is a certain non-poisonous snake must not merely 
refrain from eating the flesh of this snake ; that of seals, eels, etc., 
is also forbidden to them.^ If, driven by necessity, they do eat 
some of it, they must at least attenuate the sacrilege by expiatory 
rites, just as if they had eaten the totem itself.* Among the 

' Cf. Curr, III, p. 461 ; and Howitt, Nat. Tr., p. 146. The expressions 
Tooman and Wingo are applied to the one and the other. 
^ Howitt, Nat. Tr., p. 123. 

' Spencer and Gillen, Nat. Tr., pp. 447 flf. ; cf. Strehlow, III, pp. xii ff. 
* Fison and Howitt, Kamilaroi and Kurnai, p. 169. 
■' Curr, III, p. 462, 

Totemic Beliefs 151 

Euahlayi, where it is permitted to use the totem, but not to 
abuse it, the same rule is appUed to the other members of the 
clan.i Among the Arunta, the interdictions protecting the 
totemic animal extend over the associated animals ;2 and in any 
case, particular attention must be given to these latter.^ The 
sentiments inspired by the two are identical.^ 

But the fact that the things thus attached to the totem are 
not of a different nature from it, and consequently have a re- 
ligious character, is best proved by the fact that on certain 
occasions they fulfil the same functions. They are accessory or 
secondary totems, or, according to an expression now conse- 
crated by usage, they are sub-totems.^ It is constantly happening 
in the clans that under the influence of various sympathies, par- 
ticular affinities are forming, smaller groups and more limited 
associations arise, which tend to lead a relatively autonomous 
life and to form a new subdivision like a sub-clan within the 
larger one. In order to distinguish and individualize itself, this 
sub-clan needs a special totem or, consequently, a sub-totem.^ 
Now the totems of these secondary groups are chosen from 
among the things classified under the principal totem. So they 
are always almost totems and the slightest circumstance is 
enough to make them actually so. There is a latent totemic 
nature in them, which shows itself as soon as conditions permit 

1 Mrs. Parker, The Euahlayi Tribe, p. 20. 

* Spencer and Gillen, Nor. Tr., p. 151 ; Nat. Tr., p. 447 ; Strehlow, III, 
p. xii. 

^ Spencer and Gillen, Nat. Tr., p. 44g. 

* However, there are certain tribes in Queensland where the things thus 
attributed to a social group are not forbidden for the members of the group : 
this is notably the case with the Wakelbura. It is to be remembered that in this 
society, it is the matrimonial classes that serve as the framework of the classifica- 
tion (see above, p. 144). Not only are the men of one class allowed to eat the 
animals attributed to this class, but they may eat no others. All other food is 
forbidden them (Howitt, Nat. Tr., p. 113 ; Curr, III, p. 27). 

But we must not conclude from this that these animals are considered profane. 
In fact, it should be noticed that the individual not only has the privilege of 
eating them, but that he is compelled to do so, for he cannot nourish himself 
otherwise. Now the imperative nature of this rule is a sure sign that we are in 
the presence of things having a religious nature, only this has given rise to 
a positive obligation rather than the negative one known as an interdiction. 
Perhaps it is not quite impossible to see how this deviation came about. We 
have seen above (p. 140) that every individual is thought to have a sort of 
property-right over his totem and consequently over the things dependent upon 
it. Perhaps, under the influence of special circumstances, this aspect of the 
totemic relation was developed, and they naturally came to believe that only the 
members of the clan had the right of disposing of their totem and all that is 
connected with it, and that others, on the contrary, did not have the right of 
touching it. Under these circumstances, a tribe could nourish itself only on the 
food attributed to it. 

* Mrs. Parker uses the expression " multiplex totems." 

* As examples, see the Euahlayi tribe in Mrs. Parker's book (pp. 15 if.) and 
the Wotjobaluk (Howitt, Nat. Tr., pp. 121 fi. ; cf. the above-mentioned article 
of Mathews). 

152 Elementary Forms of Religious Life 

it or demand it. It thus happens that a single individual has two 
totems, a principal totem common to the whole clan and a sub- 
totem which is special to the sub-clan of which he is a member. 
This is something analogous to the nomen and cognomen of the 

Sometimes we see a sub-clan emancipate itself completely and 
become an autonomous group and an independent clan ; then, 
the sub-totem, on its side, becomes a regular totem. One tribe 
where this process of segmentation has been pushed to the limit, 
so to speak, is the Arunta. The information contained in the 
first book of Spencer and Gillen showed that there were some 
sixty totems among the Arunta •,^ but the recent researches of 
Strehlow have shown the number to be much larger. He counted 
no less than 442.^ Spencer and Gillen did not exaggerate at all 
when they said, " In fact, there is scarcely an object, animate 
or inanimate, to be found in the country occupied by the natives 
which does not give its name to some totemic group."* Now 
this multitude of totems, whose number is prodigious when com- 
pared to the population, is due to the fact that under special 
circumstances, the original clans have divided and sub-divided 
infinitely ; consequently nearly all the sub-totems have passed 
to the stage of totems. 

This has been definitely proved by the observations of Streh- 
low. Spencer and Gillen cited only certain isolated cases of 
associated totems.^ Strehlow has shown that this is in reality 
an absolutely general organization. He has been able to draw 
up a table where nearly all the totems of the Arunta are classified 
according to this principle : all are attached, either as associates 
or as auxiliaries, to some sixty principal totems.^ The first 
are believed to be in the service of the second.' This state of 

^ See the examples in Howitt, Nat. Tr., p. 122. 

* See our De quelques formes primitives de classification, p. 28, n. 2. 
' Strehlow, II, pp. 61-72. * Nat. Tr., p. 112. 

* See especially Nat. Tr., p. 447, and Nor. Tr., p. 151. 

' Strehlow, III, pp. xiii-xviii. It sometimes happens that the same secondary 
totems are attached to two or three principal totems at the same time. This is 
undoubtedly because Strehlow has not been able to establish with certainty 
which is the principal totem. 

Two interesting facts which appear from this table confirm certain propositions 
which we had already formulated. First, the principal totems are nearly all 
animals, with but rare exceptions. Also, stars are always only secondary or 
associated totems. This is another proof that these latter were only slowly 
advanced to the rank of totems and that at first the principal totems were 
preferably chosen from the animal kingdom. 

' According to the myth, the associate totems served as food to the men 
of the principal totem in the fabulous times, or, when these are trees, they gave 
their shade (Strehlow, III, p. xii ; Spencer and Gillen, Nat. Tr., p. 403). The 
fact that the associate totems are believed to have been eaten does not imply that 
they are considered profane ; for in the mythical period, the principal totem itself 
was consumed by the ancestors, the founders of the clan, according to the belief. 

Totemic Beliefs 153 

dependence is very probably the echo of a time when the " allies " 
of to-day were only sub-totems, and consequently when the tribe 
contained only a small number of clans subdivided into sub- 
clans. Numerous survivals confirm this hypothesis. It fre- 
quently happens that two groups thus associated have the same 
totemic emblem : now this unity of emblem is explicable only 
if the two groups were at first only one.^ The relation of the 
two clans is also shown by the part and the interest that each 
one takes in the rites of the other. The two cults are still only 
imperfectly separated ; this is very probably because they were 
at first completely intermingled. ^ Tradition explains the bonds 
which unite them by imagining that formerly the two clans 
occupied neighbouring places.^ In other cases, the myth says 
expressly that one of them was derived from the other. It is 
related that at first the associated animal belonged to the species 
still serving as principal totem ; it differentiated itself at a later 
period. Thus the chantunga birds, which are associated with 
the witchetly grub to-day, were witchetly grubs in fabulous 
times, who later transformed themselves into birds. Two 
species which are now attached to the honey-ant were formerly 
honey-ants, etc.^ This transformation of a sub-totem into a 
totem goes on by imperceptible degrees, so that in certain cases 
the situation is undecided, and it is hard to say whether one is 
dealing with a principal totem or a secondary one.^ As Howitt 
says in regard to the Wotjobaluk, there are sub-totems which are 
totems in formation.^ Thus the different things classified in a 
clan constitute, as it were, so many nuclei around which new 
totemic cults are able to form. This is the best proof of the 
religious sentiments which they inspire. If they did not have a 
sacred character, they could not be promoted so easily to the 
same dignity as the things which are sacred before all others, the 
regular totems. 

So the field of religious things extends well beyond the limits 
within which it seemed to be confined at first. It embraces not 
only the totemic animals and the human members of the clan ; but 
since no known thing exists that is not classified in a clan and 
under a totem, there is likewise nothing which does not receive 

^ Thus in the Wild Cat clan, the designs carved on the churinga represent the 
Hakea tree, which is a distinct totem to-day (Spencer and Gillen, Nat. Tr., 
pp. 147 f.). Strehlow (III, p. xii, n. 4) says that this is frequent. 

- Spencer and Gillen, Nor. Tr., p. 182 ; Nat. Tr., pp. 151 and 297. 

' Nat. Tr., pp. 151 and 158. 

* Ibid., pp. 448 and 44g. 

* Thus Spencer and Gillen speak of a pigeon called Inturrita, sometimes as 
a principal totem {Nat. Tr., p. 410), sometimes as an associate totem (ibid., 
p. 448). 

" Howitt, Further Notes, pp. 63-64. 

154 Elementary Forms of Religious Life 

to some degree something of a religious character. When, in the 
rehgions which later come into being, the gods properly so- 
called appear, each of them will be set over a special category of 
natural phenomena, this one over the sea, that one over the air, 
another over the harvest or over fruits, etc., and each of these 
provinces of nature will be believed to draw what life there is in it 
from the god upon whom it depends. This division of nature 
among the different divinities constitutes the conception which 
these religions give us of the universe. Now so long as humanity 
has not passed the phase of totemism, the different totems of 
the tribe fulfil exactly the same functions that will later fall upon 

[ the divine personalities. In the Mount Gambler tribe, which we 
have taken as our principal example, there are ten clans ; conse- 
quently the entire world is divided into ten classes, or rather 
into ten families, each of which has a special totem as its basis. 
It is from this basis that the things classed in the clan get all 
their reality, for they are thought of as variant forms of the 
totemic being ; to return to our example, the rain, thunder, 
lightning, clouds, hail and winter are regarded as different sorts 
of crows. When brought together, these ten families of things 
make up a complete and systematic representation of the world ; 
and this representation is religious, for religious notions furnish 
its basis. Far from being limited to one or two categories of 
beings, the domain of totemic religion extends to the final limits 
of the known universe. Just like the Greek religion, it puts the 
divine everywhere ; the celebrated formula Travrà 'TrKriprj dewv 
(everything is full of the gods), might equally well serve it as 

L, motto. 

However, if totemism is to be represented thus, the notion of 
it which has long been held must be modified on one essential 
point. Until the discoveries of recent years, it was made to 
consist entirely in the cult of one particular totem, and it was 
defined as the religion of the clan. From this point of view, 
each tribe seemed to have as many totemic religions, each inde- 
pendent of the others, as it had different clans. This conception 
was also in harmony with the idea currently held of the clan ; in 
fact, this was regarded as an autonomous society, ^ more or less 
closed to other similar societies, or having only external and 
superficial relations with these latter. But the reality is more 
complex. Undoubtedly, the cult of each totem has its home 
in the corresponding clan ; it is there, and only there, that it is 
celebrated ; it is members of the clan who have charge of it ; 

^ Thus it comes about that the clan has frequently been confounded with the 
tribe. This confusion, which frequently introduces trouble into the writings of 
ethnologists, has been made especially by Curr (I, pp. 6i fi.). 

Totemic Beliefs 155 

it is through them that it is transmitted from one generation 
to another, along with the beliefs which are its basis. But it is 
also true that the different totemic cults thus practised within 
a single tribe do not have a parallel development, though re- 
maining ignorant of each other, as if each of them constituted 
a complete and self-sufhcing religion. On the contrary, they 
mutually imply each other ; they are only the parts of a single 
whole, the elements of a single religion. The men of one clan 
never regard the beliefs of neighbouring clans with that in- 
difference, scepticism or hostility which one religion ordinarily 
inspires for another which is foreign to it ; they partake of these 
beliefs themselves. The Crow people are also convinced that 
the Snake people have a mythical serpent as ancestor, and that 
they owe special virtues and marvellous powers to this origin. 
And have we not seen that at least in certain conditions, a man 
may eat a totem that is not his own only after he has observed 
certain ritual formalities ? Especially, he must demand the 
permission of the men of this totem, if any are present. So for 
him also, this food is not entirely profane ; he also admits that 
there are intimate affinities between the members of a clan of 
which he is not a member and the animal whose name they bear. 
Also, this community of belief is sometimes shown in the cult. 
If in theory the rites concerning a totem can be performed only 
by the men of this totem, nevertheless representatives of different 
clans frequently assist at them. It sometimes happens that 
their part is not simply that of spectators ; it is true that they 
do not officiate, but they decorate the officiants and prepare 
the service. They themselves have an interest in its being 
celebrated ; therefore, in certain tribes, it is they who invite the 
qualified clan to proceed with the ceremonies.^ There is even a 
whole cycle of rites which must take place in the presence of the 
assembled tribe : these are the totemic ceremonies of initiation. ^ 
Finally, the totemic organization, such as we have just 
described it, must obviously be the result of some sort of an 
indistinct understanding between all the members of the tribe. 
It is impossible that each clan should have made its beliefs in 
an absolutely independent manner ; it is absolutely necessary 
that the cults of the different totems should be in some way 
adjusted to each other, since they complete one another exactly. 
In fact, we have seen that normally a single totem is not repeated 
twice in the same tribe, and that the whole universe is divided 
up among the totems thus constituted in such a way that the 
same object is not found in two different clans. So methodical 

^ This is the case especially among the Warramunga {Nor. Tr., p. 298). 
* See, for example, Spencer and Gillen, Nat. Tr., p. 380 and passim. 

156 Elementary Forms of Religious Life 

a division could never have been made without an agreement, 
tacit or planned, in which the whole tribe participated. So the 
group of beliefs which thus arise are partially (but only partially) 
a tribal affair. ^ 

To sum up, then, in order to form an adequate idea of totemism, 
we must not confine ourselves within the limits of the clan, but 
must consider the tribe as a whole. It is true that the particular 
cult of each clan enjoys a very great autonomy ; we can now see 
that it is within the clan that the active ferment of the rehgious life 
takes place. But it is also true that these cults fit into each other 
and the totemic religion is a complex system formed by their union, 
just as Greek polytheism was made by the union of all the particu- 
lar cults addressed to the different divinities. We have just shown 
that, thus understood, totemism also has it cosmology. , 

1 One might even ask if tribal totems do not exist sometimes. Thus, among 
the Arunta, there is an animal, the wild cat, which serves as totem to a particular 
clan, but which is forbidden for the whole tribe ; even the people of other clans 
can eat it only very moderately {Nat. Tr., p. 168). But we believe that it would 
be an abuse to speak of a tribal totem in this case, for it does not follow from the 
fact that the free consumption of an animal is forbidden that this is a totem. 
Other causes can also give rise to an interdiction. The religious unity of the tribe 
is undoubtedly real, but this is affirmed with the aid of other symbols. We shall 
show what these are below (Bk. 11, ch. ix). 



The Individual Totem and the Sexual Totem 

UP to the present, we have studied totemism only as a 
pubhc institution : the only totems of which we have 
spoken are common to a clan, a phratry or, in a sense, to a tribe ; ^ 
an individual has a part in them only as a member of a group. 
But we know that there is no religion which does not have an 
individual aspect. This general observation is applicable to 
totemism. In addition to the impersonal and collective totems 
which hold the first place, there are others which are peculiar 
to each individual, which express his personality, and whose cult 
he celebrates in private. 


In certain Australian tribes, and in the majority of the Indian 
tribes of North America,- each individual personally sustains 
relations with some determined object, which are comparable 
to those which each clan sustains with its totem. This is some- 
times an inanimate being or an artificial object ; but it is generally 
an animal. In certain cases, a special part of the organism, such 
as the head, the feet or the liver, fulfils this office.^ 

The name of the thing also serves as the name of the individual. 
It is his personal name, his forename, which is added to that of 
the collective totem, as the praenomen of the Romans was to the 
nomen gentilicium. It is true that this fact is not reported except 
in a certain number of societies,'* but it is probably general. In 

* The totems belong to the tribe in the sense that this is interested as a body 
in the cult which each clan owes to its totem. 

" Frazer has made a very complete collection of the texts relative to indi- 
vidual totemism in North America {Totemism and Exogamy, III, pp. 370-456). 

^ For example, among the Hurons, the Iroquois, the Algonquins (Charlevoix, 
Histoire de la Nouvelle France, VI, pp. 67-70 ; Sagard, Le grand voyage au pays 
des Hurons, p. 160), or among the Thompson Indians (Teit, The Thompson 
Indians of British Columbia, p. 355). 

^ This is the case of the Yuin (Howitt, Nat. Tr., p. 133), the Kumai {ibid., 
p. 135), several tribes of Queensland (Roth, Superstition, Magic and Medicine, 
North Queensland Ethnography, Bulletin No. 5, p. 19 ; Haddon, Head-Hunters, 
P- 193) ; among the Delaware (Heckewelder, An Account of the History . . . of 
the Indian Nations, p. 238), among the Thompson Indians (Teit, op. cit., p. 355), 
and among the Salish Statlumli (Hill Tout, Rep. of the Ethnol. of the Statliimh, 
J. A. I., XXXV. pp. 147 ff.). 


158 Elementary Forms of Religious Life 

fact, we shall presently show that there is an identity of nature 
between the individual and the thing ; now an identity of nature 
implies one of name. Being given in the course of especially 
important religious ceremonies, this forename has a sacred 
character. It is not pronounced in the ordinary circumstances 
of profane life. It even happens that the word designating this 
object in the ordinary language must be modified to a greater 
or less extent if it is to serve in this particular case.^ This is 
because the terms of the usual language are excluded from the 
religious life. 

In certain American tribes, at least, this name is reinforced 
by an emblem belonging to each individual and representing, 
under various forms, the thing designated by the name. For 
example, each Mandan wears the skin of the animal of which he 
is the namesake. 2 If it is a bird, he decorates himself with its 
feathers.^ The Hurons and Algonquins tattoo their bodies with 
its image.* It is represented on their arms.^ Among the north- 
western tribes, the individual emblem, just like the collective 
emblem of the clan, is carved or engraved on the utensils, houses,* 
etc. ; it serves as a mark of ownership.' Frequently the two 
coats-of-arms are combined together, which partially explains the 
great diversity of aspects presented by the totemic escutcheons 
among these peoples. ^ 

Between the individual and his animal namesake there exist 
the very closest bonds. The man participates in the nature of 
the animal ; he has its good qualities as well as its faults. For 
example, a man having the eagle as his coat-of-arms is believed 
to possess the gift of seeing into the future ; if he is named after 
a bear, they say that he is apt to be wounded in combat, for the 
bear is heavy and slow and easily caught ; ^ if the animal is 
despised, the man is the object of the same sentiment.^** The 
relationship of the two is even so close that it is believed that 
in certain circumstances, especially in case of danger, the man 
can take the form of the animal. ^^ Inversely, the animal is 

^ Hill Tout, loc. cit., p. 154. 

* Catlin, Manners and Customs, etc., London, 1876, I, p. 36. 
' Lettres édifiantes et curieuses, new edition, VI, pp. 172 fl. 

* Charlevoix, op. cit., VI, p. 69. 

* Dorsey, Siouan Cults, Xlth Rep., p. 443. 

* Boas, Kwakiutl, p. 323. 

"> Hill Tout, loc. cit., p. 154. 

* Boas, Kwakiutl, p. 323. 

" Miss Fletcher, The Import of the Totem, a Study from the Omaha Tribe 
{Smithsonian Rep. for 1897, p. 583). — Similar facts will be found in Teit, op. cit., 
PP- 354. 35^ ; Peter Jones, History of the Ojibway Indians, p. 87. 

*" This is the case, for example, with the dog among the Salish Statlumh, 
owing to the condition of servitude in which it lives (Hill Tout, loc. cit., p. 153). 

^* Langloh Parker, Euahlayi, p. 21. 

Totemic Beliefs 159 

regarded as a double of the man, as his alter ego^ The association 
of the two is so close that their destinies are frequently thought 
to be bound up together : nothing can happen to one without the 
other's feeling a reaction. ^ If the animal dies, the life of the man 
is menaced. Thus it comes to be a very general rule that one 
should not kill the animal, nor eat its flesh. This interdiction, 
which, when concerning the totem of the clan, allows of all sorts 
of attenuations and modifications, is now much more formal 
and absolute.^ 

On its side, the animal protects the man and serves him as 
a sort of patron. It informs him of possible dangers and of the 
way of escaping them ; * they say that it is his friend.^ Since it 
frequently happens to possess marvellous powers, it communicates 
them to its human associate, who believes in them, even under 
the proof of bullets, arrows, and blows of every sort.® This 
confidence of an individual in the efficacy of his protector is so 
great that he braves the greatest dangers and accomplishes the 
most disconcerting feats with an intrepid serenity : faith gives 
him the necessary courage and strength.' However, the relations 
of a man with his patron are not purely and simply those of 
dependence. He, on his side, is able to act upon the animal. 
He gives it orders ; he has influence over it. A Kumai having 
the shark as ally and friend believes that he can disperse the 
sharks who menace a boat, by means of a charm. ^ In other cases, 
the relations thus contracted are believed to confer upon the man 
a special aptitude for hunting the animal with success.^ 

* " The spirit of a man," says Mrs. Parker {ibid.), " is in his Yuanbeai (his 
individual totem), and his Yuanbeai is in him." 

* Langloh Parker, Euahlayi, p. 20. It is the same among certain Salish (Hill 
Tout, Ethn. Rep. on the Stseehs and Skauhts Tribes. J.A.I.. XXXIV, p. 324). 
The fact is quite general among the Indians of Central America (Brinton, 
NagucUism, a Study in Native American Folklore and History, in Proceed, of the 
Am. Philos. Soc, XXXIII, p. 32). 

* Parker, ibid. ; Howitt, Nat. Tr., p. 147 ; Dorsey, Siouan Cults; Xlth Rep., 
p. 443. Frazer has made a collection of the American cases and established the 
generality of the interdiction (Totemism and Exogamy, III, p. 450). It is true 
that in America, as we have seen, the individual must kill the animal whose skin 
serves to make what ethnologists call his medicine-sack. But this usage has been 
observed in five tribes only ; it is probably a late and altered form of the institution. 

* Howitt, Nat. Tr., pp. 135, 147, 387; Australian Medicine Men. J.A.I., XVI, 
p. 34 ; Teit, The Shuswap, p. 607. 

^ Meyer, Manners and Customs of the Aborigines of the Encounter Bay Tribe, 
in Woods, p. 197. 

' Boas, Vlth Rep. on the North-West Tribes of Canada, p. 93 ; Teit, The 
Thompson Indians, p. 336 ; Boas, Kwakiutl, p. 394. 

' Facts will be found in Hill Tout, Rep. of the Ethnol. of the Statlumh, J. A. I., 
XXXV, pp. 144, 145. Cf. Langloh Parker, op. cit., p. 29. 

" According to information given by Howitt in a personal letter to Frazer 
{Totemism and Exogamy, I, p. 495, and n. 2). 

* Hill Tout, Ethnol. Rep. on the Stseelis and Skaulits Tribes, J. A. I.. XXXIV, 
P 324- 

i6o Elementary Forms of Religious Life 

The very nature of these relations seems clearly to imply 
that the being to which each individual is thus associated is 
only an individual itself, and not a species. A man does not 
have a species as his alter ego. In fact, there are cases where it 
is certainly a certain determined tree, rock or stone that fulfils 
this function. 1 It must be thus every time that it is an animal, 
and that the existences of the animal and the man are believed 
to be connected. A man could not be united so closely to a 
whole species, for there is not a day nor, so to speak, an instant 
when the species does not lose some one of its members. Yet 
the primitive has a certain incapacity for thinking of the indi- 
vidual apart from the species ; the bonds uniting him to the 
one readily extend to the other ; he confounds the two in the 
same sentiment. Thus the entire species becomes sacred for 

This protector is naturally given different names in different 
societies : nagual among the Indians of Mexico,^ manitou among 
the Algonquins and okki among the Hurons,^ snam among certain 
Salish,^ sulia among others,* budjan among the Yuin,' yunbeai 
among the Euahlayi,^ etc. Owing to the importance of these 
beliefs and practices among the Indians of North America, some 
have proposed creating a word nagualism or manitoiiism to 
designate them.^ But in giving them a special and distinctive 
name, we run the risk of misunderstanding their relations with 
the rest of totemism. In fact, the same principle is applied in 
the one case to the clan and in the other to the individual. In 
both cases we find the same belief that there are vital connections 

^ Howitt, Australian Medicine Men, J. A. I., XVI, p. 34 ; Lafitau, Mœurs des 
Sauvages Amériquains, I, p. 370; Charlevoix, Histoire de la Nouvelle France, 
VI, p. 68. It is the same with the atai and tamaniu in Mota (Codrington, The 
Melanesians, pp. 250 f.). 

* Thus the line of demarcation between the animal protectors and fetishes, 
which Frazer has attempted to establish, does not exist. According to him, 
fetishism commences when the protector is an individual object and not a class 
{Totemism, p. 56) ; but it frequently happens in Australia that a determined 
animal talies this part (see Howitt, Australian Medicine Men, J. A. I., XVI, 
p. 34). The truth is that the ideas of fetish and fetisliism do not correspond to 
any definite thing. 

* Brinton, Nagualism, in Proceed. Amer. Philos. Soc, XXXIII, p. 32. 

* Charlevoix, VI, p. 67. 

* Hill Tout, Rep. on the Ethnol. of the Statlumh of British Columbia, J. A. I., 
XXXV, p. 142. 

* Hill Tout, Ethnol. Rep. on the Stseelis and Skaulits Tribes, J. A. I., XXXI V, 
pp. 311 ff. 

' Howitt, Nat. Tr., p. 133. " Langloh Parker, op. cit., p. 20. 

* J. W. Powell, An American View of Totemism, in Man, 1902, No. 84 ; 
Tylor, ibid.. No. i ; Andrew Lang has expressed analogous ideas in Social Origins, 
pp. 133-135. Also Frazer himself, turning from his former opinion, now thinks 
that until we are better acquainted with the relations existing between collective 
totems and " guardian spirits," it would be better to designate them by diflerent 
names [Totemism and E.xogamy, III, p. 456). 

Totemic Beliefs i6i 

between the things and the men, and that the former are endowed 
with special powers, of which their human alhes may also enjoy 
the advantage. We also find the same custom of giving the man 
the name of the thing with which he is associated and of adding 
an emblem to this name. The totem is the patron of the clan, just 
as the patron of the individual is his personal totem. So it is 
important that our terminology should make the relationship 
of the two systems apparent ; that is why we, with Frazer, 
shall give the name individual totemism to the cult rendered by 
each individual to his patron. A further justification of this 
expression is found in the fact that in certain cases the primitive 
himself uses the same word to designate the totem of the clan 
and the animal protector of the individual.^ If Tylor and Powell 
have rejected this term and demanded different ones for these 
two sorts of religious institutions, it is because the collective totem 
is, in their opinion, only a name or label, having no religious 
character. 2 But we, on the contrary, know that it is a sacred 
thing, and even more so than the protecting animal. Moreover, 
the continuation of our study will show how these two varieties 
of totemism are inseparable from each other. ^ 

Yet, howsoever close the kinship between these two institutions 
may be, there are important difterences between them. While 
the clan beheves that it is the offspring of the animal or plant 
serving it as totem, the individual does not believe that he has 
any relationship of descent with his personal totem. It is a 
friend, an associate, a protector ; but it is not a relative. He 
takes advantage of the virtues it is believed to possess ; but he 
is not of the same blood. In the second place, the members of 
a clan allow neighbouring clans to eat of the animal whose 
name they bear collectively, under the simple condition that 
the necessary formalities shall be observed. But, on the contrary, 
the individual respects the species to which his personal totem 
belongs and also protects it against strangers, at least in those 
parts where the destiny of the man is held to be bound up with 
that of the animal. 

But the chief difference between these two sorts of totems 
is in the manner in which they are acquired. 

The collective totem is a part of the civil status of each in- 
dividual : it is generally hereditary ; in any case, it is birth 

* This is the case in Australia among the Yuin (Howitt. Nat. Tr., p. 8i). and 
the Narrinyeri (Meyer, Manners and Customs of the Aborigines of the Encounter 
Bay Tribe, in Woods, pp. 197 ff.). 

* " The totem resembles the patron of the individual no more than an 
escutcheon resembles the image of a saint," says Tylor (op. cit., p. 2). Likewise, 
if Frazer has taken up the theory of Tylor, it is because he refuses all religious 
character to the totem of the clan [Totemism and Exogamy, III, p. 452). 

' See below, chapter ix of this book. 

102 Elementary Forms of Religious Life 

which designates it, and the wish of men counts for nothing. 
Sometimes the child has the totem of his mother (Kamilaroi, 
Dieri, Urabunna, etc.) ; sometimes that of his father (Narrinyeri, 
Warramunga, etc.) ; sometimes the one predominating in the 
locahty where his mother conceived (Arunta, Loritja). But, on 
the contrary, the individual totem is acquired by a deliberate 
act : ^ a whole series of ritual operations are necessary to deter- 
mine it. The method generally employed by the Indians of 
North America is as follows. About the time of puberty, as the 
time for initiation approaches, the young man withdraws into 
a distant place, for example, into a forest. There, during a 
period varying from a few days to several years, he submits 
himself to all sorts of exhausting and unnatural exercises. He 
fasts, mortifies himself and inflicts various mutilations upon 
himself. Now he wanders about, uttering violent cries and 
veritable howls ; now he lies extended, motionless and lamenting, 
upon the ground. Sometimes he dances, prays and invokes his 
ordinary divinities. At last, he thus gets himself into an extreme 
state of super-excitation, verging on delirium. When he has 
reached this paroxysm, his representations readily take on the 
character of hallucinations. " When," says Heckewelder, " a 
boy is on the eve of being initiated, he is submitted to an alter- 
nating régime of fasts and medical treatment ; he abstains 
from all food and takes the most powerful and repugnant drugs : 
at times, he drinks intoxicating concoctions until his mind 
really wanders. Then he has, or thinks he has, visions and extra- 
ordinary dreams to which he was of course predisposed by all 
this training. He imagines himself flying through the air, ad- 
vancing under the ground, jumping from one mountain-top to 
another across the valleys, and fighting and conquering giants 
and monsters." ^ If in these circumstances he sees, or, as amounts 
to the same thing, he thinks he sees, while dreaming or while 
awake, an animal appearing to him in an attitude seeming to 

' Yet according to one passage in Mathews, the individual totem is hereditary 
among the Wotjobaluk. " Each individual," he says, " claims some animal, plant 
or inanimate object as his special and personal totem, which he inherits from his 
mother " {Journ. and Proc. of the Roy. Soc. of N.S. Wales. XXXVIII, p. 291). 
But it is evident that if all the children in the same family had the personal 
totem of their mother, neither they nor she would really have personal totems at 
all. Mathews probably means to say that each individual chooses his individual 
totem from the list of things attributed to the clan of his mother. In fact, we 
shall see that each clan has its individual totems which are its exclusive property ; 
the members of the other clans cannot make use of them. In this sense, birth 
determines the personal totem to a certain extent, but to a certain extent 

* Heckewelder, An Account of the History, Manners and Customs of the Indian 
Nations who once inhabited Pennsylvania, in Transactions of the Historical and 
Literary Committee uf the American Philosophical Society, I, p. 238. 

Totemic Beliefs 163 

show friendly intentions, then he imagines that he has discovered 
the patron he awaited.^ 

Yet this procedure is rarely employed in Australia. ^ On this 
continent, the personal totem seems to be imposed by a third 
party, either at birth ^ or at the moment of initiation. ^ Generally 
it is a relative who takes this part, or else a personage invested 
with special powers, such as an old man or a magician. Sometimes 
divination is used for this purpose. For example, on Charlotte 
Bay, Cape Bedford or the Proserpine River, the grandmother or 
some other old woman takes a little piece of umbilical cord to 
which the placenta is still attached and whirls it about quite 
violently. Meanwhile the other old women propose different 
names. That one is adopted which happens to be pronounced 
just at the moment when the cord breaks.^ Among the Yarrai- 
kanna of Cape York, after a tooth has been knocked out of the 
young initiate, they give him a little water to rinse his mouth 
and ask him to spit in a bucket full of water. The old men care- 
fully examine the clot formed by the blood and saliva thus spit 
out, and the natural object whose shape it resembles becomes 
the personal totem of the young man.* In other cases, the totem 
is transmitted from one individual to another, for example from 
father to son, or uncle to nephew.' This method is also used in 
America. In a case reported by Hill Tout, the operator was 
a shaman,^ who wished to transmit his totem to his nephew. 
" The uncle took the symbol of his snam (his personal totem), 
which in this case was a dried bird's skin, and bade his nephew 
breathe upon it. He then blew upon it also himself, uttered some 
mystic words and the dried skin seemed to Paul (the nephew) to 
become a living bird, which flew about them a moment or two 

^ See Dorsey, Siouan Cults, Xlth Rep., p. 507 ; Catlin, op. cit., I, p. 37 ; 
Miss Fletcher, The Import of the Totem, in Smithsonian Rep. for 1897, p. 580 ; 
Teit, The Thompson Indians, pp. 317-320; Hill Tout, J.A.I., XXXV, p. 144. 

* But some examples are found. The Kurnai magicians see their personal 
totems revealed to them in dreams (Howitt, Nat. Tr., p. 387 ; On Australian 
Medicine Men, in J. A. I., XVI, p. 34). The men of Cape Bedford believe that 
when an old man dreams of something during the night, this thing is the personal 
totem of the first person he meets the next day (W. E. Roth, Superstition, Magic 
and Medicine, p. 19). But it is probable that only supplementary and accessory 
totems are acquired in this way ; for in this same tribe another process is used 
at the moment of initiation, as we said in the text. 

* In certain tribes of which Roth speaks (ibid.) ; also in certain tribes near to 
Maryborough (Howitt, Nat. Tr., p. 147). 

* Among the Wiradjuri (Howitt, Nat. Tr., p. 406 ; On Australian Medicine 
Men, in J. A. I.. XVI, p. 50). 

' Roth, loc. cit. » Haddon, Head Hunters, pp. 193 fï. 

' Among the Wiradjuri (same references as above, n. 4). 

■* In general, it seems as though these transmissions from father to son never 
take place except when the father is a shaman or a magician. This is also the 
case among the Thompson Indians (Teit, The Thompson Indians, p. 320) and the 
Wiradjuri, of whom we just spoke. 

164 Elementary Forms of Religious Life 

and then finally disappeared. Paul was then instructed by his 
uncle to procure that day a bird's skin of the same kind as his 
uncle's and wear it on his person. This he did, and that night 
he had a dream, in which the snani appeared to him in the shape 
of a human being, disclosed to him its mystic name by which it 
might be summoned, and promised him protection." ^ 

Not only is the individual totem acquired and not given, but 
ordinarily the acquisition of one is not obligatory. In the first 
place, there are a multitude of tribes in Australia where the custom 
seems to be absolutely unknown. ^ Also, even where it does 
exist, it is frequently optional. Thus among the Euahlayi, 
while all the magicians have individual totems from which they 
get their powers, there are a great number of laymen who have 
none at all. It is a favour given by the magician, but which he 
reserves for his friends, his favourites and those who aspire to 
becoming his colleagues.^ Likewise, among certain Salish, persons 
desiring to excel especially either in fighting or in hunting, or 
aspirants to the position of shaman, are the only ones who 
provide themselves with protectors of this sort.* So among 
certain peoples, at least, the individual totem seems to be con- 
sidered an advantage and convenient thing rather than a necessity. 
It is a good thing to have, but a man can do without one. In- 
versely, a man need not limit himself to a single totem ; if he 
wishes to be more fully protected, nothing hinders his seeking 
and acquiring several,^ and if the one he has fulfils its part badly, 
he can change it.* 

But while it is more optional and free, individual totemism 
contains within it a force of resistance never attained by the 
totemism of the clan. One of the chief informers of Hill Tout 
was a baptized Salish ; however, though he had sincerely 
abandoned the faith of his fathers, and though he had become 
a model catechist, still his faith in the efiicacy of the personal 
totems remained unshaken.' Similarly, though no visible traces 
of collective totemism remain in civilized countries, the idea 
that there is a connection between each individual and some 

* Hill Tout {J. A ./., XXXV, pp. 146 f.). The essential rite is the blowing upon 
the skin : if this were not done correctly, the transmission would not take place. 
As we shall presently see, the breath is the soul. When both breathe upon the 
skin of the animal, the magician and the recipient each exhale a part of their 
souls, which are thus fused, while partaking at the same time of the nature of 
the animal, who also takes part in the ceremony in the form of its symbol. 

2 N. W. Thomas, Further Remarks on Mr. Hill Tout's Views on Totemism, in 
Man, iQO.f, p. 85. 

* Langloh Parker, op. cit., pp. 20, 29. 

« Hill Tout, in J. A. I., XXXV, pp. 143 and 146 ; ibid., XXXIV, p. 324. 
'•> Parker, op. cit., p. 30 ; Teit, The Thompson Indians, p. 320 ; Hill Tout, in 
J.A.I.. XXXV, p. 144. 

" Charlevoix, VI, p. Og. ' Hill Tout, ibid., y>. 145. 

Totemic Beliefs 165 

animal, plant or other object, is at the bottom of many customs 
still observable in many European countries. ^ 

Between collective totemism and individual totemism there is 
ah intermediate form partaking of the characteristics of each : 
this is sexual totemism. It is found only in Australia and in a 
small number of tribes. It is mentioned especially in Victoria 
and New South Wales. ^ Mathews, it is true, claims to have 
observed it in all the parts of Australia that he has visited, but 
he gives no precise facts to support this affirmation.^ 

Among these different peoples, all the men of the tribe on the 
one hand, and all the women on the other, to whatever special 
clan they may belong, form, as it were, two distinct and even 
antagonistic societies. Now each of these two sexual corporations 
believes that it is united by mystical bonds to a determined 
animal. Among the Kurnai, all the men think they are brothers, 
as it were, of the emu-wren (Yeerung), all the women, that they 
are as sisters of the linnet (Djeetgun) ; all the men are Yeerung 
and all the women are Djeetgun. Among the Wotjobaluk and the 
Wurunjerri, it is the bat and the nightjar (a species of screech-owl) 
respectively who take this rôle. In other tribes, the woodpecker 
is substituted for the nightjar. Each sex regards the animal to 
which it is thus related as a sort of protector which must be treated 
with the greatest regard ; it is also forbidden to kill and eat it.'* 

Thus this protecting animal plays the same part in relation 
to the sexual society that the totem of the clan plays to this 
latter group. So the expression sexual totemism, which we 
borrow from Frazer,^ is justified. This new sort of totem 
resembles that of the clan particularly in that it, too, is 
collective ; it belongs to all the people of one sex indiscriminately. 
It also resembles this form in that it implies a relationship of 
descent and consanguinity between the animal patron and the 

Ï Thus at the birth of a child, a tree is planted which is cared for piously ; for 
it IS believed that its fate and the child's are united. Frazer, in his Golden Bough, 
•rives a number of customs and beliefs translating this same idea in different ways. 
(Cf. Hartland, Legend of Perseus, II, pp. 1-55.) 

* Howitt, Nat. Tr., pp. 148 fi. ; Fison and Howitt, Kamilaroi and Kurnai, 
pp. 194, 201 ff. ; Dawson, Australian Aborigines, p. 52. Pétrie also mentions it 
in Queensland [Tom Petrie's Reminiscences of Early Queensland, pp. 62 and 118). 

* Journ. and Proc. of the Roy. Soc. of N.S. Wales, XXXVIII, p. 339. Must 
we see a trace of sexual totemism in the following custom of the VVarramunga ? 
When a dead person is buried, a bone of the arm is kept. If it is a woman, the 
feathers of an emu are added to the bark in which it is wrapped up ; if it is a 
man, the feathers of an owl (Nor. Tr., p. 169). 

* Some cases are cited where each sexual group has two sexual totems ; thus 
the Wurunjerri unite the sexual totems of the Kurnai (the em>i-wren and the 
linnet) to those of the Wotjobaluk (the bat and the nightjar owl). See Howitt, 
Nat. Tr., p. 150. s Totemism, p. 51. 

1 66 Elementary Forms of Religious Life 

corresponding sex : among the Kurnai, all the men are believed to 
be descended from Yeerung and all the women from Djeetgun.^ 
The first observer to point out this curious institution described 
it, in 1834, in the following terms : " Tilmun, a little bird the 
size of a thrush (it is a sort of woodpecker), is supposed by the 
women to be the first maker of women. These birds are held in 
veneration by the women only." ^ So it was a great ancestor. 
But in other ways, this same totem resembles the individual 
totem. In fact, it is believed that each member of a sexual 
group is personally united to a determined individual of the 
corresponding animal species. The two lives are so closely 
associated that the death of the animal brings about that of the 
man. " The life of a bat," say the Wotjobaluk, " is the life of a 
man." ^ That is why each sex not only respects its own totem, 
but forces the members of the other to do so as well. Every 
violation of this interdiction gives rise to actual bloody battles 
between the men and the women.* 

Finally, the really original feature of these totems is that they 
are, in a sense, a sort of tribal totems. In fact, they result from 
men's representing the tribe as descended as a whole from one 
couple of mythical beings. Such a belief seems to demonstrate 
clearly that the tribal sentiment has acquired sufficient force 
to resist, at least to a considerable extent, the particularism of 
the clans. In regard to the distinct origins assigned to men 
and to women, it must be said that its cause is to be sought in 
the separate conditions in which the men and the women live.^ 

It would be interesting to know how tl^e sexual totems are 
related to the totems of the clans, according to the theory of the 
Australians, what relations there were between the two ancestors 
thus placed at the commencement of the tribe, and from which 
one each special clan is believed to be descended. But the ethno- 
graphical data at our present disposal do not allow us to resolve 
these questions. Moreover, however natural and even necessary 
it may appear to us, it is very possible that the natives never 
raised it. They do not feel the need of co-ordinating and syste- 
matizing their beliefs as strongly as we do.* 

* Kamilaroi and Kurnai, p. 215. 

* Threlkeld, quoted by Mathews, loc. cit., p. 339. 
' Howitt, Nat. Tr., pp. 148, 151. 

* Kamilaroi and Kurnai, pp. 200-203 ; Howitt. Nat. Tr., p. 149 ; Pétrie, 
op. cit., p. 62. Among the Kurnai, these bloody battles frequently terminate in 
marriages of which they are, as it were, a sort of ritual precursor. Sometimes 
they are merely plays (Pétrie, loc. cit.). 

■• On this point, see our study on La Prohibition de l'inceste et ses origines, in 
the Année Sociologique, I, pp. 44 fï. 

' However, as we shall presently see (ch. ix), there is a connection between 
the sexual totems and the great gods. 



Critical Examination of Preceding Theories 

THE beliefs which we have just summarized are manifestly 
of a religious nature, since they imply a division of things 
into sacred and profane. It is certain that there is no thought 
of spiritual beings, and in the course of our exposition we have 
not even had occasion to pronounce the words, spirits, genii or 
divine personalities. But if certain writers, of whom we shall 
have something more to say presently, have, for this reason, 
refused to regard totemism as a religion, it is because they have 
an inexact notion of what religious phenomena are. 

On the other hand, we are assured that this religion is the 
most primitive one that is now observable and even, in all 
probability, that has ever existed. In fact, it is inseparable from 
a social organization on a clan basis. Not only is it impossible, 
as we have already pointed out, to define it except in connection 
with the clan, but it even seems as though the clan could not 
exist, in the form it has taken in a great number of Australian 
societies, without the totem. For the members of a single clan 
are not united to each other either by a common habitat or 
by common blood, as they are not necessarily consanguineous 
and are frequently scattered over different parts of the tribal 
territory. Their unity comes solely from their having the same 
name and the same emblem, their believing that they have the 
same relations with the same categories of things, their practising 
the same rites, or, in a word, from their participating in the 
same toteraic cult. Thus totemism and the clan mutually imply 
each other, in so far, at least, as the latter is not confounded with 
the local group. Now the social organization on a clan basis is 
the simplest which we know. In fact, it exists in all its essential 
elements from the moment when the society includes two primary 
clans ; consequently, we may say that there are none more 
rudimentary, as long as societies reduced to a single clan have 
not been discovered, and we believe that up to the present no 
traces of such have been found. A religion so closely connected 
to a social system surpassing all others in simplicity may well 


i68 Elementary Forms of Religions Life 

be regarded as the most elementary religion wc can possibly 
know. If we succeed in discovering the origins of the beliefs 
which we have just analysed, we shall very probably discover 
at the same time the causes leading to the rise of the religious 
sentiment in humanity. 

But before treating this question for ourselves, we must examine 
the most authorized solutions of it which have already been 


In the first place, we find a group of scholars who believe 
that they can account for totemism by deriving it from some 
previous religion. 

For Tylor ^ and Wilken,^ totemism is a special form of the 
cult of the ancestors ; it was the widespread doctrine of the 
transmigration of souls that served as a bridge between these 
two religious systems. A large number of peoples believe that 
after death, the soul does not remain disincarnate for ever, but 
presently animates another living body ; on the other hand, 
" the lower psychology, drawing no definite line of demarcation 
between the souls of men and of beasts, can at least admit with- 
out difficulty the transmigration of human souls into the bodies 
of the lower animals." ^ Tylor cites a certain number of cases.* 
Under these circumstances, the religious respect inspired by the 
ancestor is quite naturally attached to the animal or plant 
with which he is presently confounded. The animal thus serving 
as a receptacle for a venerated being becomes a holy thing, the 
object of a cult, that is, a totem, for all the descendants of the 
ancestor, who form the clan descended from him. 

Facts pointed out by Wilken among the societies of the Malay 
Archipelago would tend to prove that it really was in this manner 
that the totemic beliefs originated. In Java and Sumatra, 
crocodiles are especially honoured ; they are regarded as benevo- 
lent protectors who must not be killed ; offerings are made to 
them. Now the cult thus rendered to them is due to their being 
supposed to incarnate the souls of ancestors. The Malays of 
the Philippines consider the crocodile their grandfather ; the 
tiger is treated in the same way for the same reasons. Similar 
beliefs have been observed among the Bantous.^ In Melanesia 

^ Primitive Culture, I, p. 402 ; II, p. 237 ; Remarks on Totemism, with 
especial reference to some modern theories concerning it, in J.A.I., XXVIII, and I, 
New Series, p. 13S. 

* Het Animisme bij den Volken van den indischen Archipel, pp. 69—75. 

* T^'lor, Primitive Culture, II, p. 6. * Tylor, ibid., II, pp. 6-18. 

* G. McCall Theal, Records of South-Eastern Africa, VII. We are acquainted 
with this work only through an article by Frazer, South African Totemism, 
published in Man, 1901, No. iii. 

Origins of these Beliefs 169 

it sometimes happens that an influential man, at the moment of 
death, announces his desire to reincarnate himself in a certain 
animal or plant ; it is easily understood how the object thus 
chosen as his posthumous residence becomes sacred for his whole 
family.^ So, far from being a primitive fact, totemism would 
seem to be the product of a more complex religion which pre- 
ceded it. 2 

But the societies from which these facts were taken had 
already arrived at a rather advanced stage of culture ; in any 
case, they had passed the stage of pure totemism. They have 
families and not totemic clans. ^ Even the majority of the animals 
to which religious honours are thus rendered are venerated, not 
by special groups of families, but by the tribes as a whole. So 
if these beliefs and practices do have some connection with 
ancient totemic cults, they now represent only altered forms of 
them ^ and are consequently not very well fitted for showing 
us their origins. It is not by studying an institution at the 
moment when it is in full decadence that we can learn how it 
was formed. If we want to know how totemism originated, 
it is neither in Java nor Sumatra nor Melanesia that we must 
study it, but in Australia. Here we find neither a cult of the 
dead^ nor the doctrine of transmigration. Of course they believe 
that the mythical heroes, the founders of the clan, reincarnate 
themselves periodically ; but this is in human bodies only ; each 
birth, as we shall see, is the product of one of these reincarnations. 
So if the animals of the totemic species are the object of rites, it 
is not because the ancestral souls are believed to reside in them. 
It is true that the first ancestors are frequently represented 
under the form of an animal, and this very common representation 
is an important fact for which we must account ; but it was 
not the belief in metempsychosis which gave it birth, for this 
belief is unknown among Australian societies. 

Moreover, far from being able to explain totemism, this 
belief takes for granted one of the fundamental principles upon 
which this rests ; that is to say, it begs the question to be ex- 
plained. It, just as much as totemism, implies that man is 

' Codrington, The Melanesians, pp. 32 f., and a personal letter by the same 
author cited by Tylor in J. A. I.. XXVIII, p. 147. 

* This is practically the solution adopted by Wundt {Mythus und Religion, 
II, p. 269). 

' It is true that according to Tylor's theory, a clan is only an enlarged family ; 
therefore whatever may be said of one of these groups is, in his theory, applicable 
to the other [J. A. I., XXVIII, p. 157). But this conception is exceedingly con- 
testable ; only the clan presupposes a totem, which has its whole meaning only 
in and through the clan. 

* For this same conception, see A. Lang, Social Origins, p. 150. 

* See above, p. 63. 

170 Elementary Forms of Religious Life 

considered a close relative of the animal ; for if these two king- 
doms were clearly distinguished in the mind, men would never 
believe that a human soul could pass so easily from one into 
the other. It is even necessary that the body of the animal be 
considered its true home, for it is believed to go there as soon as 
it regains its liberty. Now while the doctrine of transmigration 
postulates this singular affinity, it offers no explanation of it. 
The only explanation offered by Tylor is that men sometimes 
resemble in certain traits the anatomy and physiology of the 
animal. " The half-human features and actions and characters 
of animals are watched with wondering sympathy by the savage, 
as by the child. The beast is the very incarnation of familiar 
qualities of man : and such names as lion, bear, fox, owl, parrot, 
viper, worm, when we apply them as epithets to men, condense 
into a word some leading features of a human life." ^ But even 
if these resemblances are met with, they are uncertain and 
exceptional ; before all else, men resemble their relatives and 
companions, and not plants and animals. Such rare and 
questionable analogies could not overcome such unanimous 
proofs, nor could they lead a man to think of himself and his 
forefathers in forms contradicted by daily experience. So this 
question remains untouched, and as long as it is not answered, 
we cannot say that totemism is explained. ^ 

Finally, this whole theory rests upon a fundamental misunder- 
standing. For Tylor as for Wundt, totemism is only a particular 
case of the cult of animals.^ But we, on the contrary, know that 

* Primitive Culture, II, p. 17. 

* Wundt, who has revived the theory of Tylor in its essential lines, has tried 
to explain this mysterious relationship of the man and the animal in a different 
way : it was the sight of the corpse in decomposition which suggested the idea, 
when they saw worms coming out of the body, they thought that the soul was 
incarnate in them and escaped with them. Worms, and by extension, reptiles 
(snakes, lizards, etc.), were therefore the first animals to serve as receptacles for 
the souls of the dead, and consequently they were also the first to be venerated 
and to play the rôle of totems. It was only subsequently that other animals and 
plants and even inanimate objects were elevated to the same dignity. But this 
hypothesis does not have even the shadow of a proof. Wundt affirms [Mythus 
und Religion, II, p. 296) that reptiles are much more common totems than other 
animals ; from this, he concludes that they are the most primitive. But we 
cannot see what justifies this assertion, in the support of which the author cites 
no facts. The lists of totems gathered either in Australia or in America do not 
show that any special species of animal has played a preponderating rôle. Totems 
vary from one region to another with the flora and fauna. Moreover, if the circle 
of possible totems was so closely limited at first, we cannot see how totemism 
was able to satisfy the fundamental principle which says that the two clans or 
sub-clans of a tribe must have two different totems. 

* " Sometimes men adore certain animals," says Tylor, " because they regard 
them as the reincarnation of the divine souls of the ancestors ; this belief is a 
sort of bridge between the cult rendered to shades and that rendered to animals " 
(Primitive Culture, II, p. 805, cf. 309, in fine). Likewise, Wundt presents 
totemism as a section of animalism (II, p. 234). 

Origins of these Beliefs 171 

it is something very different from a sort of animal- worship. ^ 
The animal is never adored ; the man is nearly its equal and 
sometimes even treats it as his possession, so far is he from being 
subordinate to it like a believer before his god. If the animals 
of the totemic species are really believed to incarnate the ancestors, 
the members of foreign clans would not be allowed to eat their 
flesh freely. In reality, it is not to the animal as such that the 
cult is addressed, but to the emblem and the image of the totem. 
Now between this religion of the emblem and the ancestor-cult, 
there is no connection whatsoever. 

While Tylor derives totemism from the ancestor-cult, Jevons 
derives it from the nature-cult, ^ and here is how he does so. 

When, under the impulse of the surprise occasioned by the 
irregularities observed in the course of phenomena, men had 
once peopled the world with supernatural beings,^ they felt the 
need of making agreements with these redoubtable forces with 
which they had surrounded themselves. They understood that 
the best way to escape being overwhelmed by them was to ally 
themselves to some of them, and thus make sure of their aid. But 
at this period of history men knew no other form of alliance 
and association than the one resulting from kinship. All the 
members of a single clan aid each other mutually because they 
are kindred or, as amounts to the same thing, because they think 
they are ; on the other hand, different clans treat each other 
as enemies because they are of different blood. So the only way 
of assuring themselves of the support of these supernatural 
beings was to adopt them as kindred and to be adopted by them 
in the same quality : the well-known processes of the blood- 
covenant permitted them to attain this result quite easily. 
But since at this period, the individual did not yet have a real 
personality, and was regarded only as a part of his group, or 
clan, it was the clan as a whole, and not the individual, which 
collectively contracted this relationship. For the same reason, 
it was contracted, not with a particular object, but with the 
natural group or species of which this object was a part ; for 
men think of the world as they think of themselves, and just 
as they could not conceive themselves apart from their clans, so 
they were unable to conceive of anything else as distinct from 
the species to which it belonged. Now a species of things united 
to a clan by a bond of kinship is, says Jevons, a totem. 

In fact, it is certain that totemism implies the close association 
of a clan to a determined category of objects. But that this 

* See above, p. 139. 

' Introduction to the History of Religions, pp. 97 ff. * See above, p. 28. 

172 Elementary Forms of Religious Life 

association was contracted with a deliberate design and in the 
full consciousness of an end sought after, as Jevons would have 
us believe, is a statement having but little harmony with what 
history teaches. Religions are too complex, and answer to needs 
that are too many and too obscure, to have their origin in a 
premeditated act of the will. And while it sins through over- 
simplicity, this hypothesis is also highly improbable. It says 
that men sought to assure themselves of the aid of the super- 
natural beings upon which things depend. Then they should 
preferably have addressed themselves to the most powerful 
of these, and to those whose protection promised to be the most 
beneficial. 1 But quite on the contrary, the beings with whom 
they have formed this mystic kinship are often among the most 
humble which exist. Also, if it were only a question of making 
allies and defenders, they would ha'^'e tried to make as many as 
possible ; for one cannot be defended too well. Yet as a matter 
of fact, each clan systematically contents itself with a single 
totem, that is to say, with one single protector, leaving the 
other clans to enjoy their own in perfect liberty. Each group 
confines itself within its own religious domain, never seeking to 
trespass upon that of its neighbours. This reserve and moderation 
are inexplicable according to the hypothesis under consideration. 


Moreover, all these theories are wrong in omitting one question 
which dominates the whole subject. We have seen that there 
are two sorts of totemism : that of the individual and that of the 
clan. There is too evident a kinship between the two for them 
not to have some connection with each other. So we may well 
ask if one is not derived from the other, and, in the case of an 
affirmative answer, which is the more primitive ; according to 
the solution accepted, the problem of the origins of totemism 
will be posed in difterent terms. This question becomes all the 
more necessary because of its general interest. Individual 
totemism is an individual aspect of the totemic cult. Then if it 
was the primitive fact, we must say that religion is born in the 
consciousness of the individual, that before all else, it answers to 
individual aspirations, and that its collective form is merely 

The desire for an undue simplicity, with which ethnologists 
and sociologists are too frequently inspired, has naturally led 
many scholars to explain, here as elsewhere, the complex by the 

1 Jevons recognizes this himself, saying, " It is to be presumed that in the 
choice of an ally he would prefer . . . the kind or species which possessed the 
greatest power " (p. loi). 

Origins of these Beliefs 173 

simple, the totem of the group by that of the individual. Such, 
in fact, is the theory sustained by Frazer in his Golden Boiigh,^ 
by Hill Tout, 2 by Miss Fletcher,^ by Boas ^ and by Swanton.^ 
It has the additional advantage of being in harmony with the 
conception of religion which is currently held ; this is quite 
generally regarded as something intimate and personal. From 
this point of view, the totem of the clan can only be an individual 
totem which has become generalized. Some eminent man, 
having found from experience the value of a totem he chose for 
himself by his own free will, transmitted it to his descendants ; 
these latter, multiplying as tune went on, finally formed the 
extended family known as a clan, and thus the totem became 

Hill Tout believes that he has found a proof supporting this 
theory in the way totemism has spread among certain societies 
of North-western America, especially among the Salish and 
certain Indians on the Thompson River. Individual totemism 
and the clan totemism are both found among these peoples ; 
but they either do not co-exist in the same tribe, or else, when 
they do co-exist, they are not equally developed. They vary 
in an inverse proportion to each other ; where the clan totem 
tends to become the general rule, the individual totem tends to 
disappear, and vice versa. Is that not as much as to say that the 
first is a more recent form of the second, which excludes it by 
replacing it ? ^ Mythology seems to confirm this interpretation. 
In these same societies, in fact, the ancestor of the clan is not a 
totemic animal ; the founder of the group is generally repre- 
sented in the form of a human being who, at a certain time, 
had entered into familiar relations with a fabulous animal from 
whom he received his totemic emblem. This emblem, together 
with the special powers which are attached to it, was then passed 
on to the descendants of this mythical hero by right of heritage. 
So these people themselves seem to consider the collective 
totem as an individual one, perpetuated in the same family.' 

^ 2nd Edition, III, pp. 416 ff. ; see especially p. 419, n. 5. In more recent 
articles, to be analysed below, Frazer exposes a different theorj', but one which 
does not. in his opinion, completely exclude the one in the Golden Bough. 

^ The Origin of the Totemism of the Aborigines of British Columbia, in Proc. 
and Transact, of the Roy. Soc. of Canada, 2nd series, VII, § 2, pp. 3 fi. Also, 
Report on the Ethnology of the Statlumh, J. A. I., XXXV, p. 141. Hill Tout has 
replies to various objections made to his theory in Vol. IX of the Transact, of the 
Roy. Soc. of Canada, pp. 61-99. 

^ Alice C. Fletcher, The Import of the Totem, in Smithsonian Report for 1897, 
pp. 577-.5y6. * The Kwakiutl Indians, pp. 323 tï., 336-338, 393. 

' The Development of the Clan System, in .kmer. .4nthrop., N.S. VI, 1904, 
pp. 477-486. « J. A. I., XXXV, p. 142. 

' Ibid., p. 150. Cf. Vth Rep. on the . . . N.W. Tribes of Canada. B.A.A.S., 
p. 24. A myth of this sort has been quoted above. 

174 Elementary Forms of Religious Life 

Moreover, it still happens to-day that a father transmits his own 
totem to his children. So if we imagine that the collective totem 
had, in a general way, this same origin, we are assuming that the 
same thing took place in the past which is still observable to-day. ^ 

It is still to be explained whence the individual totem comes. 
The reply given to this question varies with different authors. 

Hill Tout considers it a particular case of fetishism. Feeling 
himself surrounded on all sides by dreaded spirits, the individual 
experienced that sentiment which we have just seen Jevons 
attribute to the clan : in order that he might continue to exist, 
he sought some powerful protector in this mysterious world. 
Thus the use of a personal totem became established. ^ For 
Frazer, this same institution was rather a subterfuge or trick of 
war, invented by men that they might escape from certain dangers. 
It is known that according to a belie'" which is very widespread 
in a large number of inferior societ!' j, the human soul is able, 
without great inconvenience, to quit the body it inhabits for a 
while ; howsoever far away it may be, it continues to animate 
this body by a sort of detached control. Then, in certain critical 
moments, when life is supposed to be particularly menaced, it 
may be desirable to withdraw the soul from the body and lead 
it to some place or into some object where it will be in greater 
security. In fact, there are a certain number of practices whose 
object is to withdraw the soul in order to protect it from some 
danger, either real or imaginary. For example, at the moment 
when men are going to enter a newly-built house, a magician 
removes their souls and puts them in a sack, to be saved and 
returned to their proprietors after the door-sill has been crossed. 
This is because the moment when one enters a new house is 
exceptionally critical ; one may have disturbed, and consequently 
offended, the spirits who reside in the ground and especially 
under the sill, and if precautions are not taken, these could make 
a man pay dearly for his audacity. But when this danger is 
once passed, and one has been able to anticipate their anger 
and even to make sure of their favour through the accomplish- 
ment of certain rites, the souls may safely retake their accustomed 
place. ^ It is this same belief which gave birth to the personal 
totem. To protect themselves from sorcery, men thought it 
wise to hide their souls in the anonymous crowd of some species 
of animal or vegetable. But after these relations had once been 

1 J.A.I., XXXV, p. 147. 

* Proc. and Transact., eta., VII, § 2, p. 12. 

' See The Golden Bough,'^ III, pp. 351 fi. Wilkcn had already pointed out 
similar facts in De Simsonsage, in De Gids, 1890 ; De Betrekking tusschen Menschen- 
Dieren en Plantenleven, in Indische Gids, 1884, 1888 ; Ueber das Haaropfer, in 
Revue Coloniale Internationale, 1886-1887. 

Origins of these Beliefs 175 

established, each individual found himself closely united to the 
animal or plant where his own vital principle was believed to 
reside. Two beings so closely united were finally thought to be 
practically indistinguishable : men believed that each partici- 
pated in the nature of the other. When this belief had once 
been accepted, it facilitated and hastened the transformation 
of the personal totem into an hereditary, and consequently a 
collective, totem ; for it seemed quite evident that this kinship of 
nature should be transmitted hereditarily from father to child. 

We shall not stop to discuss these two explanations of the 
individual totem at length : they are ingenious fabrications of 
the mind, but they completely lack all positive proof. If we are 
going to reduce totemism to fetishism, we must first establish 
that the latter is prior to the former ; now, not merely is no fact 
brought forward to support this hypothesis, but it is even con- 
tradicted by everything that we know. The ill-determined group 
of rites going under the name of fetishism seem to appear only 
among peoples who have already attained to a certain degree of 
civilization ; but it is a species of cult unknown in Australia. It 
is true that some have described the churinga as a fetish ; ^ but 
even supposing that this qualification were justified, it would 
not prove the priority which is postulated. Quite on the contrary, 
the churinga presupposes totemism, since it is essentially an 
instrument of the totemic cult and owes the virtues attributed 
to it to totemic beliefs alone. 

As for the theory of Frazer, it presupposes a tboroughgoing 
idiocy on the part of the primitive which known facts do not 
allow us to attribute to him. He does have a logic, however 
strange this may at times appear ; now unless he were completely 
deprived of it, he could never be guilty of the reasoning imputed 
to him. Nothing could be more natural than that he should 
believe it possible to assure the survival of his soul by hiding it 
in a secret and inaccessible place, as so many heroes of myths 
and legends are said to have done. But why should he think it 
safer in the body of an animal than in his own ? Of course, if it 
were thus lost in space, it might have a chance to escape the 
spells of a magician more readily, but at the same time it would 
be prepared for the blows of hunters. It is a strange way of 
sheltering it to place it in a material form exposing it to risks at 
every instant. 2 But above all, it is inconceivable that a whole 
people should allow themselves to be carried into such an 

^ For example, Eylmann in Die Eingeborenen der Kolonie Siidaustr alien, 
p. 199. 

* Mrs. Parker says in connection with the Euahlayi, that if the Yunbeai does 
" confer exceptional force, it also exposes one to exceptional dangers, for all 
that hurts the animal wounds the man " (Euahlayi, p. 29). 

176 Ele^nentary Forms of Religious Life 

aberration. 1 Finally, in a very large number of cases, the function 
of the individual totem is very different from that assigned it by 
Frazer ; before all else, it is a means of conferring extraordinary 
powers upon magicians, hunters or warriors. ^ As to the kinship 
of the man and the thing, with all the inconveniences it implies, 
it is accepted as a consequence of the rite ; but it is not desired 
in its and for itself. 

There is still less occasion for delaying over this controversy 
since it concerns no real problem. What we must know before 
everything else is whether or not the individual totem is really 
a primitive fact, from which the collective totem was derived ; 
for, according to the reply given to this question, we must seek 
the home of the religious life in one or the other of two opposite 

Against the hypothesis of Hill Tout, Miss Fletcher, Boas and 
Frazer there is such an array of decisive facts that one is surprised 
that it has been so readily and so generally accepted. 

In the first place, we know that a man frequently has the 
greatest interest not only in respecting, but also in making his 
companions respect the species serving him as personal totem ; 
his own life is connected with it. Then if collective totemism 
were only a generalized form of individual totemism, it too 
should repose upon this same principle. Not only should the 
men of a clan abstain from killing and eating their totem-animal 
themselves, but they should also do all in their power to force 
this same abstention upon others. But as a matter of fact, far 
from imposing such a renunciation upon the whole tribe, each 
clan, by rites which we shall describe below, takes care that the 
plant or animal whose name it bears shall increase and prosper, 

^ In a later work (The Origin of Totemism, in The Fortnightly Review, May, 
1899, pp. 844-845), Frazer raises this objection himself. " If," he says, " I 
deposit my soul in a hare, and my brother John (a member of another clan) shoots 
that hare, roasts and swallows it, what becomes of my soul ? To meet this 
obvious danger it is necessary that John should know the state of my soul, and 
that, knowing it, he should, whenever he shoots a hare, take steps to extract and 
restore to me my soul before he cooks and dines upon the animal." Now Frazer 
believes that he has found this practice in use in Central Australia. Every year, 
in the course of a ceremony which we shall describe presently, when the animals 
of the new generation arrive at maturity, the first game to be killed is presented 
to men of that totem, who eat a little of it ; and it is only after this that the 
men of the other clans may eat it freely. This, says Frazer, is a way of returning 
to the former the souls they may have confided to these animals. But, aside from 
the fact that this interpretation of the fact is wholly arbitrary, it is hard not to 
find this way of escaping the danger rather peculiar. This ceremony is annual ; 
long days may have elapsed since the animal was killed. During all this time, 
what has become of the soul which it sheltered and the individual whose life 
depended on this soul ? But it is superfluous to insist upon all the inconceivable 
things in this explanation. 

^ Parker, op. cit., p. 20 ; Howitt, Australian Medicine Men, in J.A.I., XVI, 
pp. 34, 49 f. ; Hill Tout. J.A.I., XXXV. p. 146. 

Origins of these Beliefs 177 

so as to assure an abundant supply of food for the other clans. 
So we must at least admit that in becoming collective, individual 
totemism was transformed profoundly, and we must therefore 
account for this transformation. 

In the second place, how is it possible to explain, from this 
point of view, the fact that except where totemism is in full 
decay, two clans of a single tribe always have different totems ? 
It seems that nothing prevents two or several members of a 
single tribe, even when there is no kinship between them, from 
choosing their personal totem in the same animal species and 
passing it on to their descendants. Does it not happen to-day 
that two distinct families have the same name ? The carefully 
regulated way in which the totems and sub-totems are divided 
up, first between the two phratries and then among the various 
clans of the phratry, obviously presupposes a social agreement 
and a collective organization. This is as much as to say that 
totemism is something more than an individual practice spon- 
taneously generalized. 

Moreover, collective totemism cannot be deduced from indi- 
vidual totemism except by a misunderstanding of the differences 
separating the two. The one is acquired by the child at birth ; 
it is a part of his civil status. The other is acquired during the 
course of his life ; it presupposes the accomplishment of a 
determined rite and a change of condition. Some seek to diminish 
this distance by inserting between the two, as a sort of middle 
term, the right of each possessor of a totem to transmit it to 
whomsoever he pleases. But wherever these transfers do take 
place, they are rare and relatively exceptional acts ; they cannot 
be performed except by magicians or other personages invested 
with special powers ; ^ in any case, they are possible only through 
ritual ceremonies which bring about the change. So it is necessary 
to explain how this prerogative of a few became the right of all ; 
how that which at first implied a profound change in the religious 
and moral constitution of the individual, was able to become an 
element of this constitution ; and finally, how a transmission 
which at first was the consequence of a rite was later believed to 
operate automatically from the nature of things and without the 
intervention of any human will. 

In support of his interpretation, Hill Tout claims that certain 
myths give the totem of the clan an individual origin : they 
tell how the totemic emblem was acquired by some special 
individual, who then transmitted it to his descendants. But in 

^ According to Hill Tout himself, " The gift or transmission (of a personal 
totem) can only be made or effected by certain persons, such as shamans, or 
those who possess great mystery power " (J. A. I., p. 146). Cf. Langloh Paricer, 
op. cit., pp. 29-30. 

178 Elementary Forms of Religious Life 

the first place, it is to be remarked that these myths are all taken 
from the Indian tribes of North America, which are societies 
arrived at a rather high degree of culture. How could a 
mythology so far removed from the origins of things aid in 
reconstituting the primitive form of an institution with any 
degree of certainty ? There are many chances for intermediate 
causes to have gravely disfigured the recollection which these 
people have been able to retain. Moreover, it is very easy to 
answer these myths with others, which seem much more primitive 
and whose signification is quite different. The totem is there 
represented as the very being from whom the clan is descended. 
So it must be that it constitutes the substance of the clan ; men 
have it within them from their birth ; it is a part of their very 
flesh and blood, so far are they from having received it from 
without^^ More than that, the very myths upon which Hill 
Tout relies contain an echo of this ancient conception. The 
founder who gave his name to the clan certainly had a human 
form ; but he was a man who, after living among animals of a cer- 
tain species, finally came to resemble them. This is undoubtedly 
because a. time came when the mind was too cultivated to admit 
any longer, as it had formerly done, that men might have been 
born of animals ; so the animal ancestor, now become incon- 
ceivable, is replaced by a human being ; but the idea persists 
that this man had acquired certain characteristics of the animal 
either by imitation or by some other process. Thus even this late 
mythology bears the mark of a more remote epoch when the totem 
of the clan was never regarded as a sort of individual creation. 

But this hypothesis does not merely raise grave logical diffi- 
culties ; it is contradicted directly by the following facts. 

If individual totemism were the initial fact, it should be more 
developed and apparent, the more primitive the societies are, 
and inversely, it should lose ground and disappear before the 
other among the more advanced peoples. Now it is the contrary 
which is true. The Australian tribes are far behind those of 
North America ; yet Australia is the classic land of collective 
totemism. In the great majority of the tribes, it alone is found, 
while we do not know a single one where individual totemism alone 
is practised.^ This latter is found in a characteristic form only 
in an infinitesimal number of tribes.^ Even where it is met with 

* Cf. Hartland, Totemism and some recent Discoveries, in Folk-Lore, XI, 
pp. 59 flf. 

* Except perhaps the Kurnai ; but even in this tribe, there are sexual totems 
in addition to the personal ones. 

* Among the Wotjobaluk, the Buandik, the Wiradjuri, the Yuin and the 
tribes around Maryborough (Queensland). See Howitt, Nat. Tr., pp. 114-147; 
MathevHS, /. of the R. Soc. of N.S. Wales, XXXVIII, p. 291. Cf. Thomas. 
Further Notes on Mr. Hill Tout's Views on Totemism, in Man, 1904, p. 85. 

Origins of these Beliefs 179 

it is generally in a rudimentary form. It is made up of individual 
and optional practices having no generality. Only magicians 
are acquainted with the art of creating mysterious relationships 
with species of animals to which they are not related by nature. 
Ordinary people do not enjoy this privilege. '^ In America, on 
the contrary, the collective totem is in full decadence ; in the 
societies of the North-west especially, its religious character is 
almost gone. Inversely, the individual totem plays a considerable 
rôle among these same peoples. A very great efficacy is attributed 
to it ; it has become a real public institution. This is because 
it is the sign of a higher civilization. This is undoubtedly the 
explanation of the inversion of these two forms of totemism, 
which Hill Tout believes he has observed among the Salish. 
If in those parts where collective totemism is the most fully 
developed the other form is almost lacking, it is not because the 
second has disappeared before the first, but rather, because the 
conditions necessary for its existence have not yet been fully 

But a fact which is still more conclusive is that individual 
totemism, far from having given birth to the totemism of the 
clan, presupposes this latter. It is within the frame of collective 
totemism that it is born and lives : it is an integral part of it. 
In fact, in those very societies where it is preponderating, the 
novices do not have the right of taking any animal as their 
individual totem ; to each clan a certain definite number of 
species are assigned, outside of which it may not choose. In 
return, those belonging to it thus are its exclusive property ; 
members of other clans may not usurp them.^ They are thought 
to have relations of close dependence upon the one serving as 
totem to the clan as a whole. There are even cases where it is 
quite possible to observe these relations : the individual aspect 
represents a part or a particular aspect of the collective totem. ^ 
Among the Wotjobaluk, each member of the clan considers the 

* This is the case with the Euahlayi and the facts of personal totemism cited by 
Howitt, Austrahan Medicine Men, in J. A. I., XVI, pp. 34, 35, 49-50. 

* Miss Fletcher, A Study of the Omaha Triffe, in Smithsonian Report for 1897, 
p. 586; Boas, The Kwakiutl, p. 322. Likewise, Vth Rep. of the Committee . . . 
of the N.W. Tribes of the Dominion of Canada, B.A.A.S., p. 25 ; Hill Tout, J. A. I., 
XXXV, p. 148. 

' The proper names of the génies, says Boas in regard to the Tlinkit, are 
derived from their respective totems, each gens having its special names. The 
connection between the name and the (collective) totem is not very apparent 
sometimes, but it always exists (Vth Rep. of the Committee, etc., p. 25). The fact 
that individual forenames are the property of the clan, and characterize it as 
surely as the totem, is also found among the Iroquois (Morgan, Ancient Society, 
p. 78), the Wyandot (Powell, Wyandot Government, in 1st Rep., p. 59), the 
Shawnee, Sauk and Fox (Morgan, Ancient Society, pp. 72, 76-77) and the Omaha 
(Dorsey, Omaha Sociology, in Ilird Rep., pp. 227 fï.). Now the relation between 
forenames and personal totems is already known (see above, p. 157). 

i8o Elementary Forms of Religious Life 

personal totems of his companions as being his own after a fashion ; ^ 
so they are probably sub-totems. Now the sub-totem supposes 
the totem, as the species supposes the class. "Thus the first form 
of individual religion met with in history appears, not as the 
active principle of all public rehgion, but, on the contrary, 
as a simple aspect of this latter. The cult which the individual 
organizes for himself in his own inner conscience, far from being 
the germ of the collective cult, is only this latter adapted to the 
personal needs of the individual. 


In a more recent study, ^ which the works of Spencer and 
Gillen suggested to him, Frazer has attempted to substitute a 
new explanation of totemism for the one he first proposed, 
and which we have just been discussing. It rests on the postulate 
that the totemism of the Arunta is the most primitive which we 
know ; Frazer even goes so far as to say that it scarcely differs 
from the really and absolutely original type.^ 

The singular thing about it is that the totems are attached 
neither to persons nor to determined groups of persons, but to 
localities. In fact, each totem has its centre at some definite 
spot. It is there that the souls of the first ancestors, who founded 
the totemic group at the beginning of time, are believed to have 
their preferred residence. It is there that the sanctuary is located 
where the churinga are kept ; there the cult is celebrated. It is 
also this geographical distribution pf totems which determines 
the manner in which the clans are recruited. The child has 
neither the totem of his father nor that of his mother, but the 
one whose centre is at the spot where the mother believes that 
she felt the first symptoms of approaching maternity. For it is 
said that the Arunta is ignorant of the exact relation existing 
between generation and the sexual act ; ^ he thinks that every 

^ " For example," says Mathews, " if you ask a Wartwurt man what totem 
he is, he will first tell his personal totem, and will probably then enumerate 
those of his clan " (Jour, of the Roy. Soc. of N.S. Wales, XXXVIII, p. 291). 

" The Beginnings of Religion and Totemism among the Australiayi Aborigines, 
in Fortnightly Review, July, 1905, pp. 162 lï., and Sept., p. 452. Cf. the same 
author, The Origin of Totemism, ibid., April, 1899, p. 648, and May, p. 835. These 
latter articles, being slightly older, differ from the former on one point, but the 
foundation of the theory is not essentially dilïerent. Both are reproduced in 
Totemism and Exogamy, I, pp. 89-172. In the same sense, see Spencer and 
Gillen, Some Remarks on 'Totemism as applied to Australian Tribes, in J. A. I., 
1899, pp. 275-280, and the remarks of Frazer on the same subject, ibid., pp. 

^ " Perhaps we may . . . say that it is but one remove from the original 
pattern, the absolutely original form of totemism " (Fortnightly Review, Sept., 

1905, P- 455)- 

* On this point, the testimony of Strehlow (II, p. 52) confirms that of Spencer 
and Gillen. For a contrary opinion, see A. Lang, The Secret of the Totem, p. 190. 

Origins of these Beliefs i8i 

conception is due to a sort of mystic fecundation. According 
to him, it is due to the entrance of the soul of an ancestor into 
the body of a woman and its becoming the principle of a new 
life there. So at the moment when a woman feels the first 
tremblings of the child, she imagines that one of the souls whose 
principal residence is at the place where she happens to be, has 
just entered into her. As the child who is presently born is merely 
the reincarnation of this ancestor, he necessarily has the same 
totem ; thus his totem is determined by the locality where he is 
believed to have been mysteriously conceived. 

Now, it is this local totemism which represents the original 
form of totemism ; at most, it is separated from this by a very 
short step. This is how Frazer explains its genesis. 

At the exact moment when the woman realizes that she is 
pregnant, she must think that the spirit by which she feels 
herself possessed has come to her from the objects about her, 
and especially from one of those which attract her attention at 
the moment. So if she is engaged in plucking a plant, or watching 
an animal, she believes that the soul of this plant or animal has 
passed into her. Among the things to which she will be particu- 
larly inchned to attribute her condition are, in the first place, the 
things she has just eaten. If she has recently eaten emu or yam, 
she will not doubt that an emu or yam has been born in her and 
is developing. Under these conditions, it is evident how the 
child, in his turn, will be considered a sort of yam or emu, how 
he regards himself as a relative of the plant or animal of the 
same species, how he has sympathy and regard for them, how he 
refuses to eat them, etc.^ From this moment, totemism exists 
in its essential traits : it is the native's theory of conception 
that gave rise to it, so Frazer calls this primitive totemism 

It is from this original type that all the other forms of totemism 
are derived. " When several women had, one after the other, 
felt the first premonitions of maternity at the same spot and 
under the same circumstances, the place would come to be re- 
garded as haunted by spirits of a pecuHar sort ; and so the 
whole country might in time be dotted over with totem centres 
and distributed into totem districts." - This is how the local 
totemism of the Arunta originated. In order that the totems 

^ A very similar idea had already been expressed by Haddon in his Address 
to the Anthropological Section (B.A.A.S., 1902, pp. 8 ff.). He supposes that at 
first, each local group had some food which was especially its own. The plant or 
animal thus serving as the principal item of food became the totem of the group. 

All these explanations naturally imply that the prohibitions against eating 
the totcmic animal were not primitive, but were even preceded by a contrary 

* Fortnightly Review, Sept., 1905, p. 458. 

i82 Elementary Forms of Religious Life 

may subsequently be detached from their territorial base, it is 
sufficient to think that the ancestral souls, instead of remaining 
immutably fixed to a determined spot, are able to move freely 
over the surface of the territory and that in their voyages they 
follow the men and women of the same totem as themselves. 
In this way, a woman may be impregnated by her own totem 
or that of her husband, though residing in a different totemic 
district. According to whether it is believed that it is the ancestor 
of the husband or of the wife who thus follow the family about, 
seeking occasions to reincarnate themselves, the totem of the 
child will be that of his father or mother. In fact, it is in just 
this way that the Guanji and Umbaia on the one hand, and the 
Urabunna on the other, explain their systems of filiation. 

But this theory, like that of Tylor, rests upon a begging of the 
question. If he is to imagine that human souls are the souls of 
animals or plants, one must believe beforehcind that men take 
either from the animal or vegetable world whatever is most 
essential in them. Now this belief is one of those at the founda- 
tion of totemism. To state it as something evident is therefore 
to take for granted that which is to be explained. 

Moreover, from this point of view, the religious character of 
the totem is entirely inexplicable, for the vague belief in an 
obscure kinship between the man and the animal is not enough 
to found a cult. This confusion of distinct kingdoms could never 
result in dividing the world into sacred and profane. It is true 
that, being consistent with himself, Frazer refuses to admit that 
totemism is a religion, under the pretext that he finds in it neither 
spiritual beings, nor prayers, nor invocations, nor offerings, etc. 
According to him, it is only a system of magic, by which he means 
a sort of crude and erroneous science, a first effort to discover the 
laws of things.^ But we know how inexact this conception, 
both of magic and of religion, is. We have a religion as soon 
as the sacred is distinguished fronTThe profane, and we have 
seen that totemism is a-jy^st system of sacred things. If we are 
to explain it, we must therefore show how it happened that these 
things were stamped with this character. ^ But he does not 
even raise this problem. 

But this system is completely overthrown by the fact that 
the postulate upon which it rests can no longer be sustained. 
The whole rgument of Frazer supposes that the local totemism 
of the Arunta is the most primitive we know, and especially 

* Fortn. Rev., May, 1899, p. 835, and July, 1905, pp. 162 ff. 

* Though considering totemism only a system of magic, Frazer recognizes 
that the first germs of a real religion are sometimes found in it {Fortn. Rev., July, 
1905. P- 163)- O ' the way in which he thinks religion developed out of magic, 
see The Golden Bough,^ I, pp. 75-78. 

Origins of these Beliefs 183 

that it is clearly prior to hereditary totemism, either in the 
paternal or the maternal line. Now as soon as the facts contained 
in the first volume of Spencer and Gillen were. at our disposal, 
we were able to conjecture that there had been a time in the 
history of the Arunta people when the totems, instead of being 
attached to localities, were transmitted hereditarily from mother 
to child. ^ This conjecture is definitely proved by the new 
facts discovered by Strehlow,^ which only confirm the previous 
observations of Schulze.® In fact, both of these authors tell us 
that even now, in addition to his local totem, each Arunta has 
another which is completely independent of all geographical 
conditions, and which belongs to him as a birthright : it is his 
mother's. This second totem, just like the first, is considered 
a powerful friend and protector by the natives, which looks 
after their food, warns them of possible dangers, etc. They 
have the right of taking part in its cult. When they are buried, 
the corpse is laid so that the face is turned towards the region 
of the maternal totemic centre. So after a fashion this centre 
is also that of the deceased. In fact it is given the name tmara 
altjira, which is translated : camp of the totem which is associated 
with me. So it is certain that among the Arunta, hereditary 
totemism in the uterine line is not later than local totemism, 
but, on the contrary, must have preceded it. For to-day, the 
maternal totem has only an accessory and supplementary rôle ; 
it is a second totem, which explains how it was able to escape 
observation as attentive and careful as that of Spencer and 
Gillen. But in order that it should be able to retain this secondary 
place, being employed along with the local totem, there must 
have been a time when it held the primary place in the religious 
life. It is, in part, a fallen totem, but one recalling an epoch when 
the totemic organization of the Arunta was very different from 
what it is to-day. So the whole superstructure of Frazer's system 
is undermined at its foundation.* 

* Sur le totémisme, in Année Soc, V. pp. 82-121. Cf.. on this same question, 
Hartland, Presidential Address, in Fçlk-Lore, XI, p. 75 ; A. Lang, A Theory of 
Arunta Totemism, in Man, 1904, No. 44 ; Concepiional Totemism and Exogamy, 
ibid., 1907, No. 55 ; The Secret of the Totem, ch. iv ; N. W. Thomas, Arunta 
Totemism, in Maft. 1904, No. 68 ; P. W. Schmidt, Die Stellung der Aranda unter 
der Australischen Stdmmen, in Zeitschrift fur Ethnologie, 1908, pp. 866 fi. 

» LHe Aranda, II, pp. 57-58. 3 Schulze, loc. cit., pp. 238-239. 

* In the conclusion of Totemism and Exogamy {IV , pp. 58-59), Frazer says. 
it must be admitted, that there is a totemism still more ancient than that of thé 
Arunta : it is the one observed by Rivers in the Banks Islands (Totemism in 
Polynesia and Melanesia, in J.A.I., XXXIX, p. 172). Among the Arunta it is 
the spirit of an ancestor who is believed to impregnate the mother ; in the Banks 
Islands, it is the spirit of an animal or vegetable, as the theory supposes. But 
as the ancestral spirits of the Arunta have an animal or vegetable form, the 
difference is slight. Therefore we have not mentioned it in our exposition. 

184 Elementary Forms of Religious Life 


Although Andrew Lang has actively contested this theory 
of Frazer's, the one he proposes himself in his later works,* 
resembles it on more than one point. Like Frazer, he makes 
totemism consist in the belief in a sort of consubstantiality of 
the man and the animal. But he explains it differently. 

He derives it entirely from the fact that the totem is a name. 
As soon as human groups were founded, ^ each one felt the need 
of distinguishing between the neighbouring groups with which it 
came into contact and, with this end in view, it gave them different 
names. The names were preferably chosen from the surrounding 
flora and fauna because animals and plants can easily be desig- 
nated by movements or represented by drawings. ^ The more 
or less precise resemblances which men may have with such and 
such objects determined the way in which these collective 
denominations were distributed among the groups.^ 

Now, it is a well-known fact that " to the early mind names, 
and the things known by names, are in a mystic and transcendental 
connection of rapport." ^ For example, the name of an individual 
is not considered as a simple word or conventional sign, but as 
an essential part of the individual himself. So if it were the 
name of an animal, the man would have to believe that he 
himself had the most characteristic attributes of this same animal. 
This theory would become better and better accredited as the 
historic origins of these denominations became more remote 
and were effaced from the memory. Myths arose to make this 
strange ambiguity of human nature more easily representable 
in the mind. To explain this, they imagined that the animal 
was the ancestor of the men, or else that the two were descended 
from a common ancestor. Thus came the conception of bonds 
of kinship uniting each clan to the animal species whose name 
it bore. With the origins of this fabulous kinship once explained, 
it seems to our author that totemism no longer contains a 

' Social Origins, London, 1903, especially ch. viii, entitled The Origin of 
Totem Names and Beliefs, and The Secret of the Totem, London, 1905. 

* In his Social Origins especially, Lang attempts to reconstitute by means of 
conjecture the form which these primitive groups should have ; but it seems 
superfluous to reproduce these hypotheses, which do not affect his theory of 

* On this point, Lang approaches the theory of Julius Pickler (see Pickler and 
Szomolo, Der Ursprung des Totemismus. Ein Beitrag sur materialistirchen 
Geschichtstheorie. Berlin, 36 pp. in 8vo). The difference between the two hypo- 
theses is that Pickler attributes a higher importance to the pictorial representation 
of the name than to the name itself. * Social Origins, p. 166. 

* The Secret of the Totem, p. 121 ; cf. pp. 116, 117. 

Origins of these Beliefs 185 

But whence comes the rehgious character of the totemic 
behefs and practices ? For the fact that a man considers himself 
an animal of a certain species does not explain why he attributes 
marvellous powers to this species, and especially why he renders 
a cult to the images symbolizing it. — To this question Lang gives 
the same response as Frazer : he denies that totemism is a religion-. 
" I find in Australia," he says, " no example of religious practices 
such as praying to, nourishing or burying the totem." ^ It was 
only at a later epoch, when it was already established, that 
totemism was drawn into and surrounded by a system of con- 
ceptions properly called religious. According to a remark of 
Howitt,^ when the natives undertake the explanation of the 
totemic institutions, they do not attribute them to the totems 
themselves nor to a man, but to some supernatural being such 
as Bunjil or Baiame. " Accepting this evidence," says Lang, 
" one source of the ' religious ' character of totemism is at once 
revealed. The totemist obeys the decree of Bunjil, or Baiame, as 
the Cretans obeyed the divine decrees given by Zeus to Minos." 
Now according to Lang the idea of these great divinities arose 
outside of the totemic system ; so this is not a religion in itself ; 
it has merely been given a religious colouring by contact with a 
genuine religion. 

But these very myths contradict Lang's conception of totemism. 
If the Australians had regarded totemism as something human 
and profane, it would never have occurred to them to make a 
divine institution out of it. If, on the other hand, they have 
felt the need of connecting it with a divinity, it is because they 
have seen a sacred character in it. So these mythological inter- 
pretations prove the religious nature of totemism, but do not 
explain it. 

Moreover, Lang himself recognizes that this solution is not 
sufficient. He realizes that totemic things are treated with a 
religious respect ; ^ that especially the blood of an animal, as 
well as that of a man, is the object of numerous interdictions, or, 
as he says, taboos which this comparatively late mythology 
cannot explain.^ Then where do they come from ? Here are 
the words with which Lang answers this question : "As soon as 
the animal-named groups evolved the universally diffused 
beliefs about the wakan or mana, or mystically sacred quality 
of the blood as the life, they would also develop the various 
taboos." ^ The words wakan and mana, as we shall see in the 

' The Secret of the Totem, p. 136. 

* J.A.I., Aug.. 1888, pp. 53-54 ; cf. Nat. Tr., pp. 89, 488, 498. 

* " With reverence," as Lang says (The Secret 0/ the Totem, p. iii). 

* Lang adds that these taboos are the basis of exogamic practices. 

* Ibid., p. 125. 

i86 Elementary Forms of Religions Life 

following chapter, involve the very idea of sacredness itself ; the 
one is taken from the language of the Sioux, the other from that 
of the Melanesian peoples. To explain the sacred character of 
totemic things by postulating this characteristic, is to answer the 
question by the question. What we must find out is whence this 
idea of wakan comes and how it comes to be applied to the totem 
and all that is derived from it. As long as these two questions 
remain unanswered, nothing is explained. 

We have now passed in review all the principal explanations 
which have been given for totemic beliefs,^ leaving to each of 
them its own individuality. But now that this examination is 
finished, we may state one criticism which addresses itself to all 
these systems alike. 

If we stick to the letter of the formul?e, it seems that these 
may be arranged in two groups. Some (Frazer, Lang) deny the 
religious character of totemism ; in reality, that amounts to 
denying the facts. Others recognize this, but think that they 
can explain it by deriving it from an anterior religion out of 
which totemism developed. But as a matter of fact, this dis- 
tinction is only apparent : the first group is contained within the 
second. Neither Frazer nor Lang have been able to maintain 
their principle systematically and explain totemism as if it were 
not a religion. By the very force of facts, they have been com- 
pelled to slip ideas of a religious nature into their explanations. 
We have just seen how Lang calls in the idea of sacredness, which 
is the cardinal idea of all religion. Frazer, on his side, in each of 
the theories which he has successively proposed, appeals openly 
to the idea of souls or spirits ; for according to him, totemism 
came from the fact that men thought they could deposit their 
souls in safety in some external object, or else that they attributed 
conception to a sort of spiritual fecundation of which a spirit 
was the agent. Now a soul, and still more, a spirit, are sacred 
things and the object of rites ; so the ideas expressing them are 
essentially religious and it is therefore in vain that Frazer makes 
totemism a mere system of magic, for he succeeds in explaining 
it only in the terms of another religion. 

We have already pointed out the insufficiencies of animism 
and naturism ; so one may not have recourse to them, as Tylor 

1 However, we have not spoken of the theory of Spencer. But this is because 
it is only a part of his general theory of the transformation of the ancestor-cult 
into the nature-cult. As we have described that already, it is not necessary to 
repeat it. 

Origins of these Beliefs 187 

and Jevons do, without exposing himself to these same objections. 
Yet neither Frazer nor Lang seems to dream of the possibihty 
of another hypothesis. ^ On the other hand, we know that totemism 
is tightly bound up with the most primitive social system which 
we know, and in all probability, of which we can conceive. To 
suppose that it has developed out of another religion, differing 
from it only in degree, is to leave the data of observation and 
enter into the domain of arbitrary and unverifiable conjectures. 
If we wish to remain in harmony with the results we have already 
obtained, it is necessary that while affirming the religious nature 
of totemism, we abstain from deriving it from another different 
religion. There can be no hope of assigning it non-religious ideas 
as its cause. But among the representations entering into the 
conditions from which it results, there may be some which directly 
suggest a religious nature of themselves. These are the ones we 
must look for. 

* Except that Lang ascribes another source to the idea of the great gods : as 
we have already said, he believes that this is due to a sort of primitive revelation. 
But Lang does not make use of this idea in his explanation of totemism. 




The Notion of the Totemic Principle, or Mana, and the 
Idea of Force 

SINCE individual totemism is later than the totemism of the 
clan, and even seems to be derived from it, it is to this 
latter form that we must turn first of all. But as the analysis 
which we have just made of it has resolved it into a multiplicity 
of beliefs which may appear quite heterogeneous, before going 
farther, we must seek to learn what makes its unity. 

We have seen that totemism places the figured representations 
of the totem in the first rank of the things it considers sacred ; 
next come the animals or vegetables whose name the clan bears, 
and finally the members of the clan. Since all these things are 
sacred in the same way, though to different degrees, their re- 
ligious character can be due to none of the special attributes 
distinguishing them from each other. If a certain species of 
animai or vegetable is the object of a reverential fear, this is not 
because of its special properties, for the human members of the 
clan enjoy this same privilege, though to a slightly inferior 
degree, while the mere image of this same plant or animal in- 
spires an even more pronounced respect. The similar sentiments 
inspired by these different sorts of things in the mind of the 
believer, which give them their sacred character, can evidently 
come only from some common principle partaken of alike by the 
totemic emblems, the men of the clan and the individuals of the 
species serving as totem. In reality, it is to this common principle 
that the cult is addressed. In other words, totemism is the religion, 
not of such and such animals or men or images, but of an anony- 
mous and impersonal force, found in each of these beings but 
not to be confounded with any of them. No one possesses it 
entirely and all participate in it. It is so completely independent 
of the particular subjects in whom it incarnates itself, that it 
precedes them and survives them. Individuals die, generations 

1 88 

Origins of these Beliefs 189 

pass and are replaced by others ; but this force always remains 
actual, living and the same. It animates the generations of to- 
day as it animated those of yesterday and as it will animate 
those of to-morrow. Taking the words in a large sense, we may 
say that it is the god adored by each totemic cult. Yet it is an 
impersonal god, without name or history, immanent in the world 
and diffused in an innumerable multitude of things. 

But even now we have only an imperfect idea of the real 
ubiquity of this guasi-divine entity. It is not merely found in 
the whole totemic species, the whole clan and all the objects 
symbolizing the totem : the circle of its action extends beyond 
that. In fact, we have seen that in addition to the eminently 
holy things, all those attributed to the clan as dependencies of 
the principal totem have this same character to a certain degree. 
They also have something religious about them, for some are 
protected by interdictions, while others have determined functions 
in the ceremonies of the cult. Their religiousness does not differ 
in kind from that of the totem under which they are classified ; 
it must therefore be derived from the same source. So it is 
because the totemic god — to use again the metaphorical expres- 
sion which we have just employed — is in them, just as it is in the 
species serving as totem and in the men of the clan. We may 
see how much it differs from the beings in which it resides from 
the fact that it is the soul of so many different beings. 

But the Australian does not represent this impersonal force 
in "an abstract form. Under the influence of causes which we 
must seek, he has been led to conceive it under the form of an 
animal or vegetable species, or, in a word, of a visible object. 
This is what the totem really consists in : it is only the material 
form under which the imagination represents this immaterial 
substance, this energy diffused through all sorts of heterogeneous 
things, which alone is the real object of the cult. We are now 
in a better condition for understanding what the native means 
when he says that the men of the Crow phratry, for example, are 
crows. He does not exactly mean to say that they are crows in 
the vulgar and empiric sense of the term, but that the^same 
principle is found in all of them, which is their most essential 
characteristic, which they have in common with the animals of 
the same name and which is thought of under the external form of 
a crow. Thus the universe, as totemism conceives it, is filled and 
animated by a certain number of forces which the imagination 
represents in forms taken, with only a few exceptions, from the 
animal or vegetable kingdoms : there are as many of them as there 
are clans in the tribe, and each of them is also found in certain 
categories of things, of which it is the essence and vital principle. 

igo Elementary Forms of Religious Life 

When we say that these principles are forces, we do not take 
the word in a metaphorical sense ;they act just like veritable 
forces. In one sense, they are even material forces which mechani- 
cally engender physical effects. JDoes an individual come in con- 
tact with them without having taken proper precautions ? He 
receives a shock which might be compared to the effect of an 
electric discharge. Sometimes they seem to conceive of these 
as a sort of fluid escaping by points.^ If they are introduced 
into an organism not made to receive them, they produQe_sick- 

jiess andjdeatii by a wholly automatic action. ^ Outside of men, 
they play the rôle of vital principle ; it is by acting on them, 
we shall see,^ that the reproduction of the species is assured. It 
is upon them that the universal life reposes. 

But in addition to this j)hysical aspect, they also have a moral 
character. When someone asks a native why he observes hiS" 
rites, he replies that his ancestors always have observed them, 
and he ought to follow their example.'* So if he acts in a certain 
way towards the totemic beings, it is not only because the forces 
resident in them are physically redoubtable, but because he feels 
himself morally obliged to act thus ; he has the feeling that he 

-4s obeying an imperative, that he is fullilling a duty. For these 
sacred beings, he has not merely fear, but also respect. More- 
over, the totem is the source of the moral life of the clan. All the 
beings partaking of the same totemic principle consider that 
owing to this very fact, they are morally bound to one another ; 
they have definite duties of assistance, vendetta, etc., towards 
each other ; and it is these duties which constitute kinship. 
So while the totemic principle is a totemic force, it is also a moral 
power ; so we shall see how it easily transforms itself into a 
divinity properly so-called. 

Moreover, there is nothing here which is special to totemism. 
Even in the most advanced religions, there is scarcely a god who 
has not kept something of this ambiguity and whose functions 
are not at once cosmic and moral. At the same time that it is 
a spiritual discipline, every religion is also a means enabling 
men to face the world with greater confidence. Even for the 
Christian, is not God the Father the guardian of the physical 
order as well as the legislator and the judge of human conduct ? 

* For example, in a Kwakiutl myth, an ancestral hero pierces the head of 
an enemy by pointing a finger at him (Boas, Vth Rep. on the North. Tribes of 
Canada, B.A.A.S., 1889, p. 30). 

* References supporting this assertion will be found on p. 128, n. i, and 
p. 320, n. I. 

» See Bk. Ill, ch. ii. 

* See, for example, Hewitt, Nat. Tr., p. 482 ; Schiirmann, The Aboriginal 
Tribes of Port Lincoln, in Woods, Nat. Tr. of S. Australia, p. 231. 

Origins of these Beliefs 191 


Perhaps someone will ask whether, in interpreting totemism 
thus, we do not endow the native with ideas surpassing the 
limits of his intellect. Of course we are not prepared to affirm 
thaThe represents these forces wi th the relativ e clarity which we 
have been able to give to them in our analysTsT We' are able 
to show quite clearly that this notion is implied by the whole 
system of beliefs which it dominates ; but we are unable to say 
how -far it is conscious and how far, on the contrary, it is only ■ 
implicit and confusedly felt. There is no way of determining ]\ist^;;yfj,^^-\j 
what degree of clarity an idea like this may have in obscure / ^ 
minds . Blit it is well shown, in any case, that this in no way J'^^ 
surpasses the capacities of the primitive mind, and on the con- 
trary, the results at which we have just arrived are confirmed 
by the fact that either in the societies closely related to these 
Australian tribes, or even in these tribes themselves, we find, in 
an explicit form, conceptions which differ from the preceding 
only by shades and degrees. 

The native religions of Samoa have certainly passed the 
totemic phase. Real gods are found there, who have their own 
names, and, to a certain degree, their own personal physiognomy. 
Yet the traces of totemism are hardly contestable. In fact, each 
god is attached to a group, either local or domestic, just as the 
totem is to its clan.^ Then, each of these gods is thought of as 
immanent in a special species of animal. But this does not mean 
that he resides in one subject in particular : he is immanent in 
all at once ; he is diffused in the species as a whole. When an 
animal dies, the men of the group who venerate it weep for it 
and render pious duties to it, because a god inhabits it ; but the 
god is not dead. He is eternal, like the species. He is not even 
confused with the present generation ; he has already been the 
soul of the preceding one, as he will be the soul of the one which 
is to follow. 2 So he has all the characteristics of the totemic 
principle. He is the totemic principle, re-clothed in a slightly 
personal form by the imagination. But still, we must not 
exaggerate a personality which is hardly reconcilable with this 
diflusion and ubiquity. If its contours were clearly defined, it 
could never spread out thus and enter into such a multitude of 

* Frazer has even taken many facts from Samoa which he presents as really 
totemic (See Totemism, pp. 6, 12-15, 24, etc.). It is true that we have charged 
Frazer with not being critical enough in the choice of his examples, but so many 
examples would obviously have been impossible if there had not really been 
important survivals of totemism in Samoa. 

* See Turner, Samoa, p. 21 and ch. iv and v. 

192 Elementary Forms of Religious Life 

However, it is incontestable that in this case the idea of an 
impersonal religious force is beginning to change ; but there 
are other cases where it is affirmed in all its abstract purity and 
even reaches a higher degree of generality than in Australia. 
If the different totemic principles to which the various clans of 
a single tribe address themselves are distinct from each other, 
they are, none the less, comparable to each other at bottom ; 
for all play the same rôle in their respective spheres. There are 
societies which have had the feeling of this unity with nature 
and have consequently advanced to the idea of a unique religious 
force of which all other sacred principles are only expressions and 
which makes the unity of the universe. As these societies are 
still thoroughly impregnated with totemism, and as they remain 
entangled in a social organization identical with that of the 
Australians, we may say that totemism contained this idea in 

This can be observed in a large number of American tribes, 
especially those belonging to the great Sioux family : the Omaha, 
Ponka, Kansas, Osage, Assiniboin, Dakota, Iowa, Winnebago, 
Mandan, Hidatsa, etc. Many of these are still organized in clans, 
as the Omaha^ and the Iowa ;- others were so not long since, 
and, says Dorsey, it is still possible to find among them " all the 
foundations of the totemic system, just as in the other societies 
of the Sioux. "^ Now among these peoples, above all the par- 
ticular deities to whom men render a cult, there is a pre-eminent 
power to which all the others have the relation of derived forms, 
and which is called wakan^ Owing to the preponderating place 
thus assigned to this principle in the Siouan pantheon, it is some- 
times regarded as a sort of sovereign god, or a Jupiter or Jahveh, 
and travellers have frequently translated wakan by " great 
spirit." This is misrepresenting its real nature gravely. The 
wakan is in no way a personal being ; the natives do not repre- 
sent it in a determined form. According to an observer cited by 
Dorsey, " they say that they have never seen the wakanda, so 
they cannot pretend to personify it."^ It is not even possible 

* Alice Fletcher, A Study of the Omaha Tribe, in Smithsonian Rep. for 1897, 
pp. 582 f. 

'* Dorsey, Siouan Sociology, in XVth Rep., p. 238. ' Ibid., p. 221. 

* Riggs and Dorsey, Dakota-English Dictionary, in Contrib. N. Amer. Ethnol., 
VII, p. 508. Many observers cited by Dorsey identify the word wakan with the 
words wakanda and wakanta, which are derived from it, but which really have 
a more precise signification. 

^ Xlth Rep., p. 372, § 21. Miss Fletcher, while recognizing no less clearly the 
impersonal character of the wakanda, adds nevertheless that a certain anthro- 
pomorphism has attached to this conception. But this anthropomorphism 
concerns the various manifestations of the -wakanda. Men address the trees or 
rocks where they think they perceive the wakanda, as if they were personal beings. 
But the wakanda itself is not personified (Smithsonian Rep. for 1897, p. 579). 

Origins of these Beliefs 193 

to define it by determined attributes and characteristics. " No 
word," says Riggs, " can explain the meaning of this term among 
the Dakota. It embraces all mystery, all secret power, all 
divinity."^ All the beings which the Dakota reveres, " the earth, 
the four winds, the sun, the moon and the stars, are manifesta- 
tions of this mysterious life and power " which enters into all. 
Sometimes it is represented in the form of a wind, as a breath 
having its seat in the four cardinal points and moving every- 
thing :^ sometimes it is a voice heard in the crashing of the 
thunder ,^ the sun, moon and stars are wakan.* But no enumera- 
tion could exhaust this infinitely complex idea. It is not a definite 
and definable power, the power of doing this or that ; it is Power 
in an absolute sense, with no epithet or determination of any 
sort. The various divine powers are only particular manifesta- 
tions and personifications of it ; each of them is this power seen 
under one of its numerous aspects.^ It is this which made one 
observer say, " He is a protean god ; he is supposed to appear 
to different persons in different forms."® Nor are the gods the 
only beings animated by it : it is the principle of all that lives or 
acts or moves. " All life is wakan. So also is everything which 
exhibits power, whether in action, as the winds and drifting 
clouds, or in passive endurance, as the boulder by the wayside."' 
Among the Iroquois, whose social organization has an even 
more pronouncedly totemic character, this, same idea is found 
again ; the word orenda which expresses it is the exact equiva- 
lent of the wakan of the Sioux. " The savage man," says Hewitt, 
" conceived the diverse bodies collectively constituting his en- 
vironment to possess inherently mystic potence . . . (whether 
they be) the rocks, the waters, the tides, the plants and the trees, 
the animals and man, the wind and the storms, the clouds and 
the thunders and the lightnings,"^ etc. " This potence is held 
to be the property of all things . . . and by the inchoate menta- 
tion of man is regarded as the efficient cause of all phenomena, all 
the activities of his environment."^ A sorcerer or shaman has 
orenda, but as much would be said of a man succeeding in his 
enterprises. At bottom, there is nothing in the world which does 
not have its quota of orenda ; but the quantities vary. There are 
some beings, either men or things, which are favoured ; there are 
others which are relatively disinherited, and the universal life 

* Riggs, Tah-Koo Wah-Kon, pp. 56-57, quoted from Dorsey, Xlth Rep., 
p. 433, § 95. 

2 Xlth Rep., p. 380, § 33. » Ibid., p. 38r, § 35. 

* Ibtd.. p. 376, § 28 ; p. 378, § 30 : cf. p. 449. § 138. " Ibid., p. 432, § 95. 
« Ib%d., p. 431. § 92. ' Ibid... p. 433, § 95. 

* Orenda and a Definition of Religion, in American Anthropologist. 1902, p. 33. 
" Ibid., p. 36. 

194 Elementary Forms of Religious Life 

consists in the struggles of these orenda of unequal intensity. 
The more intense conquer the weaker. Is one man more success- 
ful than his companions in the hunt or at war ? It is because he 
has more orenda. If an animal escapes from a hunter who is 
pursuing it, it is because the orenda of the former was the more 

This same idea is found among the Shoshone under the name 
of pokunt, among the Algonquin under the name of manitou,^ of 
nauala among the Kwakiutl,^ of yek among the Tlinkit^ and of 
sgâna among the Haida.^ But it is not peculiar to the Indians 
of North America ; it is in Melanesia that it was studied for the 
first time. It is true that in certain of the islands of Melanesia, 
social organization is no longer on a totemic basis ; but in all, 
totemism is still visible,^ in spite of what Codrington has said 
about it. Now among these peoples, we find, under the name of 
man a, an idea which is the exact equivalent of the wakan of the 
Sioux and the orenda of the Iroquois. The definition given by 
Codrington is as follows : Y' There is a belief in a force altogether 
distinct from physical power, which acts in all ways for good and 
evil ; and which it is of the greatest advantage to possess or con- 
trol. This is Mana. I think I know what our people mean by it. 
... It is a power-oi* influence, not physical and in a way super- 
natural ; but it shows itself in physical force, or in any kind of 
power or excellence which a man possesses. This mana is not 
fixed in anything, and can be conveyed in almost anything. 
. . . All Melanesian religion consists, in fact, in getting this 
mana for one's self, or getting it used for one's benefit."* Is this 
not the same notion of an anonymous and diffused force, the 
germs of which we recently found in the totemism of Australia ? 
Here is the same impersonality ; for, as Codrington says, we must 
be careful not to regard it as a sort of supreme being ; any such 
idea is " absolutely foreign " to Melanesian thought. Here is 
the same ubiquity ; the mana is located nowhere definitely and 
it is everywhere. All forms of life and all the effects of the action, 

* Tesa, Siudi del Thavenet, p. 17. * Boas, Kwakiutl, p. 695. 

' Swanton, Social Condition, etc., of the Tlinkit Indians, XXVIth Rep., 1905, 
p. 451, n. 2. 

* Swanton, Contributions to the Ethnology of the H aida, p. 14 ; cf. Social 
Condition, etc., p. 479. 

* In certain Melanesian societies (Banks Islands, North New Hebrides) the 
two exogaraic phratries are found which characterize the Australian organization 
(Codrington, The Melanesians, pp. 23 ff.). In Florida, there are regular totems, 
called butos (ibid., p. 31). An interesting discussion of this point will be found 
in Lang, Social Origins, pp. 176 if. On the same subject, and in the same sense, 
see W. H. R. Rivers, Totemism in Polynesia and Melanesia, in J. A. I., XXXIX, 
pp. 156 fï. 

'• The Melanesians, p. 118, n. i. Cf. Parkinson, Dreissig Jahre in der Siidsee, 
pp. 178, 392, 394. etc. 

Origins of these Beliefs 195 

either of men or of living beings or of simple minerals, are attri- 
buted to its influence.^ 

Therefore there is no undue temerity in attributing to the 
Australians an idea such as the one we have discovered in our 
analysis of totemic beliefs, for we find it again, but abstracted 
and generalized to a higher degree, at the basis of other religions 
whose roots go back into a system like the Australian one and 
which visibly bear the mark of this. The two conceptions 
are obviously related ; they differ only in degree, while the 
mana is diffused into the whole universe, what we call the god 
or, to speak more precisely, the totemic principle, is localized in 
the more limited circle of the beings and things of certain species. 
It is mana, but a little more specialized ; yet as a matter of fact, 
this specialization is quite relative. 

Moreover, there is one case where this connection is made 
especially apparent. Among the Omaha, there are totems of 
all sorts, both individual and collective ; ^ but both are only 
particular forms of wakan. " The foundation of the Indian's 
faith in the efftcacy of the totem," says Miss Fletcher, " rested 
upon his belief concerning nature and life. This conception was 
complex and involved two prominent ideas : First, that all 
things, animate and inanimate, were permeated by a common 
life ; and second, that this life could not be broken, but was 
continuous." ^ Now this common principle of life is the wakan. 
The totem is the means by which an individual is put into relations 
with this source of energy ; if the totem has any powers, it is 
because it incarnates the wakan. If a man who has violated the 
interdictions protecting his totem is struck by sickness or death, 
it is because this mysterious force against which he has thus 
set himself, that is, the wakan, reacts against him with a force 
proportionate to the shock received.'* Also, just as the totem is 
wakan, so the wakan, in its turn, sometimes shows its totemic 
origin by the way in which it is conceived. In fact. Say says that 
among the Dakota the " wahconda " is manifested sometimes 
in the form of a grey bear, sometimes of a bison, a beaver or 
some other animal.^ Undoubtedly, this formula cannot be 
accepted without reserve. The wakan repels all personitication 

* An analysis of this idea will be found in Hubert and Mauss, Théorie Générale 
de la Magxe, in Année Sociol., VII, p. io8. 

" There are not only totems of clans but also of guilds (A. Fletcher, Smith- 
sonian Rep. for 1897, pp. 581 ff.). 
' Fletcher, op. cit., pp. 57S f. 

* Ibid., p. 583. Among the Dakota, the totem is called Wakan. See Riggs 
and Dorsey, Dakota Grammar, Texts and Ethnol., in Contributions N. Amer. 
Ethn., 1893, p. 219. 

- James's Account of Long's Expedition in the Rocky Mountains, L, p. 268. 
(Quoted by Dorsey, Xlth Rep., p. 431, § 92.) 

196 Elementary Forms of Religious Life 

and consequently it is hardly prob. ole that it has ever been 
thought of in its abstract generality vvith the aid of such definite 
symbols. But Say's remark is probably applicable to the 
particular forms which it takes in specializing itself in the concrete 
reality of life. Now if there is a Dossibility that there was a time 
when these specializations of the wakan bore witness to such 
an affinity for an animal form, that would be one more proof 
of the close bonds uniting this conception to the totemic beliefs.^ 
It is possible to explain why this idea has been unable to reach 
the same degree of abstraction in Australia as in the more ad- 
vanced societies. This is not merely due to the insufficient 
aptitude of the Australian for abstracting and generalizing : 
before all, it is the nature of the social environment which has 
imposed this particularism. In fact, as long as totemism remains 
at the basis of the cultural organization, the clan keeps an 
autonomy in the religious society which, though not absolute, 
is always very marked. Of course we can say that in one sense 
each totemic group is only a chapel of the tribal Church ; 
but it is a chapel enjoying a large independence. The cult cele- 
brated there, though not a self-sufficing whole, has only external 
relations with the others ; they interchange without intermingling ; 
the totem of the clan is fully sacred only for this clan. Con- 
sequently the groups of things attributed to each clan, which are 
a part of it in the same way the men are, have the same indi- 
viduality and autonomy. Each of them is represented as ir- 
reducible into similar groups, as separated from them by a 
break of continuity, and as constituting a distinct realm. Under 
these circumstances, it would occur to no one that these hetero- 
geneous worlds were different manifestations of one and the same 
fundamental force ; on the contrary, one might suppose that 
each of them corresponded to an organically different mana 
whose action could not extend beyond the clan and the circle 
of things attributed to it. The idea of a single and universal 
mana could be born only at the moment when the tribal religion 
developed above that of the clans and absorbed them more or 
less completely. It is along with the feeling of the tribal unity 
that the feeling of the substantial unity of the world awakens. 
As we shall presently show,^ it is true that the Australian societies 
are already acquainted with a cult that is common to the tribe 
as a whole. But if this cult represents the highest form of the 

^ We do not mean to say that in principle every representation of religious 
forces in an animal form is an index of former totemism. But when we are 
dealing with societies where totemism is still apparent, as is the case with the 
Dakota, it is quite natural to think that these conceptions are not foreign to it. 

2 See below, same book, ch. ix, § 4, pp. 285 ff. 

Origins of these Beliefs 197 

Australian religions, it has not succeeded in touching and modify- 
ing the principles upon which they repose : totemism is essentially 
a federative religion which cannot go beyond a certain degree of 
centralization without ceasing to be itself. 

One characteristic fact clearly shows the fundamental reason 
which has kept the idea of the mana so specialized in Australia. 
The real religious forces, those thought of in the form of totems, 
are not the only ones with which the Australian feels himself 
obliged to reckon. There are also some over which magicians 
have particular control. While the former are theoretically 
considered healthful and beneficent, the second have it as their 
especial function to cause sickness and death. And at the same 
time that they differ so greatly in the nature of their effects, 
they are contrasted also by the relations which they sustain 
with the social organization. A totem is always a matter of 
the clan ; but on the contrary, magic is a tribal and even an 
intertribal institution. Magic forces do not belong to any 
special portion of the tribe in particular. All that is needed to 
make use of them is the possession of efficient recipes. Likewise, 
everybody is liable to feel their effects and consequently should 
try to protect himself against them. These are vague forces, 
specially attached to no determined social division, and even 
able to spread their action beyond the tribe. Now it is a remark- 
able fact that among the Arunta and Loritja, they are con- 
ceived as simple aspects and particular forms of a unique force, 
called in Arunta Arungquiltha or Arunkulta^ " This is a term," 
say Spencer and Gillen, " of somewhat vague import, but always 
associated at bottom with the possession of supernatural evil 
power. . . . The name is applied indiscriminately to the evil 
influence or to the object in which it is, for the time being, or 
permanently, resident." ^ " By arunkulta," says Strehlow, 
" the native signifies a force which suddenly stops life and 
brings death to all who come in contact with it." ^ This name 
is given to the bones and pieces of wood from which evil-working 
charms are derived, and also to poisonous animals and vegetables. 
So it may accurately be called a harmful mana. Grey mentions 
an absolutely identical notion among the tribes he observed.'* 
Thus among these different peoples, while the properly religious 

* The first spelling is that of Spencer and Gillen ; the second, that of Strehlow. 

* Nat. Tr., p. 548, n. i. It is true that Spencer and Gillen add : " The idea 
can be best expressed by saying that an Arungquiltha object is possessed of an 
evil spirit." But this free translation of Spencer and Gillen is their own unjusti- 
fied interpretation. The idea of the arungquiltha in no way implies the existence 
of spiritual beings, as is shown by the context and Strehlow's definition. 

* Die Aranda, II, p. 76, n. 

* Under the name Boyl-ya (see Grey, Journal of Two Expeditions, II, pp. 

198 Elementary Forms of Religious Life 

forces do not succeed in avoiding a certain heterogeneity, magic 
forces are thought of as being all of the same nature ; the mind 
represents them in their generic unity. This is because they 
rise above the social organization and its divisions and sub- 
divisions, and move in a homogeneous and continuous space 
where they meet with nothing to differentiate them. The others, 
on the contrary, being localized in definite and distinct social 
forms, are diversified and particularized in the image of the 
environment in which they are situated. 

From this we can see how thoroughly the idea of an impersonal 
religious force enters into the meaning and spirit of Australian 
totemism, for it disengages itself with clarity as soon as no 
contrary cause opposes it. It is true that the arungquiltha is 
purely a magic force. But between religious forces and magic 
forces there is no difference of kind : ^ sometimes they are even 
designated by the same name : in Melanesia, the magicians 
and charms have mana just like the agents and rites of the regular 
cult ; 2 the word oranda is employed in the same way by the 
Iroquois.^ So we can legitimately infer the nature of the one 
from that of the other. ^ 


The results to which the above analysis has led us do not 
concern the history of totemism only, but also the genesis of 
religious thought in general. 

Under the pretext that in early times men were dominated 
by their senses and the representations of their senses, it has 
frequently been held that they commenced by representing the 
divine in the concrete form of definite and personal beings. 
The facts do not confirm this presumption. We have just 
described a systematically united scheme of religious beliefs 
which we have good reason to regard as very primitive, yet we 

^ See above, p. ^,:. Spencer and Gillen recognize this implicitly when they 
say that the arungquiltha is a " supernatural force. ' Cf. Hubert and Mauss, 
Théorie Générale de la Magie, in Année Sociol., VII, p. 119. 

* Codrington, The Melanesians, pp. 191 ff. 
^ Hewitt, loc. cit., p. 38. 

* There is even ground for asking whether an analogous notion is completely 
lacking in Australia. The word churinga, or tjurunga as Strehlow writes, has a 
very great similarity, with the Arunta. Spencer and Gillen say that it designates 
" ail that is secret or sacred. It is applied both to the object and to the quality 
it possesses " (Nat. Tr., p. 648, s.v. churinga). This is almost a definition of 
mana. Sometimes Spencer and Gillen even use this word to designate religious 
power or force in a general way. While de.scribing a ceremony among the Kaitish, 
they say that the oificiant is " full of churinga," that is to say, they continue, of 
the " magic power emanating from the objects called churinga." Yet it does not 
seem that the notion of churinga has the same clarity and precision as that of the 
mana in Melanesia or of the wakan among the Sioux. 

Origins of these Beliefs 199 

have met with no personalities of this sort. {The real totemic 
cult is addressed neither to certain deterrnined animals nor to 
certain vegetables nor even to an animal or vegetable species, 
but to a_yague power spread through these things.^! Even in the 
most advanced religions which have developed out of totemism, 
such as those which we find among the North American Indians, 
this idea, instead of being effaced, becomes more conscious of 
itself ; it is declared with a clarity it did not have before, while 
at the same time, it attains a higher generality. It is this which 
dominates the entire religious system. 

This is the original matter out of which have been constructed 
those beings of every sort which the religions of all times have 
consecrated and adored. The spirits, demons, genii and gods 
of every sort are only the concrete forms taken by this energy, 
or " potentiality," as Hewitt calls it,^ in individualizing itself, 
in fixing itself upon a certain determined object or point in space, 
or in centring around an ideal and legendary being, though one 
conceived as real by the popular imagination. A Dakota ques- 
tioned by Miss Fletcher expressed this essential consubstanti- 
ability of all sacred things in language that is full of relief. " Every 
thing as it moves, now and then, here and there, makes stops. 
The bird as it flies stops in one place to make its nest, and in 
another to rest in its flight. A man when he goes forth stops 
when he wills. So the god has stopped. The sun, which is so 
bright and beautiful, is one place where he has stopped. The 
trees, the animals, are where he has stopped, and the Indian 
thinks of these places and sends his prayers to reach the place 
where the god has stopped and to win help and a blessing." ^ 
In other words, the wakan (for this is what he was talking about) 
comes and goes through the world, and sacred things are the 
points upon which it alights. Here we are, for once, just as far 
from naturism as from animism. If the sun, the moon and the 
stars have been adored, they have not owed this honour to their 
intrinsic nature or their distinctive properties, but to the fact 
that they are thought to participate in this force which alone is 
able to give things a sacred character, and which is also found 
in a multitude of other beings, even the smallest. If the souls of 
the dead have been the object of rites, it is not because they are 
believed to be made out of* some fluid and impalpable substance, 
nor is it because they resemble the shadow cast by a body or its 

^ Yet we shall see below (this book, ch. viii and ix) that totemism is not 
foreign to all ideas of a mythical personality. But we shall show that these con- 
ceptions are the product of secondary formations : far from being the basis of 
the beliefs we have just analysed, they are derived from them. 

^ Loc cit., p. 38. 

* Rep. Peabody Museum, III, p. 276, n. (quoted by Dorsey, Xlth Rep., p. 435). 

200 Elementary Forms of Religious Life 

reflection on a surface of water. Lightness and fluidity are not 
enough to confer sanctity ; they have been invested with this 
dignity only in so far as they contained within them something 
of this same force, the source of all religiosity. 

We are now in a better condition to understand why it has 
been impossible to define religion by the idea of mythical per- 
sonalities, gods or spirits ; it is because this way of representing 
religious things is in no way inherent in their nature. What 
we find at the origin and basis of religious thought are not 
determined and distinct objects and beings possessing a sacred 
character of themselves ; they are indefinite powers, anonymous 
forces, more or less numerous in different societies, and some- 
times even reduced to a unity, and whose impersonality is strictly 
comparable to that of the physical forces whose manifestations 
the sciences of nature study. As for particular sacred things, 
they are only individualized forms of this essential principle. 
So it is not surprising that even in the religions where there are 
avowed divinities, there are rites having an efficient virtue in 
themselves, independently of all divine intervention. It is 
because this force may be attached to words that are pronounced 
or movements that are made just as well as to corporal substances ; 
the voice or the movements may serve as its vehicle, and it may 
produce its effects through their intermediacy, without the aid of 
any god or spirit. Even should it happen to concentrate itself 
especially in a rite, this will become a creator of divinities from 
that very fact.^ This is why there is scarcely a divine personality 
who does not retain some impersonality. Those who represent 
it most clearly in a concrete and visible form, think of it, at the 
same time, as an abstract power which cannot be defined except 
by its own efficacy, or as a force spread out in space and which is 
contained, at least in part, in each of its effects. It is the power 
of producing rain or wind, crops or the light of day ; Zeus is in 
each of the raindrops which falls, just as Ceres is in each of the 
sheaves of the harvest. ^ As a general rule, in fact, this efficacy 
is so imperfectly determined that the believer is able to form 
only a very vague notion of it. Moreover, it is this indecision 
which has made possible these syncretisms and duplications in 
the course of which gods are broken up, dismembered and con- 
fused in every way. Perhaps there is not a single religion in 
which the original mana, whether unique or multiform, has been 

* See above, p. 35. 

* In the expressions such as Z(v% vti or Ceres succiditur, it is shown that 
this conception survived in Greece as well as in Rome. In his GoUemamen, 
Usener has clearl}' shown that the priniitive gods of Greece and Rome were 
impersonal forces thought of only in terms of their attributes. 

Origins of these Beliefs 201 

resolved entirely into a clearly defined number of beings who 
are distinct and separate from each other ; each of them always 
retains a touch of impersonality, as it were, which enables it to 
enter into new combinations, not as the result of a simple survival 
but because it is the nature of religious forces to be unable to 
individualize themselves completely. 

This conception, to which we have been led by the study of 
totemism alone, has the additional recommendation that many 
scholars have recently adopted it quite independently of one 
another, as a conclusion from very different sorts of studies. 
There is a tendency towards a spontaneous agreement on this 
point which should be remarked, for it is a presumption of 

As early as 1899, we pointed out the impossibility of making 
the idea of a mythical personality enter into the definition of 
religious phenomena.^ In 1900, Marrett showed the existence 
of a rehgious phase which he called preanimistic, in which the 
rites are addressed to impersonal forces like the Melanesian 
mana and the wakan of the Omaha and Dakota. ^ However, 
Marrett did not go so far as to maintain that always and in 
every case the idea of a spirit is logically and chronologically 
posterior to that of mana and is derived from it ; he even seemed 
disposed to admit that it has sometimes appeared independently 
and consequently, that religious thought flows from a double 
source.' On the other hand, he conceived the mana as an 
inherent property of things, as an element of their appearance ; 
for, according to him, this is simply the character which we 
attribute to everything out of the ordinary, and which inspires 
a sentiment of fear or admiration.* This practically amounts to 
a return to the naturist theory. ^ 

A little later, MM. Hubert and Mauss, while attempting to 
formulate a general theory of magic, established the fact that 
magic as a whole reposes on the notion of mana.^ The close 
kinship of the magic rite and the religious rite being known, it 
was even possible to foresee that the same theory should be 
applied to religion. This was sustained by Preuss in a series of 

» Definition du phénomène religieux, in Année Social. , II, pp. 14-16. 

» Preanimistic Religion, in Folk-Lore, 1900, pp. 162-182. 

^ Ibid., p. 179. In a more recent work. The Conception 0/ Mana (in Trans- 
actions 0/ the Third International Congress for the History of Religions, II, pp. 54 
ff.). Marrett tends to subordinate still further the animistic conception of mana. 
but his thought on this point remains hesitating and very reserved. 

* Ibid., p. 16S. 

* This return of preanimisra to naturism is still more marked in Clodd, 
Preanimistic Stages of Religion {Trans. Third Inter. Congress for the H. of Rel.. 
I. P- iZ)- 

* Théorie générale de la Magie, in Année Social.. VII, pp. 108 S. 

202 Elementary Forms of Religious Life 

articles in the Globus ^ that same year. Relying chiefly upon 
facts taken from American civilizations, Preuss set out to prove 
that the ideas of the soul and spirit were not developed until 
after those of power and impersonal force, that the former are 
only a transformation of the latter, and that up to a relatively 
late date they retain the marks of their original impersonality. 
In fact, he shows that even in the advanced religions, they are 
represented in the form of vague emanations disengaging them- 
selves automatically from the things in which they reside, and 
even tending to escape by all the ways that are open to them : 
the mouth, the nose and all the other openings of the body, the 
breath, the look, the word, etc. At the same time, Preuss pointed 
out their Protean forms and their extreme plasticity which 
permits them to give themselves successively and almost con- 
currently to the most varied uses.^ It is true that if we stick to 
the letter of the terminology employed by this author, we may 
believe that for him the forces have a magic, not a religious 
nature : he calls them charms {Zauber, Zaiiberkrafte). But it is 
evident that in expressing himself thus, he does not intend to 
put them outside of religion ; for it is in the essentially religious 
rites that he shows their action, for example, in the great Mexican 
ceremonies.^ If he uses these expressions, it is undoubtedly 
because he knows no others which mark better the impersonality 
of these forces and the sort of mechanism with which they operate. 
Thus this same idea tends to come to light on every side.* 
1 The impression becomes more and more prevalent that even 
/lithe most elementary mythological constructions are secondary 
I products 5 which cover over a system of beliefs, at once simpler 
and more obscure, vaguer and more essential, which form the 
solid foundations upon which the religious systems are built. 
It is this primitive foundation which our analysis of totemism 
has enabled us to reach. The various writers whose studies we 
have just mentioned arrived at this conclusion only through 

^ Der Ursprung der Religion und Kunst, in Globus, 1904, Vol. LXXXVI, pp. 
321. 355. 376, 389 ; 1905. Vol. LXXXVII, pp. 333, 347. 380, 394. 413. 

» Globus, LXXXVII, p. 381. 

' He clearly opposes them to all influences of a profane nature {Globus, 
LXXXVI, p. s'yga). 

* It is found even in the recent theories of Frazer. For if this scholar denies 
to totemism all religious character, in order to make it a sort of magic, it is just 
because the forces which the totemic cult puts into play are impersonal like those 
employed by the magician. So Frazer recognizes the fundamental fact which we 
have just established. But he draws different conclusions because he recognizes 
religion only where there are mythical personalities. 

* However, we do not take this word in the same sense as Preuss and Marrett. 
According to them, there was a time in religious evolution when men knew 
neither souls nor spirits : a preanimistic phase. But this hypothesis is very 
questionable : we shall discuss this point below (Bk. II, ch. viii and ix). 

Origins of these Beliefs 203 

facts taken from very diverse religions, some of which even 
correspond to a civilization that is already far advanced : such 
is the case, for example, with the Mexican religions, of which 
Preuss makes great use. So it might be asked if this theory is 
equally applicable to the most simple religions. But since it is 
impossible to go lower than totemism., we are not exposed to this 
risk of error, and at the same time, we have an opportunity of 
finding the initial notion from which the ideas of wakan and 
mana are derived : this is the notion of the totemic principle.^ 

^ IV 

But this notion is not only of primary importance because of 
the rôle it has played in the development of religious ideas ; it 
also has a lay aspect in which it is of interest for the history of 
scientific thought. It is the first form of the idea of force. 

In fact, the wakan plays the same rôle in the world, as the 
Sioux conceives it, as the one played by the forces with which 
science explains the diverse phenomena of nature. This, however, 
does not mean that it is thought of as an exclusively physical 
energy ; on the contrary, in the next chapter we shall see that 
the elements going to make up this idea are taken from the most 
diverse realms. But this very compositeness of its nature enables 
it to be utilized as a universal principle of explanation. It is 
from it that all life comes ; ^ " all life is wakan " ; and by this 
word life, we must understand everything that acts and reacts, 
that moves and is moved, in both the mineral and biological 
kingdoms. The wakan is the cause of all the movements which 
take place in the universe. We have even seen that the orenda 
of the Iroquois is " the efficient cause of all the phenomena 
and all the activities which are manifested around men." It is a 
power " inherent in all bodies and all things." ^ It is the orenda 
which makes the wind blow, the sun lighten and heat the earth, 
or animals reproduce and which makes men strong, clever and 
intelligent. When the Iroquois says that the life of all nature 
is the product of the conflicts aroused between the unequally 
intense orenda of the different beings, he only expresses, in his 
own language, this modern idea that the world is a system 

^ On this same question, see an article of Alessandro Bruno, Sui fenomeni 
magico-religiosi della communità primitive, in Rivista italiana di Sociologia, XII 
Year, Fasc. IV-V, pp. 568 ff., and an unpublished communication made by 
W. Bogoras to the XIV Congress of the Americanists, held at Stuttgart in 1904. 
This communication is analysed by Preuss in the Globus, LXXXVI, p. 201. 

* " All things," says Miss Fletcher, " are filled with a common principle of 
life," Smiths. Rep. for 1897, p. 579. 

' Hewitt, in American Anthropologist, 1902, p. 36. 

204 Elementary Forms of Religious Life 

of forces limiting and containing each other and making an 

The Melanesian attributes this same general efficacy to his 
mana. It is owing to his mana that a man succeeds in hunting 
or fighting, that gardens give a good return or that flocks prosper. 
If an arrow strikes its mark, it is because it is charged with mana ; 
it is the same cause which makes a net catch fish well, or a canoe 
ride well on the sea,i etc. It is true that if certain phrases of 
Codrington are taken literally, mana should be the cause to 
which is attributed " everything which is beyond the ordinary 
power of men, outside the common processes of nature." ^ But 
from the very examples which he cites, it is quite evident that 
the sphere of the mana is really much more extended. In reality, 
it serves to explain usual and everyday phenomena ; there is 
nothing superhuman or supernatural in the fact that a ship 
sails or a hunter catches game, etc. However, among these 
events of daily Hfe, there are some so insignificant and familiar 
that they pass unperceived : they are not noticed and conse- 
quently no need is felt of explaining them. The concept of mana 
is applied only to those that are important enough to cause 
reflection, and to awaken a minimum of interest and curiosity ; 
but they are not marvellous for all that. And what is true of 
the mana as well as the orenda and wakan, may be said equally 
well of the totemic principle. It is through this that the life of the 
men of the clan and the animals or plants of the totemic species, 
as well as all the things which are classified under the totem and 
partake of its nature, is manifested. 

So the idea of force is of religious origin. It is from religion 
that it has been borrowed, first by philosophy, then by the 
sciences. This has already been foreseen by Comte and this 
is why he made metaphysics the heir of " theology." But he 
concluded from this that the idea of force is destined to disappear 
from science ; for, owing to its mystic origins, he refused it all 
objective value. But we are going to show that, on the contrary, 
religious forces are real, howsoever imperfect the symbols may 
be, by the aid of which they are thought of. From this it will 
follow that the same is true of the concept of force in general. 

^ The Melanssians, pp. 1 18-120. * Ibid., p. 119. 



Origin of the Idea of the Totemic Principle or Mana 

THE proposition established in the preceding chapter 
determines the terms in which the problem of the origins 
of totemism should be posed. Since totemism is everywhere 
dominated by the idea of a quasi-divine principle, imminent 
in certain categories of men and things and thought of under 
the form of an animal or vegetable, the explanation of this 
religion is essentially the explanation of this belief ; to arrive at 
this, we must seek to learn how men have been led to construct 
this idea and out of what materials they have constructed it. 


It is obviously not out of the sensations which the things 
serving as totems are able to arouse in the mind ; we have shown 
that these things are frequently insignificant. The lizard, the 
caterpillar, the rat, the ant, the frog, the turkey, the bream-fish, 
the plum-tree, the cockatoo, etc., to cite only those names which 
appear frequently in the lists of Australian totems, are not of a 
nature to produce upon men these great and strong impressions 
which in a way resemble religious emotions and which impress 
a sacred character upon the objects they create. It is true that 
this is not the case with the stars and the great atmospheric 
phenomena, which have, on the contrary, all that is necessary 
to strike the imagination forcibly ; but as a matter of fact, 
these serve only very exceptionally as totems. It is even probable 
that they were very slow in taking this office.^ So it is not the 
intrinsic nature of the thing whose name the clan bears that 
marked it out to become the object of a cult. Also, if the senti- 
ments which it inspired were really the determining cause of the 
totemic rites and beliefs, it would be the pre-eminently sacred 
thing ; the animals or plants employed as totems would play 
an eminent part in the rehgious life. But we know that the 

> See above, p. 103. 

2o6 Elementary Forms of Religious Life 

centre of the cult is actually elsewhere. It is the figurative 
representations of this plant or animal and the totemic emblems 
and symbols of every sort, which have the greatest sanctity ; 
so it is in them that is found the source of that religious nature, 
of which the real objects represented by these emblems receive 
only a reflection. 

Thus the totem is befflrje-all-a.symbol, a material expression of 
something else.^ But of what ? 

From the analysis to which we have been giving our attention, 
it is evident that it expresses and symbolizes two different sorts 
of things. In the first place, it is the outward and visible form of 
what we have called the totemic principle or god. But it is also 
the symbol of the determined society called the clan. It is its 
flag ; it is the sign by which each clan distinguishes itself from 
the others, the visible mark of its personality, a mark borne by 
everything which is a part of the clan under any title whatsoever, 
,\y men, beasts or things, -^o if it is at once the s ymbol of the god 
*_^ and of the society, is thaîTiôt because the gba and^lhe society 
are only one ^ How could the emblem of the group have been 
able to become the figure of this quasi-divinity, if the group and 
; the divinity were two distinct realities ? The god of the clan, 
i the totemic principle, can therefore be nothing else than the K\j^\ 
I clan itself, personified and represented to the imagination under ^^ 
I the visible form of the animal or vegetable which serves as totem.'' 
But how has this apotheosis been possible, and how did it 
happen to take place in this fashion ? 


In a general way, it is unquestionable that a society has all 
that is necessary to arouse the sensation of the divine in 
minds, merely by the power that it has over them ; for to its 
1^ members it is what a god is to his worshippers. In fact, a god 
is, first of all, a being whom men think of as superior to them- 
selves, and upon whom they feel that they depend. Whether 
it be a conscious personality, such as Zeus or Jahveh, or merely 
abstract forces such as those in play in totemism, the worshipper, 
in the one case as in the other, believes himself held to certain 
manners of acting which are imposed upon him by the nature 
of the sacred principle with which he feels that he is in com- 
jnunion. Now society also gives us the sensation of a perpetual 
dependence. Since it has a nature which is peculiar to itself 
and different from our individual nature, it pursues ends which 

> Pickler, in the little work above mentioned, had already expres.oed, in a 
slightly dialectical manner, the sentiment that this is what the totem essentially is. 

Origins of these Beliefs 207 

are likewise special to it ; but, as it cannot attain them except 
through our intermediacy, it imperiously demands our aid. It 
requires that, forgetful of our own interests, we make ourselves 

its servitors, and it submits us to every sort of inconvenience,. 

privation and sacrifice, without which social life would be im- 
possible. It is because of this that at every instant we are obliged 
to submit ourselves to rules of conduct and of thought which we 
have neither made nor desired, and which are sometimes even 1 
contrary to our most fundamental inchnations and instincts. j 

Even if society were unable to obtain these concessions and 
sacrifices from us except by a material constraint, it might 
awaken in us only the idea of a physical force to which we must 
give way of necessity, instead of that of a moral power such as 
religions adore. But as a matter of fact, the empire which it 
holds over consciences is due much less to the physical supremacy 
of which it has the privilege than to the moral authority with 
which it is invested. /If we yield to its orders, it is not merely 
because it is strong enough to triumph over our resistance ; it is 
primarily because it is the object of a venerable respê ct!\ — -^""'^^ ^ 

We say that an object, whether individual or collective, inspires 
aspect when the representation expressing it in the mind is 
gifted with such a force that it automatically causes or inhibits 
actions, without regard for any consideration relative to their useful 
or injurious effects. When we obey somebody because of the 
moral authority which we recognize in him, we follow out his 
opinions, not because they seem wise, but because a certain 
sort of physical energy is imminent in the idea that we form of 
this person, which conquers our will and inclines it in the indicated 
direction. Respect is the emotion which we experience when we 
feel this interior and wholly spiritual pressure operating upon 
us. Then we are not determined by the advantages or incon- 
veniences of the attitude which is prescribed or recommended to 
us ; it is by the way in which we represent to ourselves the person 
recommending or prescribing it. This is why commands generally 
take a short, peremptory form leaving no place for hesitation ; it 
is because, in so far as it is a command and goes by its own force, 
it excludes all idea of deliberation or calculation ; it gets its 
efficacy from the intensity of the mental state in which it is placed. 
It is this intensity which creates what is called a moral ascendancy. 

Now the ways of action to which society is strongly enough 
attached to impose them upon its members, are, by that very 
fact, marked with a distinctive sign provocative of respect. 
Since they are elaborated in common, the vigour with which 
they have been thought of by each particular mind is retained 
in aU the other minds, and reciprocally. The representations 


2o8 Elementary Forms of Religious Life 

which express them within each of us have an intensity which 
no purely private states of consciousness could ever attain ; 
for they have the strength of the innumerable individual repre- 
sentations which have served to form each of them. It is society 
who speaks through the mouths of those who affirm them in our 
presence ; it is society whom we hear in hearing them ; and the 
voice of all has an accent which that of one alone could never 
have.^ The very violence with which society reacts, by way of 
blame or material suppression, against every attempted dissidence, 
contributes to strengthening its empire by manifesting the 
common conviction through this burst of ardour. ^ In a word, 
when something is the object of such a state o f opinion , the 
representation which each individual has of it"gams a power 
of action from its origins and the conditions m which it was born, 
which even those feel who do not. submit themselves to it. If" 
tends to repel the representations which contradicf"it, and it 
keeps them at a distance ; on the other hand, it commands those 
acts which will realize it, and it does so, not by a material coercion 
or by the perspective of something of this sort, but by the simple 
radiation of the mental energy which it contains. It has an 
efficacy coming solely from its psychical properties, and it is by 
just this sign that moral authority is recognized. So opinioiir* 
primarily a social thing, is a source of authority, aiTdnfrrugfit 
even be asked whether aU authority is not the daughter of 
opinion.^ It may be objected that science is often the antagonist 
of opinion, whose errors it combats and rectifies. But it cannot 
succeed in this task if it does not have sufficient authority, and 
it can obtain this authority only from opinion itself. If a people 
did not have faith in science, all the scientific demonstrations in 
the world would be without any influence whatsoever over their 
minds. Even to-day, if science happened to resist a very strong 
current of public opinion, it would risk losing its credit there.* 

* See our Division du travail social, 3rd éd., pp. 64 ff. 

* Ibid., p. 76. 

3 This is the case at least with all moral authority recognized as such by the 
group as a whole. 

* We hope that tliis analysis and those which follow will put an end to an 
inexact interpretation of our thought, from which more than one misunderstand- 
ing has resulted. Since we have made constraint the outward sign by which 
social facts can be the most easily recognized and distinguished from the facts 
of individual psychology, it has been assumed that according to our opinion, 
physical constraint is the essential thing for social life. As a matter of fact, we 
have never considered it more than the material and apparent expression of an 
interior and profound fact which is wholly ideal : this is moral authority. The 
problem of sociology — if we can speak of a sociological problem — consists in 
seeking, among the different forms of external constraint, the different sorts of 
moral authority corresponding to them and in discovering the causes which 
have determined these latter. The particular question which we are treating 
in this present work has as its principal object, the discovery of the form under 

Origins of these Beliefs 209 

Since it is in spiritual ways that social pressure exercises 
itself, it could not fail to give men the idea that outside them- 
selves there exist one or several powers, both moral and, at the 
same time, efficacious, upon which they depend. They must 
think of these powers, at least_ in part, as outside themselves, 
for these address them in a^one of command and sometimes 
even order them to do violence to their most natural inclinations. 
It is undoubtedly true that if they were able to see that these 
influences which they feel emanate from society, then the mytho- 
logical system of interpretations would never be bom. But 
social action follows ways that are too circuitous and obscure, 
and employs psychical mechanisms that are too complex to allow, 
the ordinary observer to see whence it comes. As long as scientific\ 
analysis does not come to teach it to them, men-Jcnow. well that/ 
they are acted upon, but they do not know by whom. So they\ 
must invent by themselves the idea of these powers with which 
they feel themselves in connection, and from that, we are able 
to catch a glimpse of the way by which they were led to represent 
them under forms that are really foreign to their nature and to 
transfigure them by thought. 


^ But a god is not merely an authority upon whom we depend ; 
itjs^alorce upon which our strength relies. The man who has 
obeyed his god and who, for this reason, believes the god is with 
him, approaches the world with confidence and with the feeling 
of an increased energy. Likewise, social action does not confine 
itself to demanding sacrifices, privations and efforts from us. For 
the collective force is not entirely outside of us ; it does not act 
upon us wholly from without ; but rather, since society cannot 
exist except in and through individual consciousnesses,^ this 
force must also penetrate us and organize itself wiThin us ; it 
thus becomes an integral part of our being and by that very 
fact this is elevated and magnified. 

There are occasions when this strengthening and vivifyifïg ~j^ 
action of society is especially apparent. In the midst of an 1 
assembly animated by a common passion, we become susceptible 1 
of acts and sentiments of which we are incapable when reduced |\ 

which that particular variety of moral authority which is inherent in all that is 
religious has been born, and out of what elements it is made. It will be seen 
presently that even if we do make social pressure one of the distinctive character- 
istics of sociological phenomena, we do not mean to say that it is the only one. 
We shall show another aspect of the collective life, nearly opposite to the pre- 
ceding one, but none the less real (see p. 212). 

^ Of course this does not mean to say that the collective consciousness does 
not have distinctive characteristics of its own (on this point, see Représentations 
individuelles et représentations collectives, in Revue de Métaphysique et de Morale, 
1898, pp. 273 ff.). 

210 Elementary Forms of Religious Life 

to our own forces ; and when the assembly is dissolved and when, 
/ finding ourselves alone again, we fall back to our ordinary level, 
we are then able to measure the height to which we have been 
raised above ourselves.] History abounds in examples of this 
"""■^sorfi ÎTîl^énough to think of the night of the Fourth of August, 
1789, when an assembly was suddenly led to an act of sacrifice 
and abnegation which each of its members had refused the day 
before, and at which they were all surprised the day after, ^z This 
is why all parties, political, economic or confessional, are careful 
to have periodical reunions where their members may revivify 
their common faith by manifesting it in common. To strengthen 
those sentiments which, if left to themselves, would soon weaken, 
it is sufficient to bring those who hold them together and to put 
them into closer and more active relations with one another. 
This is the explanation of the particular attitude of a man speaking 
to a crowd, at least if he has succeeded in entering into com- 
munion with it. His language has a grandiloquence that would 
be ridiculous in ordinary circumstances ; his gestures show a 
certain domination ; his very thought is impatient of all rules, 
and easily falls into all sorts of excesses. It is because he feels 
within him an abnormal over-supply of force which overflows 
and tries to burst out from him ; sometimes he even has the 
feeling that he is dominated by a moral force which is greater 
than he and of which he is only the interpreter. It is by this trait 
that we are able to recognize what has often been called the 
demon of oratorical inspiration. Now this exceptional increase 
of force is something very real ; it comes to him' from the very 
group which he addresses. The sentiments provoked by his words 
come back to him, but enlarged and amplified, and to this degree 
they strengthen his own sentiment. The passionate energies 
he arouses re-echo within him and quicken his vital tone. It is 
no longer a simple individual who speaks ; it is a group incarnate 
and personified. 

Beside these passing and intermittent states, there are other 
more durable ones, where this strengthening influence of society 
makes itself felt with greater consequences and frequently even 
with greater brilliancy. There are periods in history when, 
under the influence of some great collective shock, social inter- 
actions have become much more frequent and active. Men look 
for each other and assemble together more than ever. That 
general effervescence results which is characteristic of revo- 

^ This is proved by the length and passionate character of the debates where 
a legal form was given to the resolutions made in a moment of collective enthusi- 
asm. In the clergy as in the nobility, more than one person called this celebrated 
night the dupe's night, or, with Rivarol, the St. Bartholomew of the estates (see 
StoU, Suggestion und Hypnotismus in der V olkerpsychologie , 2nd éd.. p. 618, n. 2). 

Origins of these Beliefs 21 1 

lutionary or creative epochs. Now this greater activity results 
in a general stimulation of individual forces. Men see more and 
differently now than in normal times. Changes are not merely 
of shades and degrees ; men become different. The passions 
moving them are of such an intensity that they cannot be satisfied 
except by violent and unrestrained actions, afctions of super- 
human heroism or of bloody barbarism. This is what explains 
the Crusades/ for example, or many of the scenes, either sublime 
or savage, of the French Revolution. ^ Under the influence of 
the general exaltation, we see the most mediocre and inoffensive 
bourgeois become either a hero or a butcher.^ And so clearly 
are all these mental processes the ones that are also at the root 
of religion that the individuals themselves have often pictured the 
pressure before which they thus gave way in a distinctly religious 
form. The Crusaders believed that they felt God present in the 
midst of them, enjoining them to go to the conquest of the 
Holy Land ; Joan of Arc believed that she obeyed celestial 

But it is not only in exceptional circumstances that this 
stimulating action of society makes itself felt ; there is not, so 
to speak, a moment in our lives when some current of energy 
does not come to us from without. The man who has done his 
duty finds, in the manifestations of every sort expressing the 
sympathy, esteem or affection which his fellows have for him, 
a feeling of comfort, of which he does not ordinarily take account, 
but which sustains him, none the less. The sentiments which 
society has for him raise the sentiments which he has for himself. 
Because he is in moral harmony v^ith his comrades, he has-more 
confidence, courage and boldness in action, just like the believer 
who thinks that he feels the regard of his god turned graciously 
towards him. It thus produces, as it were, a perpetual sustenance 
for ourjnûrai-ftatttFe: — Si«ee this varies with a multitude of 
elrternal circumstances, as our relations with the groups about 
us are more or less active and as these groups themselves vary, 
we cannot fail to feel that this moral support depends upon an 
external cause ; but we do not perceive where this cause is nor 
what it is. So we ordinarily think of it under the form of a moral 
power which, though immanent in us, represents, within us 
something not ourselves : this is the moral-eonseienee, of which, 
by the way, men have never made even a slightly distinct repre- 
sentation except by the aid of religious symbols^ 

* See StoU, op. cit., pp. 353 ff. 

* Ibid., pp. 6ig, 635. ^ Ibid., pp. 622 ff. 

* The emotions of fear and sorrow are able to develop similarly and to become 
intensified under these same conditions. As we shall see, they correspond to 
quite another aspect of the religious life (Bk. Ill, ch. v). 



212 Elementary Forms of Religious Life 

In addition to these free forces which are constantly coming 
to renew our own, there are others which are fixed in the methods 
and traditions which we employ. We speak a language that we 
did not make ; we use instruments that we did not invent ; we 
invoke rights that we did not found ; a treasury of knowledge 
is transmitted to each generation that it did not gather itself, 
etc. It is to society that we owe these varied benefits of civiliza- 
tion, and if we do not ordinarily see the source from which we 
get them, we at least know that they are not our own work. 
Now it is these things that give man his own place among things ; 
a man is a man only because he is civilized. So he could not 
escape the feeling that outside of him there are active causes 
from which he gets the characteristic attributes of his nature 
and which, as benevolent powers, assist him, protect him and 
assure him of a privileged fate. And of course he must attribute 
to these powers a dignity corresponding to the great value of the 
good things he attributes to them.i 

Thus the environment in which we live seems to us to be 
peopled with forces that are at once imperious and helpful, 
august and gracious, and with which we have relations. Since 
they exercise over us a pressure of which we are conscious, we 
are forced to localize them outside ourselves, just as we do for 
the objective causes of our sensations. But the sentiments 
which they inspire in us differ in nature from those which we 
have for simple visible objects. As long as these latter are re- 
duced to their empirical characteristics as shown in ordinary 
experience, and as long as the religious imagination has not 
metamorphosed them, we entertain for them no feeling which 
resembles respect, and they contain within them nothing that 
is able to raise us outside ourselves. Therefore, the representations 
which express them appear to us to be very different from those 
aroused in us by collective influences. The two form two distinct 
and separate mental states in our consciousness, just as do the 
two forms of life to which they correspond. Consequently, we 
get the impression that we are in relations with two distinct 
sorts of reality and that a sharply drawn line of demarcation 
separates them from each other : on the one hand is the world 
of profane things, on the other, that of sacred things. 

Also, in the present day just as much as in the past, we see society 
constantly creating sacred things out of ordinary ones. If it 

* This is the other aspect of society which, while being imperative, appears at 
the same time to be good and gracious. It dominates us and assists us. If we 
have defined the social fact by the first of these characteristics rather than the 
second, it is because it is more readily observable, for it is translated into outward 
and visible signs ; but we have never thought of denying the second (see our 
Règles de la Méthode Sociologique, preface to the second edition, p. xx, n.i). 

Origins of these Beliefs 213 

happens to fall in love with a man and if it thinks it has found 
in him the principal aspirations that move it, as well as the means 
of satisfying them, this man will be raised above the others and, 
as it were, deified. Opinion will invest him with a majesty 
exactly analogous to that protecting the gods. This is what 
has happened to so many sovereigns in whom their age had 
faith : if they were not made gods, they were at least regarded 
as direct representatives of the deity. And the fact that it is 
society alone which is the author of these varieties of apotheosis, 
is evident since it frequently chances to consecrate men thus 
who have no right to it from their own merit. The simple 
deference inspired by men invested with high social functions 
is not different in nature from religious respect. It is expressed 
by the same movements : a man keeps at a distance from a 
high personage ; he approaches him only with precautions ; 
in conversing with him, he uses other gestures and language 
than those used with ordinary mortals. The sentiment felt 
on these occasions is so closely related to the religious sentiment 
that many peoples have confounded the two. In order to explain 
the consideration accorded to princes, nobles and political 
chiefs, a sacred character has been attributed to them. In 
Melanesia and Polynesia, for example, it is said that an in- 
fluential man has mana, and that his influence is due to this 
mana.^ However, it is evident that his situation is due solely 
to the importance attributed to him by public opinion. Thus 1 / 
the moral power conferred by opinion and that with which sacred / 
beings are invested are at bottom of a single origin and made up 
of the same elements. That is why a single word is able to 
designate the two. 

In addition to men, society also consecrates things, especially 
ideas. If a belief is unanimously shared by a people, then, for 
the reason which we pointed out above, it is forbidden to touch 
it, that is to say, to deny it or to contest it. Now the prohibition 
o f criticism isan interdiction like the others and proves the 
pr esence o f some thing sacred. Even to-day, howsoever great 
niay be theliberty which we accord to others, a man who should 
totally deny progress or ridicule the human ideal to which 
modem societies are attached, would produce the effect of a 
sacrilege. There is at least one principle which those the most 
devoî^4a_the free examination of everything tend to place 
above discussion and to regard as untouchable, that is to say, as 
sacred : this is the very principle of free exajnination. 

^ Codrington, The Melanesians, pp. 50, 103, 120. It is also generally thought 
that in the Polynesian languages, the word mana primitively had the sense of 
authority (see Tregear, Maori Comparative Dictionary, s.v.). 

214 Elementary Forms of Religious Life 

This aptitude of society for setting itself up as a god or for 

creating gods was never more apparent than during' 1ihe first 
years of the French Revolution. At this time, in fact, under 
the influence of the general enthusiasm, things purely laical 
by nature were transformed by public opinion into sacred things : 
these were the Fatherland, Liberty, Reason. ^ A religion tended 
to become established which had its dogmas, ^ symbols,^ altars ^ 
and feasts.^ It was to these spontaneous aspirations that the 
cult of Reason and the Supreme Being attempted to give a sort 
of official satisfaction. It is true that this religious renovation 
had only an ephemeral duration. But that was because the 
patriotic enthusiasm which at first transported the masses soon 
relaxed.* The cause being gone, the effect could not remain. 
But this experiment, though short-lived, keeps all its sociological 
interest. It remains true that in one determined case we have 
seen society and its essential ideas become, directly and with no 
transfiguration of any sort, the object of a veritable cult. 

All these facts allow us to catch glimpses of how the clan 
was able to awaken within its members the idea that outside 
of them there exist forces which dominate them and at the 
same time sustain them, that is to say in fine, religious forces : 
it is because there is no society with which the primitive is more 
directly and closely connected. The bonds uniting him to the 
tribe are much more lax and more feebly felt. Although this is 
not at all strange or foreign to him, it is with the people of his 
own clan that he has the greatest number of things in common ; 
it is the action of this group that he feels the most directly ; so 
it is this also which, in preference to all others, should express 
itself in religious symbols. 

But this first explanation has been too general, for it is ap- 
plicable to every sort of society indifferently, and consequently 
to every sort of rehgion. Let us attempt to determine exactly 
what form this collective action takes in the clan and how it 
arouses the sensation of sacredness there. For there is no place 
where it is more easily observable or more apparent in its results. 

' III 

The life of the Australian societies passes alternately through 
two distinct phases.' Sometimes the population is broken up 
into little groups who wander about independently of one another, 

' See Albert Mathiez, Les origines des cultes révolutionnaires (i 789-1792). 

^ Ibid., p. 24. ' Ihid.. pp. 29, 32. * Ibid., p. 30. ■• Ibid., p. 46. 

* See Mathiez, La Théophilanthropie et la Culte décadaire, p. 36. 

• See Spencer and Gillen, Nor. Tr., p. 33. 

Origins of these Beliefs 215 

in their various occupations ; each family Hves by itself, hunting 
and fishing, and in a word, trying to procure its indispensable 
food by all the means in its power. Sometimes, on the contrary, 
the population concentrates and gathers at determined points 
for-Brlength of time varying from several days to several months. 
This concentration takes place when a clan or a part of the tribe ^ 
is summoned to the gathering, and on this occasion they celebrate 
a religious ceremony, or else hold what is called a corrobbortj 
in the usual ethnological language. ~ 

These two phases are contrasted with each other in the sharpest 
w ay. Inthe, .first, economic activity is the preponderating one, 
and it is generally of a very mediocre intensity. Gathering the 
grains or herbs that are necessary for food, or hunting and fishing 
are not occupations to awaken very lively passions.^ The dis- 
persed condition in which the society finds itself results in making 
its life uniform, languishing and dull.* But when a corrobbori 
takes place, everything changes. Since the emotional and passional 
faculties of the primitive are only imperfectly placed under 
the control of his reason and will, he easily loses control of 
himself. Any event of some importance puts him quite outside 
himself. Does he receive good news ? There are at once trans- 
ports of enthusiasm. In the contrary conditions, he is to be seen 
running here and there like a madman, giving himself up to all 
sorts of immoderate movements, crying, shrieking, rolling in 
the dust, throwing it in every direction, biting himself, brandish- 
ing his arms in a furious manner, etc.^ The very fact of the 
concentration acts as an exceptionally powerful stimulant. 
When they are once come together, a sort of electricity is formed 
by their collecting which quickly transports them to an extra- 
ordinary degree of exaltation. Every sentiment expressed finds ^ 
a place without resistance in all the minds, which are very open 
to outside impressions ; each re-echoes the others, and is re-echoed 

^ There are even ceremonies, for example, those which take place in con- 
nection with the initiation, to which members of foreign tribes are invited. A 
whole system of messages and messengers is organized for these convocations, 
without which the great solemnities could not take place (see Howitt, Notes on 
Australian Message-Sticks and Messengers, in J. A. I., 1889; Nat. Tr., pp. 83, 
678-691 ; Spencer and Gillen, Nat. Tr., p. 159 ; Nor. Tr., p. 551). 

* The corrobbori is distinguished from the real religious ceremonies by the 
fact that it is open to women and uninitiated persons. But if these two sorts of 
collective manifestations are to be distmguished, they are, none the less, closely 
related. We shall have occasion elsewhere to come back to this relationship and 
to explain it. 

* Except, of course, in the case of the great bush-beating hunts. 

* " The peaceful monotony of this part of his life," say Spencer and Gillen 
{Nor. Tr., p. 33). 

■" Howitt, Nat. Tr., p. 683. He is speaking of the demonstrations which take 
place when an ambassador sent to a group of foreigners returns to camp with 
news of a favourable result. Cf. Brough Smyth, I, p. 138 ; Schulze, loc. cit., p. 222. 

2i6 Elementary Forms of Religiotis Life 

by the others. The initial impulse thus proceeds, growing as it 
goes, as an avalanche grows in its advance. And as such active 
passions so free from all control could not fail to burst out, on 
every side one sees nothing but violent gestures, cries, veritable 
howls, and deafening noises of every sort, which aid in intensifying 
still more the state of mind which they manifest. And since a 
collective sentiment cannot express itself collectively except on 
the condition of observing a certain order permitting co-operation 
and movements in unison, these gestures and cries naturally 
tend to become rhythmic and regular ; hence come songs and 
dances. But in taking a more regular form, they lose nothing 
"^of their natural violence ; a regulated tumult remains tumult. 
The human voice is not sufficient for the task ; it is reinforced 
by means of artificial processes : boomerangs are beaten against 
each other ; bull-roarers are whirled. It is probable that these 
instruments, the use of which is so general in the Australian 
religious ceremonies, are used primarily to express in a more 
adequate fashion the agitation felt. But while they express it, 
they also strengthen it. This effervescence often reaches such a 
point that it causes unheard-of actions. The passions released 
are of such an impetuosity that they can be restrained by nothing. 
They are so far removed from their ordinary conditions of life, 
and they are so thoroughly conscious of it, that they feel that 
they must set themselves outside of and above their ordinary 
morals. The sexes unite contrarily to the rules governing sexual 
relations. Men exchange wives with each other. Sometimes 
even incestuous unions, which in normal times are thought 
abominable and are severely punished, are now contracted 
openly and with impunity. ^ If we add to all this that the cere- 
monies generally take place at night in a darkness pierced here 
and there by the light of fires, we can easily imagine what effect 
such scenes ought to produce on the minds of those who partici- 
pate. They produce such a violent super-excitation of the 
whole physical and mental life that it cannot be supported very 
long : the actor taking the principal part finally falls exhausted 
on the ground. 2 

To illustrate and make specific this necessarily schematic 
picture, let us describe certain scenes taken from Spencer and 

* See Spencer and Gillen, Nat. Tr., pp. 96 f. ; Nor. Tr., p. 137 ; Brough 
Smyth, II, p. 319. — This ritual promiscuity is found especially in the initiation 
ceremonies (Spencer and Gillen, Nat. Tr.. pp. 267, 381 ; Howitt, Nat. Tr., 
p. 657), and in the totemic ceremonies {Nor. Tr., pp. 214, 298, 237). In these 
latter, the ordinary exogamic rules are violated. Sometimes among the Arunta, 
unions between father and daughter, mother and son, and brothers and sisters 
(that is in every case, relationship by blood) remain forbidden {Nat. Tr., pp. 96 f.). 

* Howitt, Nat. Tr., pp. 535, 545. This is extremely common. 

Origins of these Beliefs 217 

One of the most important religious ceremonies among the 
Warramunga is the one concerning the snake Wollunqua. It 
consists in a series of ceremonies lasting through several days. 
On the fourth day comes the following scene. 

According to the ceremonial used among the Warramunga, 
representatives of the two phratries take part, one as officiants, 
the other as preparers and assistants. Only the members of the 
Uluuru phratry are qualified to celebrate the rite, but the members 
of the Kingilli phratry must decorate the actors, make ready 
the place and the instruments, and play the part of an audience. 
In this capacity, they were charged with making a sort of mound 
in advance out of wet sand, upon which a design is marked 
with red down which represents the snake Wollunqua. The 
real ceremony only commenced after nightfall. Towards ten or 
eleven o'clock, the Uluuru and Kingilli men arrived on the ground, 
sat down on the mound and commenced to sing. Everyone was 
evidently very excited. A little later in the evening, the Uluuru 
brought up their wives and gave them over to the Kingilli, ^ 
who had intercourse with them. Then the recently initiated 
young men were brought in and the whole ceremony was explained 
to them in detail, and until three o'clock in the morning singing 
went on without a pause. Then followed a scene of the wildest 
excitement. While fires were lighted on all sides, making the 
whiteness of the gum-trees stand out sharply against the sur- 
rounding darkness, the Uluuru knelt down one behind another 
beside the mound, then rising from the ground they went around 
it, with a movement in unison, their two hands resting upon 
their thighs, then a little farther on they knelt down again, and 
so on. At the same time they swayed their bodies, now to the 
right and now to the left, while uttering at each movement a 
piercing cry. a veritable yell, " Yrrsh ! Yrrsh ! Yrrsh ! " In 
the meantime the Kingilli, in a state of great excitement, clanged 
their boomerangs and their chief was even more agitated than 
his companions. When the procession of the Uluuru had tv\àce 
gone around the mound, quitting the kneehng position, they 
sat down and commenced to sing again ; at moments the singing 
died away, then suddenly took up again. When day commenced 
to dawn, all leaped to their feet ; the fires that had gone out 
were relighted and the Uluuru, urged on by the Kingilli, attacked 
the mound furiously with boomerangs, lances and clubs ; in a 
few minutes it was torn to pieces. The fires died away and pro- 
found silence reigned again. 2 

* These women were Kingilli themselves, so these unions violated the 
exogamic rules. 

* Nor. Tr., p. 237. 

2i8 Elementary Forms of Religious Life 

A still more violent scene at which these same observers 
assisted was in connection with the fire ceremonies among the 

Commencing at nightfall, all sorts of processions, dances and 
songs had taken place by torchlight ; the general effervescence 
was constantly increasing. At a given moment, twelve assistants 
each took a great lighted torch in their hands, and one of them 
holding his like a bayonet, charged into a group of natives. 
Blows were warded off with clubs and spears. A general mêlée 
followed. The men leaped and pranced about, uttering savage 
yells all the time ; the burning torches continually came crashing 
d( wn on the heads and bodies of the men, scattering lighted 
sparks in every direction. " The smoke, the blazing torches, 
the showers of sparks falling in all directions and the masses 
of dancing, yelling men," say Spencer and Gillen, " formed 
altogether a genuinely wild and savage scene of which it is im- 
possible to convey any adequate idea in words." ^ 

One can readily conceive how, when arrived at this state of 
exaltation, a man does not recognize himself any longer. Feeling 
himself dominated and carried away by some sort of an external 
power which makes him think and act differently than in normal 
times, he naturally has the impression of being himself no longer. 
It seems to him that he has becoma-a-newieing : the decorations 
he puts on and the masks that cover his face figure materially 
in this interior transformation, and to a still greater extent, 
they aid in determining its nature. And as at the same time all 
his companions feel themselves transformed in the same way 
and express this sentiment by their cries, their gestures and their 
general attitude, everything is just as though he really were 
transported into a special world, entirely different from the 
one where he ordinarily lives, and into an environment filled 
with exceptionally intense forces that take hold of him and 
metamorphose him. How could such experiences as these, 
especially when they are repeated every day for weeks, fail to 
leave in him the conviction that there really exist two hetero- 
geneous and mutually incomparable worlds ? One is that where 
his daily life drags wearily along ; but he cannot penetrate into 
the other without at once entering into relations with extra- 
ordinary powers that excite him to the point of frenzy .__Ihe_first 
is the profane world, the second, that of sacred things^^ 

So it is in the midst of these effervescent social environments 
and out of this effervescence itself that the rehgious idea seems 

^ Nor. Tr., p. 391. Other examples of this collective effervescence during 
the religious ceremonies will be found in Nat. Tr., pp. 244-246. 365-366, 374, 509- 
510 (this latter in connection with a funeral rite). Cf. Nor. Tr., pp. 213, 351. 

Origins of these Beliefs 219 

bom. The theory that this is really its origin is confirmed 
by the fact that in Australia the really religious activity is almost 
entirely confined to the moments when these assemblies are 
held. To be sure, there is no people among whom the great 
solemnities of the cult are not more or less periodic ; but in the 
more advanced societies, there is not, so to speak, a day when 
some prayer or offering is not addressed to the gods and some 
ritual act is not performed. But in Australia, on the contrary, 
apart from the celebrations of the clan and tribe, the time is 
nearly all filled with lay and profane occupations. Of course 
there are prohibitions that should be and are preserved even 
during these periods of temporal activity ; it is never permissible 
to kill or eat freely of the totemic animal, at least in those parts 
where the interdiction has retained its original vigour ; but 
almost no positive rites are then celebrated, and there are no 
ceremonies of any importance. These take place only in the 
midst of assembled groups. The religious life of the Australian 
passes through successive phases of complete lull and of super- 
excitation, and social life oscillates in the same rhythm. This 
puts clearly into evidence the bond uniting them to one another, 
but among the peoples called civilized, the relative continuity 
of the two blurs their relations. It might even be asked whether 
the violence of this contrast was not necessary to disengage the 
feeling of sacredness in its first form. By concentrating itself 
almost entirely in certain determined moments, the. collective 
life has been able to attain its greatest intensity and efficacy, 
and consequently to give men a more active sentiment of the 
double existence they lead and of the double nature in which 
they participate. 

But this explanation is still incomplete. We have shown 
how the clan, by the manner in which it acts upon its members, 
awakens within them the idea of external forces which dominate 
them and exalt them ; but we must still demand how it happens 
that these forces are thought of under the form of totems," that 
is^to say, in the shape of an animal or plant. 

It is because this animal or plant has given its name to the 
clan and serves it as emblem. In fact, it is a well-known Taw 
that the sentiments aroused in us by something spontaneously 
attach themselves to the symbol which represents them. For 
us, black is a sign of mourning ; it also suggests sad impressions 
and ideas. This transference of sentiments comes simply from 
the fact that the idea of a thing and the idea of its symbol are 
closely united in our minds ; the result is that the emotions pro- 
voked by the one, extend contagiously to the other. But this 

220 Elementary Forms of Religious Life 

contagion, which takes place in every case to a certain degree, 
is much more complete and more marked when the symbol is 
something simple, definite and easily representable, while the 
thing itself, owing to its dimensions, the number of its parts and 
the complexity of their arrangement, is difficult to hold in the 
mind. For we are unable to consider an abstract entity, which 
we can represent only laboriously and confusedly, the source 
of the strong sentiments which we feel. We cannot explain 
them to ourselves except by connecting them to some concrete 
object of whose reality we are vividly aware. Then if the thing 
itself does not fulfil this condition, it cannot serve as the accepted 
basis of the sentiments felt, even though it may be what really 
aroused them. Then some sign takes its place ; it is to this that 
we connect the emotions it excites. It is this which is loved, 
feared, respected ; it is to this that we are grateful ; it is for 
this that we sacrifice ourselves. The soldier who dies for his flag, 
dies for his country ; but as a matter of fact, in his own con- 
sciousness, it is the flag that has the first place. It sometimes 
happens that this even directly determines action. Whether 
one isolated standard remains in the hands of the enemy or not 
does not determine the fate of the country, yet the soldier allows 
himself to be killed to regain it. He loses sight of the fact that 
the flag is only a sign, and that it has no value in itself, but only 
brings to mind the reality that it represents ; it is treated as if 
it were this reality itself. 

Now the totem is the flag of the clan. It is therefore natural 
that the impressions aroused by the clan in individual minds — 
impressions of dependence and of increased vitality — should 
fix themselves to the idea of the totem rather than that of the 
clan : for the clan is too complex a reality to be represented 
clearly in all its complex unity by such rudimentary intelligences. 
More than that, the primitive does not even see that these 
impressions come to him from the group, jHe does not know 
that the coming together of a number of men associated in the 
same life results in disengaging new energies, which transform 
each of the mTj All that he knows is that he is raised above 
himself and tîîât he sees a different life from the one he ordinarily 
leads. However, he must connect these sensations to some 
external object as their cause. Now what does he see about 
him ? On every side those things which appeal to his senses 
and strike his imagination are the numerous images of the totem. 
They are the waninga and the nurtunja, which are symbols of 
the sacred being. They are churinga and bull-roarers, upon which 
are generally carved combinations of lines having the same 
significance. They are the decorations covering the different 

Origins of these Beliefs 221 

parts of his body, which are totemic marks. How could this 
image, repeated everywhere and in all sorts of forms, fail to 
stand out with exceptional relief in his mind ? Placed thus 
in the centre of the scene, it becomes representative. The senti- 
ments experienced fix themselves upon it, for it is the only 
concrete object upon which they can fix themselves. It continues 
to bring them to mind and to evoke them even after the assembly 
has dissolved, for it survives the assembly, being carved upon the 
instruments of the cult, upon the sides of rocks, upon bucklers, 
etc. By it, the emotions experienced are perpetually sustained 
and revived. Everything happens just as if they inspired them 
directly. It is still more natural to attribute them to it for, 
since they are common to the group, they can be associated only 
with something that is equally common to all. Now the totemic 
emblem is the only thing satisfying this condition. By definition, 
it is common to all. During the ceremony, it is the centre of all 
regards. While generations change, it remains the same ; it is 
the permanent element of the social life. So it is from it that 
those mysterious forces seem to emanate with which men feel 
that they are related, and thus they have been led to represent 
these forces under the form of the animate or inanimate being 
whose name the clan bears. 

When this point is once established, we are in a position to 
understand all that is essential in the totemic beliefs. —^ 

Since religious force is nothing other than the collective and \ 
anonymous force of the clan, and since this can be represented 
in the mind only in the form of the totem, the totemic emblem 
is like the visible body of the god ^Therefore, it is from it that 
those kindly or dreadful actions seem to emanate, which the 
cult seeks to provoke or prevent ; consequently, it is to it that 
the cult is addressed. This is the explanation of why it holds 
the first place in the series of sacred things. J 

But the clan, like every other sort of society, can live only in 
and through the individual consciousnesses that compose it. 
So if religious force, in so far as it is conceived as incorporated 
in the totemic emblem, appears to be outside of the individuals 
and to be endowed with a sort of transcendence over them, 
it, like the clan of which it is the symbol, can be realized only in 
and through them ; in this sense, it is imminent in them and 

they necessarily represent it as such. They feel it present and_ 

active within them, for it is this which raises thëm to a superior 
life. This is why men have believed that they contain within 
them a principle comparable to the one residing in the totem, 
and consequently, why they have attributed a sacred character to 
themselves, but one less marked than that of the emblem. It is 

222 Elementary Forms of Religious Life 

because the emblem is the pre-eminent source of the reUgious 
life ; the man participates in it only indirectly, as he is well 
aware ; he takes into account the fact that the force that trans- 
ports him into the world of sacred things is not inherent in him, 
but comes to him from the outside. 

But for still another reason, the animals or vegetables of the 
totemic species should have the same character, and even to a 
higher degree. If the totemic principle is nothing else than the 
clan, it is the clan thought of under the material form of the 
totemic emblem ; now this form is also that of the concrete 
beings whose name the clan bears. Owing to this resemblance, 
they could not fail to evoke sentiments analogous to those aroused 
by the emblem itself. Since the latter is the object of a religious 
respect, they too should inspire respect of the same sort and 
appear to be sacred. Having external forms so nearly identical, 
it would be impossible for the native not to attribute to them 
forces of the same nature. It is therefore forbidden to kill or 
eat the totemic animal, since its flesh is believed to have the 
positive virtues resulting from the rites ; it is because it resembles 
the emblem of the clan, that is to say, it is in its own image. 
And since the animal naturally resembles the emblem more 
than the man does, it is placed on a superior rank in the hierarchy 
of sacred things. Between these two beings there is undoubtedly 
a close relationship, for they both partake of the same essence : 
both incarnate something of the totemic principle. However, 
since the principle itself is conceived under an animal form, the 
animal seems to incarnate it more fully than the man. There- 
fore, if men consider it and treat it as a brother, it is at least 
as an elder brother. ^ 

But even if the totemic principle has its preferred seat in a 
determined species of animal or vegetable, it cannot remain 
localized there. A sacred character is to a high degree contagious ; ^ 
it therefore spreads out from the totemic being to everything that 
is closely or remotely connected with it. The religious sentiments 
inspired by the animal are communicated to the substances 
upon which it is nourished and which serve to make or remake 
its flesh and blood, to the things that resemble it, and to the 
different beings with which it has constant relations. Thus, 
little by little, sub-totems are attached to the totems and from 

* Thus we see that this fraternity is the logical consequence of totemism, 
rather than its basis. Men have not imagined their duties towards the animals 
of the totemic species because they regarded them as kindred, but have imagined 
the kinship to explain the nature of the beliefs and rites of which they were the 
object. The animal was considered a relative of the man because it was a sacred 
being like the man , but it was not treated as a sacred being because it was 
regarded as a relative. » See below, Bk. Ill, ch i, § 3. 

Origins of these Beliefs 223 

the cosmological systems expressed by the primitive classifica- 
tions. At last, the whole worid is divided up among the totemic 
principles of each tribe. -, 

We are now able to explain the origin, of the ambiguity of 
religious forces as they appear in history, and how they are 
physical as well as human, moral as well as material. They are 
moral powers because they are made up entirely of the impressions 
this moral being, the group, arouses in those other moral beings, 
its individual members ; they do not translate the manner in 
which physical things affect our senses, but the way in which 
the collective consciousness acts upon individual consciousnesses. 
Their authority is only one form of the moral ascendancy of 
society over its members. But, on the other hand, since they 
are conceived of under material forms, they could not fail to be 
regarded as closely related to material things.^ Therefore they 
dominate the two worlds. Their residence is in men, but at the 
same time they are the vital principles of things. They animate 
minds and discipline them, but it is also they who make plants 
grow and animals reproduce. It is this double nature which 
has enabled religion to be like the womb from which come all 
the leading germs of human civilization. Since it has been made 
to embrace all of reality, the physical world as well as the moral 
one, the forces that move bodies as well as those that move 
minds have been conceived in a religious form. That is how 
the most diverse methods and practices, both those that make 
possible the continuation of the moral life (law, morals, beaux- 
arts) and those serving the material life (the natural, technical 
and practical sciences), are either directly or indirectly derived 
from religion.^ 


The first religious conceptions have often been attributed 
to feelings of weakness and dependence, of fear and anguish 
which seized men when they came into contact with the world. 
Being the victims of nightmares of which they were themselves 

^ At the bottom of this conception there is a well-founded and persistent 
sentiment. Modern science also tends more and more to admit that the duality 
of man and nature does not exclude their unity, and that physical and moral 
forces, though distinct, are closely related. We undoubtedly have a different 
conception of this unity and relationship than the primitive, but beneath these 
different symbols, the truth affirmed by the two is the same. 

* We say that this derivation is sometimes indirect on account of the 
industrial methods which, in a large number of cases, seem to be derived from 
religion through the intermediacy of magic (see Hubert and Mauss, Théorie générale 
de la Magie, Année Social.. Vil, pp. 144 fî.) ; for, as we believe, magic forces 
arc only a special form of religious forces. We shall have occasion to return to 
this point several times. 


224 Elementary Forms of Religious Life 

the creators, they believed themselves surrounded by hostile 
and redoubtable powers which their rites sought to appease. 
We have now shown that the first religions were of a wholly 
different origin. The famous formula Primus in orbe deos fecit 
iimor is in no way justified by the facts. The primitive does not 
regard his gods as foreigners, enemies or thoroughly and neces- 
sarily malevolent beings whose favours he must acquire at any 
price ; quite on the contrary, they are rather friends, kindred or 
natural protectors for him. Are these not the names he gives to 
the beings of the totemic species ? The power to which the cult 
is addressed is not represented as soaring high above him and 
overwhelming him by its superiority ; on the contrary, it is 
very near to him and confers upon him very useful powers which 
he could never acquire by himself. Perhaps the deity has never 
been nearer to men than at this period of history, when it is 
present in the things filling their immediate environment and is, 
in part, imminent in himself. In fine, the sentiments at the root 
of totemism are those of happy confidence rather than of terror 
and compression. If we set aside the funeral rites — the sober 
side of every religion — we find the totemic cult celebrated in 
the midst of songs, dances and dramatic representations. As we 
shall see, cruel expiations are relatively rare ; even the painful 
and obligatory mutilations of the initiations are not of this 
character. The terrible and jealous gods appear but slowly in 
the religious evolution. This is because primitive societies are 
not those huge Leviathans which overwhelm a man by the 
enormity of their power and place him under a severe discipline ; ^ 
he gives himself up to them spontaneously and without resistance. 
As the social soul is then made up of only a small number of 
ideas and sentiments, it easily becomes wholly incarnate in each 
individual consciousness. The individual carries it all inside of 
him ; it is a part of him and consequently, when he gives himself 
up to the impulses inspired by it, he does not feel that he is giving 
way before compulsion, but that he is going where his nature calls 
him. 2 

This way of understanding the origins of religious thought 
escapes the objections raised against the most accredited classical 

We have seen how the naturists and animists pretend to con- 
struct the idea of sacred beings out of the sensations evoked in 
us by different phenomena of the physical or biological order, 

^ At least after he is once adult and fully initiated, for the initiation rites, 
introducing the young man to the yocial life, are a severe discipline in themselves. 

^ Upon this particular aspect of primitive societies, see our Division du travail 
social, 3rd éd., pp. 123, 149, 173 ff. 

Origins of these Beliefs 225 

and we have shown how this enterprise is impossible and even 
self-contradictory. Nothing is worth nothing. The impressions 
produced in us by the physical world can, by definition, contain 
nothing that surpasses this world. Out of the visible, only the 
visible can be made ; out of that which is heard, we cannot 
make something not heard. Then to explain how the idea of 
sacredness has been able to take form under these conditions, 
the majority of the theorists have been obliged to admit that 
men have superimposed upon reality, such as it is given by 
observation, an unreal world, constructed entirely out of the 
fantastic images which agitate his mind during a dream, or else 
out of the frequently monstrous aberrations produced by the 
mythological imagination under the bewitching but deceiving 
influence of language. But it remained incomprehensible that 
humanity should have remained obstinate in these errors through 
the ages, for experience should have very quickly proven them 

But from our point of view, these difficulties disappear. 
Religion ceases to be an inexplicable hallucina.tion and takes a 
fpothold in reality. In fact, we can say that the believer is not 
deceived when he believes in the existence of a mural power 
upon which he depends and from which he receives all that 
is best in himself : this_^ower exists, it is society. When the 
Australian is carriei3~outside himself and feels a new life flowing 
within him whose intensity surprises him, he is not the dupe of 
an illusion ; this exaltation is real and is really the effect of forces 
outside of and superior to the individual. It is true that he is 
wrong in thinking that this increase of vitality is the work of a 
power in the form of some animal or plant. But this error is 
merely in regard to the letter of the symbol by which this being 
is represented to the mind and the external appearance which 
the imagination has given it, and not in regard to the fact of its 
existence. Behind these figures and metaphors, be they gross or 
refined, there" is à concrete and living reality. Thus religion 
acquires a meaning and a reasonableness that the most intransigent 
rationalist cannot misunderstand. (Jts primary object is not to 
give men a representation of the physical world ; for if that were 
its essential task, we could not understand how it has been able 
to survive, for, on this side, it is scarcely more than a fabric of 
errors. Before all, it is a system of ideas with which the individuals 
represent to themselves the society of which they are members, , 
andjtiljg_i)i)Sçure but intiinate relations which they have with it^ 
This is its primary function ; and though metapHorical and 
symbolic, this representation is not unfaithful. Quite on the 
contrary, it translates everything essential in the relations which 

226 Elementary Forms of Religious Life 

are to be explained : for it is an eternal truth that outside of us 
there exists something greater than us, with which we enter into 

That is why we can rest assured in advance that the practices 
of the cult, whatever they may be, are something more than 
movements without importance and gestures without efficacy. 
By the mere fact that their apparent function is to strengthen 
the bonds attaching the believer to his god, they at the same 
time really strengthen the bonds attaching the individual to 
the society of which he is a member, since the god is only a 
figurative expression of the society. We are even able to under- 
stand how the fundamental truth thus contained in religion has 
been able to compensate for the secondary errors which it almost 
necessarily implies, and how believers have consequently been 
restrained from tearing themselves off from it, in spite of the 
misunderstandings which must result from these errors. It is 
undeniably true that the recipes which it recommends that men 
use to act upon things are generally found to be ineffective. 
But these checks can have no profound influence, for they do 
not touch religion in its fundamentals. ^ 

However, it may be objected that even according to this 
hypothesis, religion remains the object of a certain delirium. 
What other name can we give to that state when, after a collective 
effervescence, men believe themselves transported into an entirely 
different world from the one they have before their eyes ? 

It is certainly true that religious life cannot attain a certain 
degree of intensity without implying a psychical exaltation not 
far removed from delirium. That is why the prophets, the 
founders of religions, the great saints, in a word, the men whose 
religious consciousness is exceptionally sensitive, very frequently 
give signs of an excessive nervousness that is even pathological : 
these physiological defects predestined them to great religious 
rôles. The ritual use of intoxicating liquors is to be explained 
in the same way.^ Of course this does not mean that an ardent 
religious faith is necessarily the fruit of the drunkenness and 
mental derangement which accompany it ; but as experience 
soon informed people of the similarities between the mentality 
of a delirious person and that of a seer, they sought to open a 
way to the second by artificially exciting the first. But if, for 
this reason, it may be said that religion is not without a certain 
delirium, it must be added that this delirium, if it has the causes 
which we have attributed to it, is well-founded. The images out 

1 We provisionally limit ourselves to this general indication : we shall return 
to this idea and give more explicit proof, when we speak of the rites (Bk. III). 
* On this point, see Achelis. Die Ekstase, Berlin, 1902, especially ch. i. 

Origins of these Beliefs 227 

of which it is made are not pure illusions like those the naturists 
and animists put at the basis of religion ; they correspond to 
something in reality. Of course it is only natural that the moral 
forces they express should be unable to affect the human mind 
powerfully without pulling it outside itself and without plunging 
it into a state that may be called ecstatic, provided that the word 
be taken in its etymological sense {eKcrrao-tç) ; but it does not 
follow that they are imaginary. Quite on the contrary, the mental 
agitation they cause bears witness to their reality. It is merely 
one more proof that a very intense social life always does a sort 
of violence to the organism, as well as to the individual conscious- 
ness, which interferes with its normal functioning. Therefore it 
can last only a limited length of time.^ 

Moreover, if we give the name delirious to every state in which 
the mind adds to the immediate data given by the senses and 
projects its own sentiments and feelings into things, then nearly 
every collective representation is in a sense delirious ; religious 
beliefs are only one particular case of a very general law. Our 
whole social environment seems to us to be tilled with forces 
which really exist only in our own minds. We know what the 
flag is for the soldier ; in itself, it is only a piece of cloth. Human 
blood is only an organic liquid, but even to-day we cannot see it 
flowing without feeling a violent emotion which its physico- 
chemical properties cannot explain. From the physical point of 
view, a man is nothing more than a system of cells, or from the 
mental point of view, than a system of representations ; in 
either case, he differs only in degree from animals. Yet society 
conceives him, and obliges us to conceive him, as invested with 
a character sui generis that isolates him, holds at a distance 
all rash encroachments and, in a word, imposes respect. This 
dignity which puts him into a class by himself appears to us as 
one of his distinctive attributes, although we can find nothing 
in the empirical nature of man which justifies it. A cancelled 
postage stamp may be worth a fortune ; but siKely this value 
is in no way implied in its natural properties. In a sense, our 
representation of the external world is undoubtedly a mere 
fabric of hallucinations, for the odours, tastes and colours that 
we put into bodies are not really there, or at least, they are 
not such as we perceive them. However, our olfactory, gustatory 
and visual sensations continue to correspond to certain objective 
states of the things represented ; they express in their way 
the properties, either of material particles or of ether waves, 
which certainly have their origin in the bodies which we perceive 

^ Cf. Mauss, Essai sur les variations saisonnières des sociétés eskimos, in Année 
Social.. IX, p. 127. 

228 Elementary Forms of Religious Life 

as fragrant, sapid or coloured. But collective representations very 
frequently attribute to the things to which they are attached quali- 
ties which do not exist under any form or to any degree. Out of the 
commonest object, they can make a most powerful sacred being. 
Yet the powers which are thus conferred, though purely 
ideal, act as though they were real ; they determine the conduct 
of men with the same degree of necessity as physical forces. 
The Arunta who has been rubbed with his churinga feels himself 
stronger ; he is stronger. If he has eaten the flesh of an animal 
which, though perfectly healthy, is forbidden to him, he will 
feel himself sick, and may die of it. Surely the soldier who falls 
while defending his flag does not believe that he sacrifices himself 
for a bit of cloth. This is all because social thought, owing to the 
imperative authority that is in it, has an efficacy that individ\ial 
thought could never have ; by the power which it has over our 
minds, it can make us see things in whatever light it pleases ; 
it adds to reality or deducts from it according to the circum- 
stances. Thus there is one division of nature where the formula 
of idealism is applicable almost to the letter : this is the social 
kingdom. Here more than anywhere else, the idea is the reality. 
Even in this case, of course, idealism is not true without modifica- 

rtion. We can never escape the duality of our nature and free 
ourselves completely from physical necessities : in order to express 
our own ideas to ourselves, it is necessary, as has been shown 
above, that we fix them upon material things which S3nnbolize 
them. But here the part of matter is reduced to a minimum. 
The object serving as support for the idea is not much in 
comparison with the ideal superstructure, beneath which it 
disappears, and also, it counts for nothing in the superstructure. 
This is what that pseudo-delirium consists in, which we find at 
the bottom of so many collective representations : it is. only a 
form of this essential idealism.* So it is not properly called a 
delirium, for the ideas thus objectified are well founded, not in 
the nature of the material things upon which they settle them- 
^ selves, but in the nature of society. 

We are now able to understand how the totemic principle, 

1 Thus we see how erroneous those theories are which, like the geographical 
materialism of Ratzel (see especially his Politische Géographie), seek to derive all 
social life from its material foundation (either economic or territorial). They 
commit an error precisely similar to the one committed by Maudsley in individual 
psychology. Just as this latter reduced all the psychical life of the individual to 
a mere epiphenomenon of his physiological basis, they seek to reduce the whole 
psyc hical life of the group to its physical basis. But they forget that ideas are 
realities and forces, and that collective representations are forces even more 
powerful and active than individueil representations. On this point, see our 
Représentations individuelles et représentations collectives, in the Revue de Méta- 
physique et de Morale, May, 1898. 

Origins of these Beliefs 229 

and in general, every religious force, comes to be outside of the 
object in which it resides.^ It is because the idea of it is in no 
way made up of the impressions directly produced by this thing 
upon our senses or minds. Religious force is only the sentiment"^ 
inspired by the group in its members, but projected outside of 
the consciousnesses that experience them, and objectified. To 
be objectified, they are fixed upon some object which thus becomes 
sacred ; but any object might fulfil this function. In principle_^ 
there are none whose nature predestines them to it to the ex- 
clusion of all others ; but also there are none that are necessarily 
impossible. 2 Everything depends upon the circumstances which 
lead the sentiment creating religious ideas to establish itself 
here or there, upon this point or upon that one. Therefore, the 
sacred character assumed by an object is not implied in the ' 
intrinsic properties of this latter : it is added to them. The world 
of religious things is not one particular aspect of empirical nature ; 
it is superimposed upon it. J 

This conception of the religious, finally, allows us to explain'^ 
an important principle found at the bottom of a multitude of 
myths and rites, and which may be stated thus : when a sacred 
thing is subdivided, each of its parts remains equal to the thing 
itself. In other words, as far as religious thought is concerned, 
the part is equal to the whole ; it has the same powers, the same 
efficacy. The debris of a relic has the same virtue as a relic in 
good condition. The smallest drop of blood contains the same 
active principle as the whole thing. The soul, as we shall see, 
may be broken up into nearly as many pieces as there are organs 
or tissues in the organism ; each of these partial souls is worth 
a whole soul. This conception would be inexplicable if the 
sacredness of something were due to the constituent properties 
of the thing itself ; for in that case, it should vary with this thing, 
increasing and decreasing with it. But if the virtues it is believed 
to possess are not intrinsic in it, and if they come from certain 
sentiments which it brings to mind and symbolizes, though these 
originate outside of it, then, since it has no need of determined 
dimensions to play this rôle of reminder, it will have the same 
value whether it is entire or not. Since the part makes us think 
of the whole, it evokes the same sentiments as the whole. A mere ^ 
fragment of the flag represents the fatherland just as well as the 
flag itself : so it is sacred in the same way and to the same degree.^ 

^ See above, pp. i88 and 1Q4. 

* Even the excreta have a religious character. See Preuss, Der Ursprung der 
Religion und Kunst, especially ch. ii, entitled Der Zauber der DefakcUion [Globus, 
LXXXVI. pp. 325 ff.). 

' This principle has passed from religion into magic : it is the totem ex parte 
of the alchemists. 

230 Elementary Forms of Religious Life 

But if this theory of totemism has enabled us to explain 
the most characteristic beliefs of this religion, it rests upon a 
fact not yet explained. When the idea of the totem, the emblem 
of the clan, is given, all the rest follows ; but we must still investi- 
gate how this idea has been formed. This is a double question 
and may be subdivided as follows : What has led the clan to 
choose an emblem ? and why have these emblems been borrowed 
from the animal and vegetable worlds, and particularly from the 
former ? 

That an emblem is useful as a rallying-centre for any sort of 
a group it is superfluous to point out. By expressing the social 
unity in a material form, it makes this more obvious to all, and 
for that very reason the use of emblematic symbols must have 
spread quickly when once thought of. But more than that, 
this idea should spontaneously arise out of the conditions of 
common life ; for the emblemjs not merely a convenient process 
for clarifying the sentiment society has of itself : it also serves to 
^ create this sentiment ; it is one of its constituent elements. 

In fact, if left to themselves, individual consciou sness es a re 
closed to each other ; they can, communicate only by means of 
signs which express their internal states. If the communication 
established between them is to become a- real communion, that 
is to say, a fusion of all particjjlar sentiments into one common 
sentiment, the signs expressing them must themselves be fused 
into one single and unique resultant. It is the appearance of 
this that informs individuals that they are in harmony and makes 
X them conscious of their moral unity. It is by uttering the same 
cry, pronouncing the same word, or performing the same gesture 
in regard to some object that they become and feel themselves 
to be in unison. It is true that individual representations also 
cause reactions in the organism that are not without importance ; 
however, they can be thought of apart from these physical 
reactions which accompany them or follow them, but which do 
not constitute them. But it is quite another matter with collective 
/ representations. They presuppose that minds act and react 
upon one another ; they are the product of these actions and 
reactions which are themselves possible only through material 
' intermediaries. These latter do not confine themselves to reveal- 
ing the mental state with which they are associated ; they aid in 
creating it. Individual minds cannot come in contact and com- 
municate with each other except by coming out of themselves ; 
but they cannot do this except by movements. So it is the homo- 
geneity of these movements that gives the group consciousness 

Origins of these Beliefs 231 

of itself and consequently makes it exist. When this homo- 
geneitjTis once established and these movements have once taken 
a stereotyped form, they serve to symbolize the corresponding 
representations. But they symbolize them only because they 
have aided in forming them. 

Moreover, without symbols, social sentiments_could have only 
a precarious existence. Though very strong as long as men are 
together and influence each other reciprocally, they exist only 
in the form of recollections after the assembly has ended, and 
when left to themselves, these become feebler and feebler ; for 
since the group is now no longer present and active, individual 
temperaments easily regain the upper hand. The violent passions 
which may have been releas ed in the heart of a crowd fall away 
and are extinguished when this is dissolved, and men ask them- 
selves with astonishment how they could ever have been so 
carried away from their normal character. But if the movements 
by which these sentiments are expressed are connected with 
something that endures, the sentiments t hemselves become more 
durable. These other things are constantly bringing them to 
piind and arousing them ; it is as though the cause wHich excited 
tKëm in the first place continued to act. Thus these systems 
of emblems, which are necessary if society is to become conscious 
of itself, are no less indispensable for assuring the cont inuation 
of this conscipusness. 

So we must refrain from regarding these symbols as simple 
artifices , as sorts of Jabels attached to representations already 
made, in order to make them more manageable : they are an 
intégral part of them. Even the fact that collective sentiments 
are thus attached to things completely foreign to them is not 
purely conventional : it illustrates under a conventional form 
a real characteristic of social facts, that is, their t rans cendence 
over individual minds. In fact, it is known that social phenomena 
are bom, not in individuals, but i n the group. Whatever part we 
may take in their origin, each of us receives them from without.^ 
So when we represent them to ourselves as emanating from a 
niaterial object, we do ■ not completely misunderstand their 
nature."" Of course they do not come from the specific thing 
to which we connect them, but nevertheless, it is true that their 
origin is outside of us. If the moral force sustaining the believer 
does not come from the idol he adores or the emble m he venerates, 
still it is from outside ofhim, as he is well aware. The objectivity 
of its_symbol_only translates its externalness. 

Thus sociaMife, in all its aspects and in every period of its 
history, is made possible only by a vast symbolism. The material 

* On this point see Règles de la méthode sociologique, pp. 5 ff. 

232 Elementary Forms of Religious Life 

emblems and figurative representations with which we are more 
especially concerned in our present study, are one form of this ; 
but there are many others. Collective sentiments can just 
as well become incarnate in persons or formulae : some formulae 
are flags, while there are persons, either real or niythical, who 
are symbols. But there is one sort of emblem which should 
make an early appearance without reflection or calculation : 
this is tattooing. Indeed, well-known facts demonstrate that it 
is produced almost automatically in certain conditions. When 
men of an inferior culture are associated in a common life, they 
are frequently led, by an instinctive tendency, as it were, to 
paint or cut upon the body, images that bear witness to their 
common existence. According to a text of Procopius, the early 
Christians printed on their skin the name of Christ or the sign 
of the cross ; ^ for a long time, the groups of pilgrims going to 
Palestine were also tattooed on the arm or wrist with designs 
representing the cross or the monogram of Christ. ^ This same 
usage is also reported among the pilgrims going to certain holy 
places in Italy. ^ A curious case of spontaneous tattooing is 
given by Lombroso : twenty young men in an Italian college, 
when on the point of separating, decorated themselves with 
tattoos recording, in various ways, the years they had spent 
together.* The same fact has frequently been observed among 
the soldiers in the same barracks, the sailors in the same boat, 
or the prisoners in the samQ_jaiL^ It will be understood that 
especially where methods are still rudimentary, tattooing should 
be the most direct and expressive means by which thç„communion 
of minds can be affirmed. The best way of proving to one's 
self and to others that one is a member of a certain group is to 
place a distinctive mark on the body. The proof that this is the 
reason for the existence of the totemic image is the fact, which 
we have already mentioned, that it does not seek to reproduce 
the aspect of the thing it is supposed to represent. It is made 
up of lines and points to which a wholly conventional significance 
is attributed.^ Its object is not to represent or bring to mind 
a determined object, but to bear witness to the fact that a certain 
number of individuals participate in the same moral life. 

Moreover, the clan is a society which is less able than any 
other to do without an emblem or symbol, for there is almost 

* Procopius of Gaza, Commentarii in Isaiam, 496. 

* See Thévenot, Voyage au Levant, Paris, i68g, p. 638. The fact was still 
lound in 1862, 

' Lacassagne, Les Tatouages, p. 10. 

* Lombroso, L'homme criminel, I, p. 292. 

•" Lombroso, ibid., I, pp. 268, 285, 291 f. ; Lacassagno, op. cit., p. 97. 
" See above, p. 127, 

Origins of these Beliefs 233 

no other so lacking in consistency. The clan cannot be defined 
by its chief, for if central authority is not lacking, it is at least 
uncertain and unstable.^ Nor can it be defined by the territory 
it occupies, for the population, being nomad, ^ is not closely 
attached to any special locality. Also, owing to the exogamic 
law, husband and wife must be of different totems ; so wherever 
the totem is transmitted in the maternal line — and this system 
of filiation is still the most general one * — the children are of a 
difterent clan from their father, though living near to him. 
Therefore we find representatives of a number of different 
clans in each family, and still more in each locality. The unity 
of the group is visible, therefore, only in the collective name 
borne by all the members, and in the equally collective emblem 
reproducing the object designated by this name. A clan is 
essentially a reunion of individuals who bear the same name 
and rally around the same sign. Take away the name and the 
sign which materializes it, and the clan is no longer representable. 
Since the group is possible only on this condition, both the 
institution of the emblem and the part it takes in the life of the 
group are thus explained. 

It remains to ask why these names and emblems were taken 
almost exclusively from the animal and vegetable kingdoms, 
but especially from the former. 

It seems probable to us that the emblem has played a more 
important part than the name. In any case, the written sign 
still holds a more central place in the life of the clan to-day 
than does the spoken sign. Now the basis of an emblematic 
image can be found only in something susceptible of being 
represented by a design. On the other hand, these things had 
to be those with which the men of the clan were the most im- 
mediately and habitually coming in contact. Animals fulfilled 
this condition to a pre-eminent degree. For these nations of 
hunters and fishers, the animal constituted an essential element 
of the economic environment. In this connection j)lants had 
only a secondary place, for they can hold only a secondary place 

* For the authority of the chiefs, see Spencer and Gillen, Nat. Tr., p. lo ; 
Nor. Tr., p. 25 ; Howitt, Nat. Tr., pp. 295 ff. 

* At least in Australia. In America, the population is more generally 
sedentary ; but the American clan represents a relatively advanced form of 

* To make sure of this, it is sufficient to look at the chart arranged by Thomas. 
Kinship and Marriage in Australia, p. 40. To appreciate this chart properly, it 
should be remembered that the author has extended, for a reason unknown to 
us, the system of totemic filiation in the paternal line clear to the western coast 
of Australia, though we have almost no information about the tribes of this 
region, which is, moreover, largely a desert. 

234 Elementary Forms of Religious Life 

as food as long as they are not cultivated. Moreover, the animal 
is more closely associated with the life of men than the plant is, 
if only because of the natural kinship uniting these two to each 
other. On the other hand, the sun, moon and stars are too far 
away, they give the effect of belonging to another world. * Also, 
as long as the constellations were not distinguished and classified, 
the starry vault did not offer a sufficient diversity of clearly 
differentiated things to be able to mark all the clans and sub- 
clans of a tribe ; but, on the contrary, the variety of the flora, 
and especially of the fauna, was almost inexhaustible. Therefore 
celestial bodies, in spite of their brilliancy and the sharp impression 
they make upon the senses, were unfitted for the rôle of totems» 
while animals and plants seemed predestined to it. 

Ah observation of Strehlow even allows us to state precisely 
the way in which these emblems were probably chosen. He says 
that he has noticed that the totemic centres are generally situated 
near a mountain, spring or gorge where the animals serving as 
totems to the group gather in abundance, and he cites a certain 
number of examples of this fact.^ Now these totemic centres 
are surely the consecrated places where the meetings of the clan 
are held. So it seems as though each group had taken as its 
insignia the animal or plant that was the commonest in the 
vicinity of the place where it had the habit of meeting. ^ 


This conception of totemism will give us the explanation 
of a very curious trait of human mentality which, even though 
more marked formerly than to-day, has not yet disappeared and 
which, in any case, has been of considerable importance in the 
history of human thought. It will furnish still another occasion 
for showing how logical evolution is closely connected with 
religious evolution and how it, like this latter, depends upon 
social conditions.* 

1 The stars are often regarded, even by the Australians, as the land of souls 
and mythical personages, as will be established in the next chapter : that means 
that they pass as being a very different world from that of the living. 

^ Op. cit., I, p. 4. Cf. Schulze, loc. cit., p. 243. 

' Of course it is to be understood that, as we have already pointed out (see 
above, p. 155), this choice was not made without a more or less formal agree- 
ment between the groups that each should take a different emblem from its 

* The mental state studied in this paragraph is identical to the one called by 
Lévy-Bruhl the law of participation {Les fonctions mentales dans les sociétés 
inférieures, pp. 76 ff.) . The following pages were written when this work appeared 
and we publish them without change ; we confine ourselves to adding certain 
explanations showing in what we differ from M. Lévy-Bruhl in our understanding 
of the facts. 

Origins of these Beliefs 235 

If there is one truth which appears to be absolutely certain 
to-day, it is that beings differing not only in their outward appear- 
ance but also in their most essential properties, such as minerals, 
plants, animals and men, cannot be considered equivalent and 
interchangeable. Long usage, which scientific culture has still 
more firmly embedded in our minds, has taught us to establish 
barriers between the kingdoms, whose existence transformism 
itself does not deny ; for though this admits that life may have 
arisen from non-living matter and men from animals, still, it 
does not fail to recognize the fact that living beings, once formed, 
are different from minerals, and men different from animals. 
Within each kingdom the same barriers separate the different 
classes : we cannot conceive of one mineral having the same 
distinctive characteristics as another, or of one animal species 
having those of another species. But these distinctions, which 
seem so natural to us, are in no way primitive. In the beginning, 
all the kingdoms are confounded with each other. Rocks have 
a sex ; they have the power of begetting ; the sun, moon and 
stars are men or women who feel and express human sentiments, 
while men, on the contrary, are thought of as animals or 
plants. This state of confusion is found at the basis of all 
mythologies. Hence comes the ambiguous character of the beings 
portrayed in the mythologies ; they can be classified in no 
definite group, for they participate at the same time in the 
most opposed groups. It is also readily admitted that they 
can go from one into another ; and for a long time men believed 
that they were able to explain the origin of things by these 

That the anthropomorphic instinct, with which the animists 
have endowed primitive men, cannot explain their mental 
condition is shown by the nature of the confusions of which they 
are guilty. In fact, these do not come from the fact that men 
have immoderately extended the human kingdom to the point 
of making all the others enter into it, but from the fact that they 
confound the most disparate kingdoms. They have not conceived 
the world in their own image any more than they have conceived 
themselves in the world's image : they have done both at the 
same time. Into the idea they have formed of things, they have 
undoubtedly made human elements enter ; but into the idea 
they have formed of themselves, they have made enter elements 
coming from things. 

Yet there is nothing in experience which could suggest these 
connections and confusions. As far as the observation of the 
senses is able to go, everything is different and disconnected. 
Nowhere do we really see beings mixing their natures and 

236 Elementary Forms of Religious Life 

metamorphosing themselves into each other. It is therefore 
necessary that some exceptionally powerful cause should have 
intervened to transfigure reality in such a way as to make it 
appear under an aspect that is not really its own. 

It is religion that was the agent of this transfiguration ; it is 
religious beliefs that have substituted for the world, as it is per- 
ceived by the senses, another different one. This is well shown 
by the case of totemism. The fundamental thing in this religion 
is that the men of the clan and the different beings whose form 
the totemic emblems reproduce pass as being made of the same 
essence. Now when this belief was once admitted, the bridge 
between the different kingdoms was already built. The man was 
represented as a sort of animal or plant ; the plants and animals 
were thought of as the relatives of men, or rather, all these 
beings, so different for the senses, were thought of as partici- 
pating in a single nature. So this remarkable aptitude for con- 
fusing things that seem to be obviously distinct comes from the 
fact that the first forces with which the human intellect has 
peopled the world were elaborated by religion. Since these 
were made up of elements taken from the different kingdoms, 
men conceived a principle common to the most heterogeneous 
things, which thus became endowed with a sole and single 

But we also know that these religious conceptions are the 
result of determined social causes. Since the clan cannot exist 
without a name and an emblem, and since this emblem is always 
before the eyes of men, it is upon this, and the objects whose 
image it is, that the sentiments which society arouses in its 
members are fixed. Men were thus compelled to represent 
the collective force, whose action they felt, in the form of the 
thing serving as flag to the group. Therefore, in the idea of this 
force were mixed up the most different kingdoms ; in one sense, 
it was essentially human, since it was made up of human ideas 
and sentiments ; but at the same time, it could not fail to appear 
as closely related to the animate or inanimate beings who gave 
it its outward form. Moreover, the cause whose action we observe 
here is not peculiar to totemism ; there is no society where it 
is not active. In a general way, a collective sentiment can become 
conscious of itself only by being fixed upon some material 
object ; ^ but by this very fact, it participates in the nature 
of this object, and reciprocally, the object participates in its 
nature. So it was social necessity which brought about the 
fusion of notions appearing distinct at first, and social life 
has facilitated this fusion by the great mental effervescences it 

* See above, p. 230. 

Origins of these Beliefs 237 

determines.^ This is one more proof that logical understanding 
is a function of society, for it takes the forms and attitudes 
that this latter presses upon it. 

It is true that this logic is disconcerting for us. Yet we must 
be careful not to depreciate it : howsoever crude it may appear 
to us, it has been an aid of the greatest importance in the in- 
tellectual evolution of humanity. In fact, it is through it that 
the first explanation of the world has been made possible. Of 
course the mental habits it implies prevented men from seeing 
reality as their senses show it to them ; but as they show it, 
it has the grave inconvenience of allowing of no explanation. 
For to explain is to attach things to each other and to establish 
relations between them which make them appear to us as func- 
tions of each other and as vibrating sympathetically according 
to an internal law founded in their nature. But sensations, 
which see nothing except from the outside, could never make 
them disclose these relations and internal bonds ; the intellect 
alone can create the notion of them. When I learn that A regu- 
larly precedes B, my knowledge is increased by a new fact ; 
but my intelligence is not at all satisfied with a statement which 
does not show its reason. I commence to understand only if 
it is possible for me to conceive B in a way that makes it appear 
to me as something that is not foreign to A, and as united to A 
by some relation of kinship. The great service that religions have 
rendered to thought is that they have constructed a first repre- 
sentation of what these relations of kinship between things 
may be. In the circumstances under which it was attempted, 
the enterprise could obviously attain only precarious results. 
But then, does it ever attain any that are definite, and is it not 
always necessary to reconsider them ? And also, it is less im- 
portant to succeed than to try. The essential thing was not to 
leave the mind enslaved to visible appearances, but to teach it 
to dominate them and to connect what the senses separated ; 
for from the moment when men have an idea that there are 
internal connections between things, science and philosophy 
become possible. Religion opened up the way for them. But if 
it has been able to play this part, it is only because it is a social 
affair. In order to make a law for the impressions of the senses 
and to substitute a new way of representing reality for them, 

^ Another cause has contributed much to this fusion ; this is the extreme 
contagiousness of religious forces. They seize upon every object within their 
reach, whatever it may be. Thus a single religious force may animate the most 
diverse things which, by that very fact, become closely connected and classified 
within a single group. We shall return again to this contagiousness, when we shall 
show that it comes from the social origins of the idea of sacredness (Bk. Ill, ch. i, 
in fine). 

238 Elementary Forms of Religious Life 

thought of a new sort had to be founded : this is collective 
thought. If this alone has had this efhcacy, it is because of the 
fact that to create a world of ideals through which the world of 
experienced reahties would appear transfigured, a super-excitation 
of the intellectual forces was necessary, which is possible only in 
and through society. 

So it is far from true that this mentality has no connection 
with ours. Our logic was born of this logic. The explanations 
of contemporary science are surer of being objective because 
they are more methodical and because they rest on more carefully 
controlled observations, but they do not differ in nature from 
those which satisfy primitive thought. To-day, as formerly, to 
explain is to show how one thing participates in one or several 
others. It has been said that the participations of this sort 
implied by the mythologies violate the principle of contradiction 
and that they are by that opposed to those implied by scientific 
explanations.^ Is not the statement that a man is a kangaroo 
or the sun a bird, equal to identifying the two with each other ? 
But our manner of thought is not different when we say of heat 
that it is a movement, or of light that it is a vibration of the ether, 
etc. Every time that we unite heterogeneous terms by an internal 
bond, we forcibly identify contraries. Of course the terms we 
unite are not those which the Australian brings together ; we 
choose them according to different criteria and for different 
reasons ; but the processes by which the mind puts them in 
connection do not differ essentially. 

It is true that if primitive thought had that sort of general 
and systematic indifference to contradictions which has been 
attributed to it,^ it would be in open contradiction on this 
point with modem thought, which is always careful to remain 
consistent with itself. But we do not believe that it is possible 
to characterize the mentality of inferior societies by a single 
and exclusive inclination for indistinction. If the primitive 
confounds things which we distinguish, he also distinguishes 
things which we connect together, and he even conceives these 
distinctions in the form of sharp and clear-cut oppositions. Be- 
tween two things which are classified in two different phratries, 
there is not only separation, but even antagonism.^ For this 
reason, the same Austrahan who confounds the sun and the 
white cockatoo, opposes this latter to the black cockatoo as to its 
contrary. The two seem to him to belong to two separate 
classes between which there is nothing in common. A still 
more marked opposition is that existing between sacred things 

1 Lévy-Bruhl, op. cit., pp. 77 ff. * Ibid., p. 79- 

* See above, p. 14O. 

Origins of these Beliefs 239 

and profane things. They repel and contradict each other with 
so much force that the mind refuses to think of them at the same 
time. They mutually expel each other from the consciousness. 

Thus between the logic of religious thought and that of scientific 
thought there is no abyss. The two are made up of the same 
elements, though inequally and differently developed. The 
special characteristic of the former seems to be its natural 
taste for immoderate confusions as well as sharp contrasts. It 
is voluntarily excessive in each direction. When it connects, 
it confounds ; when it distinguishes, it opposes. It knows no 
shades and measures, it seeks extremes ; it consequently employs 
logical mechanisms with a certain awkwardness, but it ignores 
none of them. 



IN the preceding chapters we have been studying the funda- 
mental principles of the totemic religion. We have seen 
that no idea of soul or spirit or mythical personality is to be 
found among these. Yet, even if the idea of spiritual beings is 
not at the foundation of totemism or, consequently, of religious 
thought in general, still, there is no religion where this notion 
is not met with. So it is important to see hov^ it is formed. To 
make sure that it is the product of a secondary formation, we 
must discover the way in which it is derived from the more 
essential conceptions which we have just described and explained. 
Among the various spiritual beings, there is one which should 
receive our attention first of all because it is the prototype 
after which the others have been constructed : this is the soul. 


Just as there is no known society without a religion, so there 
exist none, howsoever crudely organized they may be, where 
we do not find a whole system of collective representations 
concerning the soul, its origin and its destiny. So far as we 
are able to judge from the data of ethnology, the idea of the 
soul seems to have been contemporaneous with humanity 
itself, and it seems to have had all of its essential characteristics 
so well formulated at the very outset that the work of the more 
advanced religions and philosophy has been practically confined 
to refining it, while adding nothing that is really fundamental. 
In fact, all the Australian societies admit that every human 
body shelters an interior being, the principle of the life which 
animates it : this is the soul. It sometimes happens, it is true, 
that women form an exception to this general rule : there are 
tribes where they are believed to have no souls. ^ If Dawson 
is to be believed, it is the same with young children in the 

' This is the case with the Gnanji ; see Nor. Tr., pp. 170, 546 ; cf. a similar 
case in Brough Smyth, 11, p. 269. 


The Idea of the Soul 241 

tribes that he has observed.^ But these are exceptional and 
probably late cases ; ^ the last one even seems to be suspect 
and may well be due to an erroneous interpretation of the 
facts. 3 

It is not easy to determine the idea which the Australian makes 
of the soul, because it is so obscure and floating ; but we should 
not be surprised at this. If someone asked our own contem- 
poraries, or even those of them who believe most firmly in the 
existence of the soul, how they represented it, the replies that 
he would receive would not have much more coherence and 
precision. This is because we are dealing with a very complex 
notion, into which a multitude of badly analysed impressions 
enter, whose elaboration has been carried on for centuries, 
though men have had no clear consciousness of it. Yet from this 
come the most essential, though frequently contradictory, 
characteristics by which it is defined. 

In some cases they tell us that it has the external appearance 
of the body.'* But sometimes it is also represented as having 
the size of a grain of sand ; its dimensions are so reduced that it 
can pass through the smallest crevices or the finest tissues.^ We 
shall also see that it is represented in the appearance of animals. 
This shows that its form is essentially inconsistent and un- 
determined ; * it varies from one moment to another with the 
demands of circumstances or according to the exigencies of the 
myth and the rite. The substance out of which it is made is no 
less indefinable. It is not without matter, for it has a form, 
howsoever vague this may be. And in fact, even during this 
life, it has physical needs : it eats, and inversely, it may be 
eaten. Sometimes it leaves the body, and in the course of its 

^ Australian Aborigines, p. 51. 

* There certainly was a time when the Gnanji women had souls, for a large 
number of women's souls still exist to-day. However, they never reincarnate 
themselves ; since in this tribe the soul animating a new-born child is an old 
reincarnated soul, it follows from the fact that women's souls do not reincarnate 
themselves, that women cannot have a soul. Moreover, it is possible to explain 
whence this absence of reincarnation comes. Filiation among the Gnanji, rffter 
having been uterine, is now in the paternal line : a mother no longer transmits 
her totem to her child. So the woman no longer has any descendants to per- 
petuate her ; she is the finis families suce. To explain this situation, there are 
only two possible hypotheses ; either women have no souls, or else they are 
destroyed after death. The Gnanji have adopted the former of these two 
explanations ; certain peoples of Queensland have preferred the latter (see Roth, 
Super stiUon, Magic and Medicine, in .V. Queensland Ethnog., No. 5, § 68). 

' " The children below four or five years of age have neither soul nor future 
life," says Dawson. But the fact he thus relates is merely the absence of funeral 
rites for young children. We shall see the real meaning of this below. 

* Dawson, p. 51 ; Parker, The Euahlayi, p. 35 ; Eylmann, p. 188. 

' Nor. Tr., p. 542 ; Schurmann, The Abariginal tribes of Port Lincoln, in 
Woods, p. 235. 

* This is the expression used by Dawson, p. 50. 

242 Elementary Forms of Religious Life 

travels it occasionally nourishes itself on foreign souls. ^ After 
it has once been completely freed from the organism, it is thought 
to lead a life absolutely analogous to the one it led in this world ; 
it eats, drinks, hunts, etc.^ When it flutters among the branches 
of trees, it causes rustlings and crackings which even profane 
ears hear.^ But at the same time, it is believed to be invisible 
to the vulgar.* It is true that magicians or old men have the 
faculty of seeing souls ; but it is in virtue of special powers 
which they owe either to age or to a special training that they 
perceive things which escape our senses. According to Dawson, 
ordinary individuals enjoy the same privilege at only one moment 
of their existence : when they are on the eve of a premature 
death. Therefore this quasi-miraculous vision is considered a 
sinister omen. Now, invisibility is generally considered one of 
the signs of spirituality. So the soul is conceived as being 
immaterial to a certain degree, for it does not affect the senses 
in the way bodies do : it has no bones, as the tribes of the TuUy 
River say.^ In order to conciliate all these opposed characteristics, 
they represent it as made of some infinitely rare and subtle 
matter, like something ethereal,' and comparable to a shadow 
or breath. ' 

It is distinct and independent of the body, for during this 
life it can leave it at any moment. It .does leave it during sleep, 
fainting spells, etc.^ It may even remain absent for some time 
without entailing death ; however, during these absences life 
is weakened and even stops if the soul does not return home.^ 
But it is especially at death that this distinction and independence 
manifest themselves with the greatest clarity. While the body 
no longer exists and no visible traces of it remain, the soul 
continues to live : it leads an autonomous existence in another 

But howsoever real this duality may be, it is in no way absolute. 
It would show a grave misunderstanding to represent the body 

^ Strehlow, I, p. 15, n. i ; Schulze, loc. cit., p. 246 ; this is the theme of the 
myth of the vampire. 

^ Strehlow, I, p. 15 ; Schulze, p. 244 ; Dawson, p. 51. It is true that it is 
sometimes said that souls have nothing corporeal ; according to certain testi- 
mony collected by Eylmann (p. 188), they are ohne Fleisch und Blut. But these 
radical negations leave us sceptical. The fact that ofierings are not made to the 
souls of the dead in no way implies, as Roth thinks {Superstition, Magic, etc., 
§ 65), that they do not eat. 

^ Roth, ib%d.. § 65 ; Nor. Tr., p. 530. It sometimes happens that the soul 
emits odours (Roth, tbid., §68). 

* Roth, ibid., ^ 67 ; Dawson, p. 51. ' Roth, ibid., ^ 65. 

* Schiirmann, Aborig. Tr. of Port Lincoln, in Woods, p. 235. 

' Parker, The Euahlayi, pp. 29, 35 ; Roth, ibid., §§ 65, 67, 68. 

• Roth, ibid., § 65 ; Strehlow, I, p. 15. 

• Strelxlow, I, p. 14, n. i. 

The Idea of the Soul 243 

as a sort of habitat in which the soul resides, but with which it 
has only external relations. Quite on the contrary, it is united 
to it by the closest bonds ; it is separable from it only imperfectly 
and with difficulty. We have already seen that it has, or at least 
is able to have, its external aspect. Consequently, everything 
that hurts the one hurts the other ; every wound of the body 
spreads to the soul.^ It is so intimately associated with the life 
of the organism that it grows with it and decays with it. This 
is why a man who has attained a certain age enjoys privileges 
refused to young men ; it is because the religious principle 
within him has acquired greater force and efficacy as he has 
advanced in life. But when senility sets in, and the old man 
is no longer able to take a useful part in the great religious 
ceremonies in which the vital interests of the tribe are concerned, 
this respect is no longer accorded to him. It is thought that 
weakness of the body is communicated to the soul. Having 
the same powers no longer, he no longer has a right to the same 

There is not only a close union of soul and body, but there is 
also a partial confusion of the two. Just as there is something 
of the body in the soul, since it sometimes reproduces its form, 
so there is something of the soul in the body. Certain regions 
and certain products of the organism are believed to have a 
special affinity with it : such is the case with the heart, the 
breath, the placenta, ^ the blood,* the shadow,^ the liver, the fat 
of the liver, the kidneys,* etc. These various material substrata 
are not mere habitations of the soul ; they are the soul itself 
seen from without. When blood flows, the soul escapes with 
it. The soul is not in the breath ; it is the breath. It and the 
part of the body where it resides are only one. Hence comes 
the conception according to which a man has a number of souls. 
Being dispersed in various parts of the organism, the soul is 
differentiated and broken up into fragments. Each organ has 
individualized, as it were, the portion of the soul which it con- 
tains, and which has thus become a distinct entity. The soul of 

* Frazer, On Certain Burial Customs, as Illustrative of the Primitive Theory of 
the Soul, in J. A. I., XV, p. 66. 

* This is the case with the Kaitish and the Unmatjer? ; see Nor. Tr., p. 506 ; 
and Nat. Tr., p. 512, 

3 Roth, ibid., §§ 65, 66, 67, 68. 

* Roth, ibid., § 68 ; this says that when someone faints after a loss of blood, 
it is because the soul is gone. Cf. Parker, The Euahlayi, p. 38. 

* Parker, The Euahlayi, pp. 29, 35 ; Roth, ibid., § 65. 

* Strehlow, I, pp. 12, 14. In these passages he speaks of evil spirits which kill 
little children and eat their souls, livers and fat, or else their souhj, livers and 
kidneys. The fact that the soul is thus put on the same plane as the different 
viscera and tissues and is made a food like them shows the close connection it 
has with them. Cf. Schulze, p. 245. 

244 Elementary Forms of Religious Life 

the heart could not be that of the breath or the shadow or the 
placenta. While they are all related, still they are to be dis- 
tinguished, and even have different names. ^ 

Moreover, even if the soul is localized especially in certain 
parts of the organism, it is not absent from the others. In varying 
degrees, it is diffused through the whole body, as is well shown 
by the funeral rites. After the last breath has been expired and 
the soul is believed to be gone, it seems as though it should profit 
by the liberty thus regained, to move about at will and to return 
as quickly as possible to its real home, which is elsewhere. Never- 
theless, it remains near to the corpse ; the bond uniting them 
has been loosened, but not broken. A whole series of special 
rites are necessary to induce it to depart definitely. It is invited 
to go by gestures and significant movements. ^ The way is laid 
open for it, and outlets are arranged so that it can go more easily.^ 
This is because it has not left the body entirely ; it was too 
closely united to it to break away all at once. Hence comes 
the very frequent rite of funeral anthropophagy ; the flesh of 
the dead is eaten because it is thought to contain a sacred principle, 
which is really nothing more than the soul.* In order to drive it 
out definitely, the flesh is melted, either by submitting it to the 
heat of the sun,^ or to that of an artificial fire.® The soul departs 
with the liquids which result. But even the dry bones still retain 
some part of it. Therefore they can be used as sacred objects 
or instruments of magic ; ' or if someone wishes to give complete 
liberty to the principle which they contain, he breaks these.® 

But a moment does arrive when the final separation is accom- 
plished ; the liberated soul takes flight. But by nature it is so 
intimately associated with the body that this removal cannot take 
place without a profound change in its condition. So it takes 
a new name also.^ Although keeping all the distinctive traits 
of the individual whom it animated, his humours and his good 

^ For example, among the peoples on the Pennefather River (Roth, ibid., 
§ 68), there is a name for the soul residing in the heart {Ngai), another for the one 
in the placenta {Cho-i), and a third for the one which is confounded with the 
breath (Wanji). Among the Euahlayi, there are three or even four souls (Parker, 
The Euahlayi, p. 35). 

* See the description of the Urpmilchima rite among the Arunta {Nat. Tr., 
pp. 503 ff.). 

' Spencer and Gillen, Nat. Tr., pp. 497 and 508. 

* Nor. Tr.. pp. 547, 548. 

* Ibid., pp. 506, 527 ff. 

* Meyer, The Encounter Bay Tribe, in Woods, p. 198. 

' Nor. Tr., pp. 551, 463 ; Nat. Tr., p. 553. * Nor. Tr., p. 540. 

* Among the Arunta and Loritja, for example (Strehlow, I, p. 15, n. 2 ; II, 
p. 77). During life, the soul is called gumna, and Itana after death. The Itana of 
Strehlow is identical with the ulthana of Spencer and Gillen [Nat. Tr., pp. 514 ff.). 
The same is true of the tribes on the Bloomfield River (Roth, Superstition, etc., 

The Idea of the Soul 245 

and bad qualities,^ still it has become a new being. From that 
moment a new existence commences for it. 

It goes to the land of souls. This land is conceived differently 
by different tribes ; sometimes different conceptions are found 
existing side by side in the same society. For some, it is situated 
under the earth, where each totemic group has its part. This is 
at the spot where the first ancestors, the founders of the clan, 
entered the ground at a certain time, and where they live since 
their death. In the subterranean world there is a geographical 
disposition of the dead corresponding to that of the living. 
There, the sun always shines and rivers flow which never run 
dry. Such is the conception which Spencer and Gillen attribute 
to the central tribes, Arunta,^ Warramunga,^ etc. It is found 
again among the Wotjobaluk.* In other places, all the dead, 
no matter what their totems may have been, are believed to live 
together in the same place, which is more or less vaguely localized 
as beyond the sea, in an island,^ or on the shores of a lake.* 
Sometimes, finally, it is into the sky, beyond the clouds, that the 
souls are thought to go. " There," says Dawson, " there is a 
delectable land, abounding in kangaroos and game of every sort, 
where men lead a happy life. Souls meet again there and recog- 
nize one another." ' It is probable that certain of the features 
of this picture have been taken from the paradise of the Christian 
missionaries ; ® but the idea that souls, or at least some souls, 
enter the skies after death appears to be quite indigenous ; for 
it is found again in other parts of the continent.^ 

In general, all the souls meet the same fate and lead the same 
life. However, a different treatment is sometimes accorded 
them based on the way they have conducted themselves upon 
earth, and we can see the first outlines of these two distinct 
and even opposed compartments into which the world to come 
will later be divided. The souls of those who have excelled, 
during life, as hunters, warriors, dancers, etc., are not confounded 

* Eylmann, p. i88. « Nat. Tr., pp. 524, 491, 496. 
3 Nor. Tr., pp. 542, 504. 

* Mathews, Ethnol. Notes on the Aboriginal Tribes of N.S. Wales and Victoria, 
in Journal and Proc. of the Roy. Soc. of N.S. Wales, XXXVIII, p. 287. 

* Strehlow, I, pp. 15 fiF. Thus, according to Strehlow, the dead live in an 
island in the Arunta theory, but according to Spencer and Gillen, in a subterranean 
place. It is probable that the two myths coexist and are not the only ones. We 
shall see that even a third has been found. On this conception of an island of the 
dead, see Howitt, Nat. Tr., p. 498 ; Schiirmann, Aborig. Tr. of Port Lincoln, in 
Woods, p. 235 ; Eylmann, p. 189. 

« Schulze, p. 244. '' Dawson, p. 51. 

' In these same tribes evident traces of a more ancient myth will be found, 
according to which the dead live in a subterranean place (Dawson, ibid.). 

» Taplin, The Narrinyeri, pp. 18 f. ; Howitt, Nat. Tr., p. 473 ; Strehlow, I, 
p. 16. 

246 Elementary Forms of Religious Life 

with the common horde of tne others ; a special place is granted 
to them.i Sometimes, this is the sky.^ Strehlow even says that 
according to one myth, the souls of the wicked are devoured by 
dreadful spirits, and destroyed. ^ Nevertheless, these conceptions 
always remain very vague in Australia ; * they begin to have 
a clarity and determination only in the more advanced societies, 
such as those of America. ^ 


Such are the beliefs relative to the soul and its destiny, in 
their most primitive form, and reduced to their most essential 
traits. We must now attempt to explain them. What is it that 
has been able to lead men into thinking that there are two beings 
in them, one of which possesses these very special characteristics 
which we have just enumerated ? To find the reply to this 
question, let us begin by seeking the origin attributed to this 
spiritual principle by the primitive himself : if it is well analysed, 
his own conception will put us on the way towards the solution. 

Following out the method which we have set before ourselves, 
we shall study these ideas in a determined group of societies 
where they have been observed with an especial precision ; these 
are the tribes of Central Australia. Though not narrow, the area 
of our observations will be limited. But there is good reason 
for believing that these same ideas are quite generally held, in 
various forms, even outside of Australia. It is also to be noted 
that the idea of the soul, as it is found among these central 
tribes, does not differ specifically from the one found in other 
tribes ; it has the same essential characteristics ever3rwhere. 
As one effect always has the same cause, we may well think 
that this idea, which is everywhere the same, does not result 
from one cause here and another there. So the origin which 
we shall be led to attribute to it as a result of our study of these 
particular tribes with which we are going to deal, ought to be 
equally true for the others. These tribes will give us a chance 
to make an experiment, as it were, whose results, like those of 
every well-made experiment, are susceptible of generalization. 
The homogeneity of the Australian civilization would of itself 

1 Howitt, Nat. Tr., p. 498. 

* Strehlow, I, p. 16 ; Eylmann, p. 189 ; Howitt, Nat. Tr., p. 473. 

» These a; the spirits of the ancestors of a special clan, the clan of a certain 
poisonous gland {Giftdrùsenmànner). 

* Sometimes the work of the missionaries is evident. Dawson speaks of a real 
hell opposed to paradise ; but he too tends to regard this as a European importa- 

* Dorsey, Xlth Rep., pp. 419-420, 422, 485. Cf. Marillier, La survivance de 
l'âme et l'idée de justice chez les peuples non-civilisés, Rapport de l'Ecole des Hautes 
Études, 1893. 

The Idea of the Soul 247 

be enough to justify this generalization ; but we shall be careful 
to verify it afterwards with facts taken from other peoples, both 
in Australia and America. 

As the conceptions which are going to furnish us with the 
basis of our demonstration have been reported in different 
terms by Spencer and Gillen on the one hand and Strehlow on the 
other, we must give these two versions one after the other. 
We shall see that when, they are well understood, they differ in 
form more than in matter, and that they both have the same 
sociological significance. 

According to Spencer and Gillen, the souls which, in each 
generation, come to animate the bodies of newly-born children, 
are not special and original creations ; all these tribes hold that 
there is a definite stock of souls, whose number cannot be aug- 
mented at all,^ and which reincarnate themselves periodically. 
When an individual dies, his soul quits the body in which it dwelt, 
and after the mourning is accomplished, it goes to the land of the 
souls ; but after a certain length of time, it returns to incarnate 
itself again, and these reincarnations are the cause of conception 
and birth. At the beginning of things, it was these fundamental 
souls which animated the first ancestors, the founders of the clan. 
At an epoch, beyond which the imagination does not go and which 
is considered the very beginning of time, there were certain 
beings who were not derived from any others. For this reason, 
the Arunta call them the Altjirangamitjina,^ the uncreated ones, 
those who exist from all eternity, and, according to Spencer 
and Gillen, they give the name Alcheringa ^ to the period when 
these fabulous beings are thought to have lived. Being organized 
in totemic clans just as the men of to-day are, they passed their 
time in travels, in the course of which they accomplished all sorts 
of prodigious actions, the memory of which is preserved in the 
myths. But a moment arrived when this terrestrial life came to 
a close ; singly or in groups, they entered into the earth. But 
their souls live for ever ; they are immortal. They even continue 
to frequent the places where the existence of their former hosts 
came to an end. Moreover, owing to the memories attached to 
them, these places have a sacred character ; it is here that the 
oknanikilla are located, the sorts of sanctuaries where the churinga 
of the clan is kept, and the centres of the different totemic cults. 
When one of the souls which wander about these sanctuaries 
enters into the body of a woman, the result is a conception and 

* They may be doubled temporarily, as we shall see in the next chapter : but 
these duplications add nothing to the number of the souls capable of reincarna- 

* Strehlow, I, p. 2. » Nat. Tr., p. 73, n. i 

248 Elementary Forms of Religious Life 

later a birth. ^ So each individual is considered as a new appear- 
ance of a determined ancestor : it is this ancestor himself, come 
back in a new body and with new features. Now, what were 
these ancestors ? 

In the first place, they were endowed with powers infinitely 
superior to those possessed by men to-day, even the most re- 
spected old men and the most celebrated magicians. They are 
attributed virtues which we may speak of as miraculous : " They 
could travel on, or above, or beneath the ground ; by opening a 
vein in the arm, each of them could flood whole tracts of country 
or cause level plains to arise ; in rocky ranges they could make 
pools of water spring into existence, or could make deep gorges 
and gaps through which to traverse the ranges, and where they 
planted their sacred poles (nurtunja), there rocks or trees arose 
to mark the spot."- It is they who gave the earth the form it 
has at present. They created all sorts of beings, both men and 
animals. They are nearly gods. So their souls also have a 
divine character. And since the souls of men are these ancestral 
souls reincarnated in the human body, these are sacred beings too. 

In the second place, these ancestors were not men in the proper 
sense of the word, but animals or vegetables, or perhaps mixed 
beings in which the animal or vegetable element predominated : 
" In the Alcheringa," say Spencer and Gillen, " lived ancestors 
who, in the native mind, are so intimately associated with the 
animals or plants the name of which they bear that an Alcheringa 
man of, say, the kangaroo totem may sometimes be spoken of 
either as a man-kangaroo or a kangaroo-man. The identity of 
the human individual is often sunk in that of the animal or plant 
from which he is supposed to have originated."^ Their immortal 
souls necessarily have the same nature ; in them, also, the 
human element is wedded to the animal element, with a certain 
tendency for the latter to predominate over the former. So they 
are made of the same substance as the totemic principle, for 
we know that the special characteristic of this is to present this 
double nature, and to synthesize and confound the two realms 
in itself. 

Since no other souls than these exist, we reach the conclusion 
that, in a general way, the soul is nothing other than the totemic 
principle incarnate in each individual. And there is nothing to 

* On this set of conceptions, see Nat. Tr.. pp. 119, 123-127, 387 ff. ; Nor. 
Tr., pp. 145-174. Among the Gnanji, it is not necessarily near the oknanikilla 
that the conception takes place. But they believe that each couple is accom- 
panied in its wanderings over the continent by a swarm of souls of the husband's 
totem. When the time comes, one of these souls enters the body of the wife and 
fertilizes it, wherever she may be (Nor. Tr., p. 169). 

* Nat. Tr., pp. 512 f. ; cf. ch. x and xi. ' Nat. Tr., p. 119- 

The Idea of the Soul 249 

surprise us in this derivation. We already know that this 
principle is immanent in each of the members of the clan. But in 
penetrating into these individuals, it must inevitably individualize 
itself. Because the consciousnesses, of which it becomes thus an 
integral part, differ from each other, it differentiates itself accord- 
ing to their image ; since each has its own physiognomy, it takes 
a distinct physiognomy in each. Of course it remains something 
outside of and foreign to the man, but the portion of it which 
each is believed to possess cannot fail to contract close affinities 
with the particular subject in which it resides ; it becomes his 
to a certain extent. Thus it has two contradictory character- 
istics, but whose coexistence is one of the distinctive features of 
the notion of the soul. To-day, as formerly, the soul is what is 
best and most profound in ourselves, and the pre-eminent part of 
our being ; yet it is also a passing guest which comes from the 
outside, which leads in us an existence distinct from that of the 
body, and which should one day regain its entire independence. 
In a word, just as society exists only in and through individuals, 
the totemic principle exists only in and through the individual 
consciousnesses whose association forms the clan. If they did 
not feel it in them it would not exist ; it is they who put it into 
things. So it must of necessity be divided and distributed among 
them. Each of these fragments is a soul. 

A myth which is found in a rather large number of the societies 
of the centre, and which, moreover, is only a particular form of 
the preceding ones, shows even better that this is really the 
matter out of which the idea of the soul is made. In these tribes, 
tradition puts the origin of each clan, not in a number of ancestors, 
but in only two,^ or even in one.^ This unique being, as long as 
he remained single, contained the totemic principle within him 
integrally, for at this moment there was nothing to which this 
principle could be communicated. Now, according to this same 
tradition, all the human souls which exist, both those which 
now animate the bodies of men and those which are at present 
unemployed, being held in reserve for the future, have issued 
from this unique personage ; they are made of his substance. 
While travelling over the surface of the ground, or moving about, 
or shaking himself, he made them leave his body and planted 
them in the various places he is believed to have passed over. 
Is this not merely a symbolic way of saying that they are parts 
of the totemic divinity ? 

^ Among the Kaitish (Nor. Tr., p. 154) and the Urabunna (Nor. Tr., p. 146). 

* This is the case among the Warramunga and the related tribes, the Walpari, 
Wulmala, Worgaia, Tjingilli (Nor. Tr., p. 161), and also the Umbaia and the 
Gnanji (ibid., p. 170). 

250 Elementary Forms of Religious Life 

. But this conclusion presupposes that the tribes of which we 
have just been speaking admit the doctrine of reincarnation. 
Now according to Strehlow, this doctrine is unknown to the 
Arunta, the society which Spencer and Gillen have studied the 
longest and the best. If, in this particular case, these two 
observers have misunderstood things to such an extent, their 
whole testimony would become suspect. So it is important to 
determine the actual extent of this divergence. 

According to Strehlow, after the soul has once been definitely 
freed from the body by the rites of mourning, it never reincar:;;^ 
nates itself again. It goes off to the isles of the dead, where it 
passes its days in sleeping and its nights in dancing, until it 
returns again to earth. Then it comes back into the midst of 
the living and plays the rôle of protecting genius to the young 
sons, or if such are lacking, to the grandsons whom the dead 
man left behind him ; it enters their body and aids their growth. 
It remains thus in the midst of its former family for a year or 
two, after which it goes back to the land of the souls. But after 
a certain length of time it goes away once more to make another 
sojourn upon earth, which is to be the last. A time will como 
when it must take up again, and with no hope of return this time, 
the route to the isles of the dead ; then, after various incidents, 
the details of which it is useless to relate, a storm will overtake it, 
in the course of which it will be struck by a flash of lightning. 
Thus its career is definitely terminated.^ 

So it cannot reincarnate itself ; nor can conceptions and births 
be due to the reincarnation of souls which periodically commence 
new existences in new bodies. It is true that Strehlow, as Spencer 
and Gillen, declares that for the Arunta commerce of the sexes 
is in no way the determining condition of generation, ^ which is 
considered the result of mystic operations, but different from the 
ones which the other observers told us about. It takes place in 
one or the other of the two following ways : 

Wherever an ancestor of the Alcheringa^ times is believed to 
have entered into the ground, there is either a stone or a tree 
representing his body. The tree or rock which has this mystic 
relation with the departed hero is called nanja according to 

^ strehlow, I. pp. 15-16. For the Loritja, see Strehlow, p. 7. 

^ Strehlow even goes so far as to say that sexual relations are not even 
thought to be a necessary condition or sort of preparation for conception (II, 
p. 52, n. 7). It is true that he adds a lew lines below that the old men know 
perfectly well the connection which unites sexual intercourse and generation, 
and that as far as animals are concerned, the children themselves know it. This 
lessens the value of his first assertion a little. 

* In general, we employ the terminology of Spencer and Gillen rather than 
that of Strehlow because it is now consecrated by long usage. 

The Idea of the Soul 251 

Spencer and Gillen,i or ngarra according to Strehlow.^ Some- 
times it is a water-hole which is beUeved to have been formed in 
this way. Now, on each of these trees or rocks and in each of 
these water-holes, there live embryo children, called ratapa,^ 
which belong to exactly the same totem as the corresponding 
ancestor. For example, on a gum-tree representing an ancestor 
of the kangaroo totem there are ratapa, all of which have the 
kangaroo as their totem. If a woman happens to pass it, and 
she is of the matrimonial class to which the mothers of these 
ratapa should belong,^ one of them may enter her through the 
hip. The woman learns of this act by the characteristic pains 
which are the first symptoms of pregnancy. The child thus 
conceived will of course belong to the same totem as the ancestor 
upon whose mystical body he resided before becoming incarnate. ^ 

In other cases, the process employed is slightly different : the 
ancestor himself acts in person. At a given moment he leaves 
his subterranean retreat and throws on to the passing woman a 
little churinga of a special form, called namatuna.^ The churinga 
enters the body of the woman and takes a human form there, 
while the ancestor disappears again into the earth.' 

These two ways of conception are believed to be equally 
frequent. The features of the child will reveal the manner in 
which he was conceived ; according to whether his face is broad 
or long, they say that he is the incarnation of a ratapa or a 
namatuna. Beside these two means of fecundation, Strehlow 
places a third, which, however, is much more rare. After his 
namatuna has penetrated into the body of the woman, the an- 
cestor himself enters her and voluntarily submits to a new birth. 
So in this case, the conception is due to a real reincarnation of 
the ancestor. But this is very exceptional, and when a man who 

* Nat. Tr., pp. 124, 513. 

* I, p. 5. Ngarra means eternal, according to Strehlow. Among the Loritja. 
only rocks fulfil this function. 

* Strehlow translates it by Kinderkeime (children-germs). It is not true that 
Spencer and Gillen have ignored the myth of the ratapa and the customs con- 
nected with it. They explicitly mention it in Nat. Tr., pp. 336 fif. and 552. They 
noticed, at different points of the Arunta territory, the existence of rocks called 
Erathipa from which the spirtt children, or the children's souls, disengage them- 
selves, to enter the bodies of women and fertilize them. According to Spencer 
and Gillen, Erathipa means child, though, as they add, it is rarely used m this 
sense in ordinary conversation {ibid., p. 338). 

* The Arunta are divided into four or eight matrimonial classes. The class 
of a child is determined by that of his father ; inversely, that of the latter may 
be deduced from the former (see Spencer and Gillen, Nat. Tr., pp. 70 fï. ; Strehlow. 
I, pp. 6 ff.). It remains to be seen how the ratapa has a matrimonial class ; we 
shall return to this point again. 

* Strehlow, II, p. 52. It happens sometimes, though rarely, that difeputes 
arise over the nature of the child's totem. Strehlow cites such a case (II. p. 53). 

* This is the same word as the namatwinna found in Spencer and Gillen [Nat. 
Tr., p. 541)- ' Strehlow, II, p. ^^. 

252 Elementary Forms of Religious Life 

has been conceived thus dies, the ancestral soul which animated 
him goes away, just like ordinary souls, to the isles of the dead 
where, after the usual delays, it is definitely annihilated. So it 
cannot undergo any further reincarnations. ^ 

Such is the version of Strehlow.^ In the opinion of this author 
it is radically opposed to that of Spencer and Gillen. But in 
reality it differs only in the letter of the formulae and symbols, 
while in both cases we find the same mythical theme in slightly 
different forms. 

In the first place, all the observers agree that every conception 
is the result of an incarnation. Only according to Strehlow, that 
which is incarnated is not a soul but a ratapa or a namatuna. 
But what is a ratapa ? Strehlow says that it is a complete 
embryo, made up of a soul and a body. But the soul is always 
represented in material forms ; it sleeps, dances, hunts, eats, etc. 
So it, too, has a corporal element. Inversely, the ratapa is in- 
visible to ordinary men ; no one sees it as it enters the body of 
the woman ; ^ this is equivalent to saying that it is made of a 
matter quite similar to that of the soul. So it hardly seems pos- 
sible to differentiate the two clearly in this regard. In reality, 
these are mythical beings which are obviously conceived after 
the same model. Schulze calls them the souls of children.^ 
Moreover, the ratapa, just like the soul, sustains the closest 
relations with the ancestor of which the sacred tree or rock is the 
materialized form. It is of the same totem as this ancestor, of 
the same phratry and of the same matrimonial class. ^ Its place 
in the social organization of the tribe is the very one that its 
ancestor is believed to have held before it. It bears the same 
name,^ which is a proof that these two personalities are at least 
very closely related to one another. 

But there is more than this ; this relationship even goes as 
far as a complete identification. In fact, it is on the mystic body 
of the ancestor that the ratapa is formed ; it comes from this ; 
it is like a detached portion of it. So it really is a part of the 

1 strehlow, II, p. 56. 

^ Mathews attributes a similar theory of conception to the Tjingilli {alias 
Chingalee) {Proc. Roy. Geogr. Trans, and Soc. Queensland, XXII (1907), pp. 75-76). 

' It sometimes happens that the ancestor who is believed to have thrown the 
namatuna shows himself to the woman in the form of an animal or a man ; this 
is one more proof of the affinity of the ancestral soul for a material form. 

* Schulze, loc. cit., p. 237. 

^ This results from the fact that the ratapa can incarnate itself only in the 
body of a woman belonging to the same matrimonial class as the mother of the 
mythical ancestor. So we cannot understand how Strehlow could say (I, p. 42, 
Anmerkung) that, except in one case, the myths do not attribute determined 
matrimonial classes to the Alcheringa ancestors. His own theory of conception 
proves the contrary (cf. II, pp. 53 fi.). 

« Strehlow. II, p. 58. 

The Idea of the Soul 253 

ancestor which penetrates into the womb of the mother and 
which becomes the child. Thus we get back to the conception 
of Spencer and Gillen : birth is due to the reincarnation of an 
ancestral personage. Of course it is not the entire person that 
is reincarnated, it is only an emanation from him. But this 
difference has only a secondary interest, for when a sacred being 
divides and duplicates itself, all of its essential characteristics 
are to be found again in each of the fragments into which it is 
broken up. So really the Alcheringa ancestor is entire in each 
part of himself which becomes a ratapa.^ 

The second mode of conception distinguished by Strehlow has 
the same significance. In fact, the churinga, and more especially 
the particular churinga that is called the namatuna, is considered 
a transformation of the ancestor ; according to Strehlow,^ it is 
his body, just as the nanja tree is. In other words, the person- 
ality of the ancestor, his churinga and his nanja tree, are sacred 
things, inspiring the same sentiments and to which the same 
religious value is attributed. So they transmute themselves into 
one another : in the spot where an ancestor lost his churinga, a 
sacred tree or rock has come out of the soil, just the same as in 
those places where he entered the ground himself.^ So there is 
a mythological equivalence of a person of the Alcheringa and his 
churinga ; consequently, when the former throws a namatuna 
into the body of a woman, it is as if he entered into it himself. 
In fact, we have seen that sometimes he does enter in person 
after the namatuna ; according to other stories he precedes it ; 
it might be said that he opens up the way for it.* The fact that 
these two themes exist side by side in the same myth completes 
the proof that one is only a doublet of the other. 

Moreover, in whatever way the conception may have taken 
place, there can be no doubt that each individual is united to 
some determined ancestor of the Alcheringa by especially close 

^ The difference between the two versions becomes still smaller and is reduced 
to almost nothing, if we observe that, when Spencer and Gillen tell us that the 
ancestral soul is incarnated in the woman, the expressions they use are not to be 
taken literally. It is not the whole soul which comes to fertilize the mother, but 
only an emanation from this soul. In fact, according to their own statement, a 
soul equal or even superior in power to the one that is incarnated continues to 
live in the nanja tree or rock (see Nat. Tr., p. 514) ; we shall have occasion to 
come back to this point again (cf. below, p. 275). 

- II, pp. 76, 81. According to Spencer and Gillen, the churinga is not the soul 
of the ancestor, but the object in which his soul resides. At bottom, these two 
mythological interpretations are identical, and it is easy to see how one has been 
able to pass into the other : the body is the place where the soul resides. 

* Strehlow, I, p. 4. 

* Strehlow, I, pp. 53 f. In these stories, the ancestor begins by introducing 
himself into the body of the woman and causing there the troubles characteristic 
of pregnancy. Then he goes out, and only then does he leave his namatuna. 

254 Elementary Forms of Religious Life 

bonds. In the first place, each man has his appointed ancestor ; 
two persons cannot have the same one simultaneously. In other 
words, a being of the Alcheringa never has more than one repre- 
sentative among the living. ^ More than that, the one is only an 
aspect of the other. In fact, as we already know, the churinga left 
by the ancestor expresses his personality ; if we adopt the inter- 
pretation of Strehlow, which, perhaps, is the more satisfactory, 
we shall say that it is his body. But this same churinga is related 
in the same way to the individual who is believed to have been 
conceived under the influence of this ancestor, and who is the 
fruit of his mystic works. When the young initiate is introduced 
into the sanctuary of the clan, he is shown the churinga of his 
ancestor, and someone says to him, " You are this body ; you 
are the same thing as this."^ So, in Strehlow's own expression, 
the churinga is " the body common to the individual and his 
ancestor."^ Now if they are to have the same body it is neces- 
sary that on one side at least their two personalities be con- 
founded. Strehlow recognizes this explicitly, moreover, when 
he says, " By the tjurunga (churinga) the individual is united to 
his personal ancestor."* 

So for Strehlow as well as for Spencer and Gillen, there is a 
mystic, religious principle in each new-born child, which emanates 
from an ancestor of the Alcheringa. It is this principle which 
forms the essence of each individual, therefore it is his soul, or in 
any case the soul is made of the same matter and the same sub- 
stance. Now it is only upon this one fundamental fact that we 
have relied in determining the nature and origin of the idea of 
the soul. The different metaphors by means of which it may 
have been expressed have only a secondary interest for us.^ 

Far from contradicting the data upon which our theory rests, 
the recent observations of Strehlow bring new proofs confirming 
it. Our reasoning consisted in inferring the totemic nature of 
the human soul from the totemic nature of the ancestral 

^ Strehlow, II, p. 76. 

* Ibid., p. St. This is the word for word translation of the terms employed, as 
Strehlow gives them : Dies du Kiirper bist ; dies du der nnmliche. In the myth, a 
civilizing hero, Mangarkunjerkunja, says as he presents to each man the churinga 
of his ancestor : " You are born of this churinga " {ibid., p. 76). 

* Strehlow, II, p. 76. * Strehlow, ibid. 

* At bottom, the only real difference between Strehlow and Spencer and 
Gillen is the following one. For these latter, the soul of the individual, after 
death, returns to the nanja tree, where it is again confounded with the ancestor's 
soul (Nat. Tr., p. 513) ; for Strehlow, it goes to the isle of the dead, where it is 
finally annihilated. In neither myth does it survive individually. We are not 
going to seek the cause of this divergence. It is possible that there has been an 
error of observation on the part of Spencer and Gillen, who do not speak of the 
isle of the dead. It is also possible that the myth is not the same among the 
eastern Arunta, whom Spencer and Gillen observed particularly, as in the other 
parts of the tribe. 

The Idea of the Soul 255 

soul, of which the former is an emanation and a sort of replica. 
Now, some of the new facts which we owe to Strehlow show 
this character of both even more categorically than those we 
had at our disposal before do. In the first place, Strehlow, like 
Spencer and Gillen, insists on " the intimate relations uniting 
each ancestor to an animal, to a plant, or to some other natural 
object." Some of these Altjirangamitjina (these are Spencer 
and Gillen's men of the Alcheringa) " should," he says, " be 
manifested directly as animals ; others take the animal form in 
a way."^ Even now they are constantly transforming themselves 
into animals. 2 In any case, whatever external aspect they may 
have, " the special and distinctive qualities of the animal clearly 
appear in each of them." For example, the ancestors of the 
Kangaroo clan eat grass just like real kangaroos, and flee before 
the hunter ; those of the Emu clan run and feed like emus,^ etc. 
More than that, those ancestors who had a vegetable as totem 
become this vegetable itself on death. ^ Moreover, this close 
kinship of the ancestor and the totemic being is so keenly felt 
by the natives that it is shown even in their terminology. Among 
the Arunta, the child calls the totem of his mother, which serves 
him as a secondary totem, ^ altjira. As filiation was at first in the 
uterine line, there was once a time when each individual had no 
other totem than that of his mother ; so it is very probable that 
the term altjira then designated the real totem. Now this clearly 
ente^-s into the composition of the word which means great 
ancestor, altjirangatnitjina.^ 

The idea of the totem and that of the ancestor are even so 
closely kindred that they sometimes seem to be confounded. 
Thus, after speaking of the totem of the mother, or altjira, 
Strehlow goes on to say, " This altjira appears to the natives in 
dreams and gives them warnings, just as it takes information 
concerning them to their sleeping friends."' This altjira, which 
speaks and which is attached to each individual personally, is 
evidently an ancestor ; yet it is also an incarnation of the totem. 
A certain text in Roth, which speaks of invocations addressed to 
the totem, should certainly be interpreted in this sense. ^ So it 
appears that the totem is sometimes represented in the mind in 
the form of a group of ideal beings or mythical personages who 
are more or less indistinct from the ancestors. In a word, the 
ancestors are the fragments of the totem. ^ 

» strehlow, II, p. 51. « Ibid.. II, p. 56. » Ibid., I, pp. 3-4. 

♦ /iid., II, p. 61. ^ Seeabove, p. 1S3. « Strehlow, II, p. 57 ; I, p. 2. 
' Strehlow, II, p. 57. « Roth, Superstition, Magic, etc., § 74. 

* In other words, the totemic species is made up of the group of ancestors and 
the mythological species much more than of the regular animal or vegetable 

256 Elementary Forms of Religious Life 

But if the ancestor is so readily confused with the totemic 
being, the individual soul, which is so near the ancestral soul, 
cannot do otherwise. Moreover, this is what actually results 
from the close union of each man with his churinga. In fact, we 
know that the churinga represents the personality of the indi- 
vidual who is believed to have been born of it ;i but it also 
expresses the totemic animal. When the civilizing hero, Man- 
garkunjerkunja, presented each member of the Kangaroo clan 
with his personal totem, he spoke as follows : " Here is the body 
of a kangaroo. "2 Thus the churinga is at once the body of the 
ancestor, of the individual himself and of the totemic animal ; 
so, according to a strong and very just expression of Strehlow, 
these three beings form a " solid unity. "^ They are almost 
equivalent and interchangeable terms. This is as much as to 
say that they are thought of as different aspects of one and the 
same reality, which is also defined by the distinctive attributes 
of the totem. Their common essence is the totemic principle. 
The language itself expresses this identity. The word ratapa, 
and the aratapi of the Loritja language, designate the mythical 
embryo which is detached from the ancestor and which becomes 
the child ; now these same words also designate the totem of 
this same child, such as is determined by the spot where the 
mother believes that she conceived.* 


Up to the present we have studied the doctrine of reincarna- 
tion only in the tribes of Central Australia ; therefore the bases 
upon which our inference rests may be deemed too narrow. 
But in the first place, for the reasons which we have pointed out, 
the experiment holds good outside of the societies which we have 
observed directly. Also, there are abundant facts proving that 
the same or analogous conceptions are found in the most diverse 
parts of Australia or, at least, have left very evident traces there. 
They are found even in America. 

Howitt mentions them among the Dieri of So.ith Australia.^ 
The word Mura-mura, which Gason translates with Good Spirit 
and which he thinks expresses a belief in a god creator,^ is really 
a collective word designating the group of ancestors placed by 
the myth at the beginning of the tribe. They continue to exist 

1 See above, p. 254. " Strehlow, II, p. 76. * Strehlow, ibid. 

* Strehlow, II, pp. 57, 60, 61. Strehlow calls the list of totems the list of 

* Howitt, Nat. Tr., pp. 475 ff. 

* The Manners and Customs of the Dieyerie Tribe of Australian Aborigines, 
in Curr, II, p. 47. 

The Idea of the Soul 257 

to-day as formerly. " They are believed to live in trees, which 
are sacred for this reason." Certain irregularities of the ground, 
rocks and springs are identified with these Mura-mura/ which 
consequently resemble the Altjirangamitjina of the Arunta in a 
singular way. The Kurnai of Gippsland, though retaining only 
vestiges of totemism, also believe in the existence of ancestors 
called Muk-Kurnai, and which they think of as beings inter- 
mediate between men and animals. ^ Among the Nimbaldi, 
Taplin has observed a theory of conception similar to that which 
Strehlow attributes to the Arunta.^ We find this belief in re- 
incarnation held integrally by the Wotjobaluk in Victoria. 
" The spirits of the dead," says Mathews, " assemble in the 
miyur'^ of their respective clans ; they leave these to be born 
again in human form when a favourable occasion presents 
itself."^ Mathews even affirms that " the belief in the reincar- 
nation or transmigration of souls is strongly enrooted in all the 
Australian tribes."* 

If we pass to the northern regions we find the pure doctrine 
of the Arunta among the Niol-Niol in the north-west ; every 
birth is attributed to the incarnation of a pre-existing soul, 
which introduces itself into the body of a woman.' In northern 
Queensland myths, diftering from the preceding only in form, 
express exactly the same ideas. Among the tribes on the Penne- 
father River it is believed that every man has two souls : the 
one, called ngai, resides in the heart ; the other, called choi, 
remains in the placenta. Soon after birth the placenta is buried 
in a consecrated place. A particular genius, named Anje-a, who 
has charge of the phenomena of procreation, comes to get this 
choi and keeps it until the child, being grown up, is married 
When the time comes to give him a son, Anje-a takes a bit of 
the choi of this man, places it in the embryo he is making, and 
inserts it into the womb of the mother. So it is out of the soul 
of the father that that of the child is made. It is true that the 
child does not receive the paternal soul integrally at first, for 
the ngai remains in the heart of the father as long as he lives. 
But when he dies the ngai, being liberated, also incarnates itself 
in the bodies of the children ; if there are several children it is 
divided equally among them. Thus there is a perfect spiritual 

1 Howitt, Nat. Tr., p. .182. * Ibid., p. 487. 

^ Taplin, Folk-Lore, Customs, Manners, etc., of the South Australian Aborig., 
p. 88. 

* The clan of each ancestor has its special camp underground ; this camp is 
the miyur. 

* Mathews, in four, of Roy. Soc. of N.S. Wales, XXXVIII, p. 293. He points 
out the same belief among other tribes of Victoria (ibid., p. 197). 

* Mathews, ibui., p. 349. 

' J. Bishop, Die Niol-Niol, in Anthropos, III, p. 35. 

258 Elementary Forms of Religious Life 

continuity between the generations ; it is the same soul which is 
transmitted from a father to his children and from these to their 
children, and this unique soul, always remaining itself in spite 
of its successive divisions and subdivisions, is the one which 
animated the first ancestor at the beginning of all things. ^ 
Between this theory and the one held by the central tribes there 
is only one difference of any importance ; this is that the reincar- 
nation is not the work of the ancestors themselves but that of 
a special genius who takes charge of this function professionally. 
But it seems probable that this genius is the product of a syn- 
cretism which has fused the numerous figures of the first ancestors 
into one single being. This hypothesis is at least made probable 
by the fact that the words Anje-a and Anjir are evidently very 
closely related ; now the second designates the first man, the 
original ancestor from whom all men are descended. ^ 

These same ideas are found again among the Indian tribes 
of America. Krauss says that among the Tlinkit, the souls of 
the departed are believed to come back to earth and introduce 
themselves into the bodies of the pregnant women of their families. 
" So when a woman dreams, during pregnancy, of some deceased 
relative, she believes that the soul of this latter has penetrated 
into her. If the young child has some characteristic mark which 
the dead man had before, they believe that it is the dead man 
himself come back to earth, and his name is given to the child." ^ 
This belief is also general among the Haida. It is the shaman 
who reveals which relative it was who reincarnated himself in 
the child and what name should consequently be given to him.^ 
Among the Kwakiutl it is believed that the latest member of a 
family who died comes back to life in the person of the first child 
to be born in that family.* It is the same with the Hurons, the 
Iroquois, the Tinneh, and many other tribes of the United States.* 

The universality of these conceptions extends, of course, to 
the conclusion which we have deduced from them, that is, to 
the explanation of the idea of the soul which we have proposed. 
Its general acceptability is also proved by the following facts. 

* Roth, Superstition, etc., § 68 ; cf. § ôga, gives a similar case from among the 
natives on the Proserpine River. To simplify the description, we have left aside 
the complications due to differences of sex. The souls of daughters are made out 
of the choi of their mother, though these share with their brothers the ngai of 
their father. This peculiarity, coming perhaps from two systems of filiation 
which have been in use successively, has nothing to do with the principle of the 
perpetuity of the soul. 

* Ibid., p. 10. * Die Tlinkit- 1 ndianer, p. 2S2. 

* Swanton, Contributions to the Ethnology of the Haida, pp. 117 ff. 

' Boas, Sixth Rep. of the Comm. on the N.W. Tribes of Canada, p. 59. 
' Lafitau, Mœurs des sauvages Amériquains, II, p. 434 ; Petitot, Monographie 
des Dénè-Dindjié, p. 59. 

The Idea of the Soul 259 

We know * that each individual contains within him something 
of that anonymous force which is diffused in the sacred species ; 
he is a member of this species himself. But as an empirical 
and visible being, he is not, for, in spite of the symbolic designs 
and marks with which he decorates his body, there is nothing 
in him to suggest the form of an animal or plant. So it must be 
that there is another being in him, in whom he recognizes himself, 
but whom he represents in the form of an animal or vegetable 
species. Now is it not evident that this double can only be the 
soul, since the soul is, of itself, already a double of the subject 
whom it animates ? The justification of this identification is 
completed by the fact that the organs where the fragment of the 
totemic principle contained in each individual incarnates itself 
the most eminently are also those where the soul resides. This 
is the case with the blood. The blood contains something of 
the nature of the totem, as is proved by the part it takes in the 
totemic ceremonies.- But at the same time, the blood is one of 
the seats of the soul ; or rather, it is the soul itself, seen from 
without. When blood flows, life runs out and, in the same process, 
the soul escapes. So the soul is confused with the sacred principle 
which is imminent in the blood. 

Regarding matters from another point of view, if our explana- 
tion is well-founded, the totemic principle, in penetrating into 
the individual as we suppose, should retain a certain amount 
of autonomy there, since it is quite distinct from the subject in 
whom it is incarnated. Now this is just what Howitt claims to 
have observed among the Yuin : " That in this tribe the totem 
is thought to be in some way part of a man is clearly seen by the 
case of Umbara, before mentioned, who told me that, many 
years ago, someone of the Lace-lizard totem sent it while he 
was asleep, and that it went down his throat and almost ate 
his totem, which was in his breast, so that he nearly died." ^ 
So it is quite true that the totem is broken up in individualizing 
itself and that each of the bits thus detached plays the part of 
a spirit or soul residing in the body.^ 

But there are other more clearly demonstrative facts. If the 
soul is only the totemic principle individualized, it should have, 
in certain cases at least, rather close relations with the animal 
or vegetable species whose form is reproduced by the totem. 

* See above, pp. 134 ff. 

* See above-, p. 137. 

^ Howitt, Sat. Tr., p. 147 ; cf. ibid., p. 76g. 

* Strehlow (I, p. 15. n. 2) and Schulze {loc. cit., p. 246) speak of the soul, as 
Howitt here speaks of the totem, as leaving the body to go to eat another soul. 
Likewise, as we have seen above, the altjira or maternal totem shows itself in 
dreams, just as a soul or spirit does. 

200 Elementary Forms of Religious Life 

And, in fact, " the Geawe-Gal (a tribe of New South Wales) 
had a superstition that everyone had within himself an affinity 
to the spirit of some bird, beast or reptile. Not that he sprung 
from the creature in any way, but that the spirit which was in 
him was akin to that of the creature."^ 

There are even cases where the soul is believed to emanate 
directly from the animal or vegetable serving as totem. Among 
the Arunta, according to Strehlow, when a woman has eaten a 
great deal of fruit, it is believed that she will give birth to a child 
who will have this fruit as totem. If, at the moment when she 
felt the first tremblings of the child, she was looking at a 
kangaroo, it is believed that the ratapa of the kangaroo has 
entered her body and fertilized her.^ H. Basedow reported the 
same fact from the Wogait.^ We know, also, that the ratapa 
and the soul are almost indistinguishable things. Now, such 
an origin could never have been attributed to the soul if men did 
not think that it was made out of the same substances as the 
plants and animals of the totemic species. 

Thus the soul is frequently represented in an animal form. 
It is known that in inferior societies, death is never considered 
a natural event, due to the action of purely physical causes ; it 
is generally attributed to the evil workings of some sorcerer. 
In a large number of Australian societies, in order to determine 
who is the responsible author of this murder, they work on the 
principle that the soul of the murderer must inevitably come to 
visit its victim. Therefore, the body is placed upon a scaffolding ; 
then, the ground under the corpse and all around it is carefully 
smoothed off so that the slightest mark becomes easily perceptible. 
They return the next day ; if an animal has passed by there 
during the interval, its tracks are readily recognizable. Their 
form reveals the species to which it belongs, and from that, 
they infer the social group of which the guilty man is a member. 
They say that it is a man of such a class or such a clan,^ according 

1 Fison and Howitt, Kurnai and Kamilaroi, p. 280. 

" Globus, Vol. CXI, p. 289. In spite of the objections of Leonhardi, Strehlow 
maintains his attirmations on this point (see Strehlow, III, p. xi). Leonhardi 
finds a contradiction between this assertion and the theory according to which 
the ratapa emanate from trees, rocks or churinga. But the totemic animal 
incarnates the totem just as much as the nanja-tree or rock does, so they may 
fulfil the same function. The two things are mythological equivalents. 

3 Notes on the West Coastal Tribes of the Northern Territory of S. Australia, in 
Trans, of the Roy. Soc. of S. Aust., XXXI (1907), p. 4. Cf. Man, igog, No. 86. 

* Among the Wakelbura, where, according to Curr and Howitt, each matri- 
monial class has its own totems, the animal shows the class (see Curr, 111, p. 28) ; 
among the Buandik, it reveals the clan (Mrs. James S. Smith, The Bitandik Tribes 
of S. Australian Aborigines, p, 128). Cf. Howitt, On Some Australian Beliefs, in 
J.A.I., XIII, p. 191 ; XIV, p. 362 ; Thomas, An American View of Totemtsm, in 
Man, 1902, No. 85 ; Mathews, Journ. of the Roy. Soc. of N.S. Wales, XXXVIIl, 
pp. 347-348 ; Brough Smyth, I, p. no ; Nor. Tr., p. 513. 

The Idea of the Soul 261 

to whether the animal is the totem of this or that class or clan. 
So the soul is believed to have come in the form of the totemic 

In other societies where totemism has weakened or disappeared, 
the soul still continues to be thought of in an animal form. 
The natives of Cape Bedford (North Queensland) believe that 
the child, at the moment of entering the body of its mother, 
is a curlew if it is a girl, or a snake if it is a boy.^ It is only later 
that it takes a human form. Many of the Indians of North 
America, says the Prince of Wied, say that they have an animal 
in their bodies. 2 The Bororo of Brazil represent the soul in the 
form of a bird, and therefore believe that they are birds of the 
same variety. ^ In other places, it is thought of as a snake, a 
lizard, a fly, a bee, etc.* 

But it is especially after death that this animal nature of the 
soul is manifested. During Hfe, this characteristic is partially 
veiled, as it were, by the very form of the human body. But 
when death has once set it free, it becomes itself again. Among 
the Omaha, in at least two of the Buffalo clans, it is believed 
that the souls of the dead go to rejoin the buffalo, their ancestors.^ 
The Hopi are divided into a certain number of clans, whose 
ancestors were animals or beings with animal forms. Now 
Schoolcraft tells us that they say that at death, they take their 
original form again ; each becomes a bear or deer, according to 
the clan to which he belongs.^ Very frequently the soul is 
believed to reincarnate itself in the body of an animal.' It is 
probably from this that the widely-spread doctrine of metem- 
psychosis was derived. We have already seen how hard pressed 
Tylor is to account for it.^ If the soul is an essentially human 

1 Roth, Superstition, etc., § 83. This is probably a form of sexual totemism. 

* Prinz zu Wied, Reise in das innere Nord-Amerika, II, p. 190. 

* K. von den Steinen, Unter den Naturvolkern Zentral-Brasiliens, 1894, 
pp. 511, 512. 

' See Frazer, Golden Bough^, I, pp. 250, 253. 256, 257, 258. 

* Third Rep., pp. 229, 233. • Indian Tribes, IV, p. 86. 

' For example, among the Batta of Sumatra (see Golden Bough^, III, p. 420), 
in Melanesia (Codrington, The Melanesians, p. 178), in the Malay Archipelago 
(Tylor, Remarks on Totemism, in J. A. I., New Series, I, p. 147)- It is to be 
remarked that the cases where the soul clearly presents itself after death in an 
animal form all come from the societies where totemism is more or less perverted. 
This is because the idea of the soul is necessarily ambiguous wherever the totemic 
beliefs are relatively pure, for totemism implies that it participate in the two 
kingdoms at the same time. So it cannot become either one or the other ex- 
clusively, but takes one aspect or the other, according to the circumstances. As 
totemism develops, this ambiguity becomes less necessary, while at the same 
time, spirits more actively demand attention. Then the marked affinities of the 
soul for the animal kingdom are manifested, especially after it is freed from the 
human body. 

® See above, p. 170. On the generality of the doctrine of metempsychosis, 
see Tylor, II, pp. 8 tf. 

202 Elementary Forms of Religious Life 

principle, what could be more curious than this marked predi- 
lection which it shows, in so large a number of societies, for 
the animal form ? On the other hand, everything is explained 
if, by its very constitution, the soul is closely related to the 
animal, for in that case, when it returns to the animal world 
at the close of this life, it is only returning to its real nature. 
Thus the generality of the belief in metempsychosis is a new 
proof that the constituent elements of the idea of the soul have 
been taken largely from the animal kingdom, as is presupposed 
by the theory which we have just set forth. 


Thus the notion of the soul is a particular application of the 
beliefs relative to sacred beings. This is the explanation of the 
religious character which this idea has had from the moment 
when it first appeared in history, and which it still retains to-day. 
In fact, the soul has always been considered a sacred thing ; on 
this ground, it is opposed to the body which is, in itself, profane. 
It is not merely distinguished from its material envelope as the 
inside from the outside ; it is not merely represented as made 
out of a more subtle and fluid matter ; but more than this, it 
inspires those sentiments which are everywhere reserved for that 
which is divine. If it is not made into a god, it is at least regarded 
as a spark of the divinity. This essential characteristic would be 
inexplicable if the idea of the soul were only a pre-scientific 
solution given to the problem of dreams ; for there is nothing 
in the dream to awaken religious emotions, so the cause by which 
these are explained could not have such a character. But if 
the soul is a part of the divine substance, it represents something 
not ourselves that is within us ; if it is made of the same mental 
matter as the sacred beings, it is natural that it should become 
the object of the same sentiments. 

And the sacred character which men thus attribute to them- 
selves is not the product of a pure illusion either ; like the notions 
of religious force and of divinity, the notion of the soul is not 
without a foundation in reality. It is perfectly true that we are 
made up of two distinct parts, v/hich are opposed to one another 
as the sacred to the profane, and we may say that, in a certain 
sense, there is divinity in us. For society, this unique source of 
all that is sacred, does not limit itself to moving us from without 
and affecting us for the moment ; it establishes itself within us 
in a durable manner. It arouses within us a whole world of ideas 
and sentiments which express it but which, at the same time, 
form an integral and permanent part of ourselves. When the 
Australian goes away from a religious ceremony, the representations 

The Idea of the Soul 263 

which this communal Ufe has aroused or re-aroused within 
him are not obHterated in a second. The figures of the great 
ancestors, the heroic exploits whose memory these rites perpetuate, 
the great deeds of every sort in which he, too, has participated 
through the cult, in a word, all these numerous ideals which he 
has elaborated with the co-operation of his fellows, continue 
to live in his consciousness and, through the empotions which are 
attached to them and the ascendancy which they hold over his 
entire being, they are sharply distinguished from the vulgar 
impressions arising from his daily relations with external things. 
Moral ideas have the same character. It is society which forces 
them upon us, and as the respect inspired by it is naturally 
extended to all that comes from it, its imperative rules of conduct 
are invested, by reason of their origin, with an authority and a 
dignity which is shared by none of our internal states : therefore, 
we assign them a place apart in our psychical life. Although 
our moral conscience is a part of our consciousness, we do not 
feel ourselves on an equality with it. In this voice which makes 
itself heard only to give us orders and establish prohibitions, 
we cannot recognize our own voices ; the very tone in which it 
speaks to us warns us that it expresses something within us 
that is not of ourselves. This is the objective foundation of the 
idea of the soul : those representations whose flow constitutes 
our interior life are of two different species which are irreducible 
one into another. Some concern themselves with the external 
and material world ; others, with an ideal world to which we 
attribute a moral superiority over the first. So we are really 
made up of two beings facing in different and almost contrary 
directions, one of whom exercises a real pre-eminence over the 
other. Such is the profound meaning of the antithesis which 
all men have more or less clearly conceived between the body 
and the soul, the material and the spiritual beings who coexist 
within us. Moralists and preachers have often maintained 
that no one can deny the reality of duty and its sacred character 
without falling into materialism. And it is true that if we 
have no idea of moral and religious imperatives, our psychical 
life will all be reduced to one level, ^ all our states of consciousness 

^ Even if we believe that religious and moral representations constitute the 
essential elements of the idea of the soul, still we do not mean to say that they 
are the only ones. Around this central nucleus are grouped other states of 
consciousness having this same character, though to a slighter degree. This is 
the case with all the superior forms of the intellectual life, owing to the special 
price and dignity attributed to them by society. When we devote our lives to 
science or art, we feel that we are moving in a circle of things that are above 
bodily sensations, as we shall have occasion to show more precisely in our con- 
clusion. This is why the highest functions of the intelligence have always been 
considered specific manifestations of the soul. But they would probably not 
have been enough to establish the idea of it. 

264 Elementary Forms of Religions Life 

will be on the same plane, and all feeling of duality will perish. 
To make this duality intelligible, it is, of course, in no way 
necessary to imagine a mysterious and unrepresentable substance, 
under the name of the soul, which is opposed to the body. But 
here, as in regard to the idea of sacredness, the error concerns 
the letter of the symbol employed, not the reality of the fact 
symbolized. It remains true that our nature is double ; there 
really is a particle of divinity in us because there is within us 
us a particle of these great ideas which are the soul of the 

/ So the individual soul is only a portion of the collective soul 
of the group ; it is the anonymous force at the basis of the 
cult, but incarnated in an individual whose personality it 

(^espouses ; it is mana individualized. Perhaps dreams aided in 
determining certain secondary characteristics of the idea. The 
inconsistency and instability of the images which fill our minds 
during sleep, and their remarkable aptitude for transforming 
themselves into one another, may have furnished the model 
for this subtile, transparent and Protean matter out of which 
the soul is believed to be made. Also, the facts of swooning, 
catalepsy, etc., may have suggested the idea that the soul was 
mobile, and quitted the body temporarily during this life ; this, 
in its turn, has served to explain certain dreams. But all these 
experiences and observations could have had only a secondary 
and complimentary influence, whose very existence it is difficult 
to establish. All that is really essential in the idea comes from 

But does not this genesis of the idea of the soul misunder- 
stand its essential characteristic ? If the soul is a particular 
form of the impersonal principle which is diffused in the group, 
the totemic species and all the things of every sort which are 
attached to these, at bottom it is impersonal itself. So, with 
differences only of degree, it should have the same properties 
as the force of which it is a special form, and particularly, the 
same diffusion, the same aptitude for spreading itself contagiously 
and the same ubiquity. But quite on the contrary, the soul is 
voluntarily represented as a concrete, definite being, wholly 
contained within itself and not communicable to others ; it is 
made the basis of our personality. 

But this way of conceiving the soul is the product of a late and 
philosophic elaboration. The popular representation, as it is 
spontaneously formed from common experience, is very different, 
especially at first. For the Australian, the soul is a very vague 

The Idea of the Soul 265 

thing, undecided and wavering in form, and spread over the 
whole organism. Though it manifests itself especially at certain 
points, there are probably none from which it is totally absent. 
So it has a diffusion, a contagiousness and an omnipresence 
comparable to those of the mana. Like the mana, it is able to 
divide and duplicate itself infinitely, though remaining entire 
in each of its parts ; it is from these divisions and duplications 
that the plurality of souls is derived. On the other hand, the 
doctrine of reincarnation, whose generality we have established, 
shows how many impersonal elements enter into the idea of the 
soul and how essential those are. For if the same soul is going 
to clothe a new personality in each generation, the individual 
forms in which it successively develops itself must all be equally 
external to it, and have nothing to do with its true nature. It 
is a sort of generic substance which individualizes itself only 
secondarily and superficially. Moreover, this conception of the 
soul is by no means completely gone. The cult of relics shows 
that for a host of believers even to-day, the soul of a saint, 
with all its essential powers, continues to adhere to his different 
bones ; and this implies that he is believed to be able to diffuse 
himself, subdivide himself and incorporate himself in all sorts 
of different things simultaneously. 

Just as the characteristic attributes of the mana are found 
in the soul, so secondary and superficial changes are enough to 
enable the mana to individualize itself in the form of a soul. 
We pass from the first idea to the second with no break of con- 
tinuity. Every religious force which is attached in a special 
way to a determined being participates in the characteristics 
of this being, takes on its appearance and becomes its spiritual 
double. Tregear, in his Maori-Polynesian dictionary, has thought 
it possible to connect the word mana with another group of words, 
such as manawa, manamana, etc., which seem to belong to 
the same family, and which signify heart, life, consciousness.* 
Is this not equivalent to saying that some sort of kinship ought 
to exist between the corresponding ideas as well, that is to say, 
between the idea of impersonal force and those of internal life, 
mental force and, in a word, of the soul ? This is why the question 
whether the churinga is sacred because it serves as the residence 
of a soul, as Spencer and Gillen believe, or because it has imper- 
sonal virtues, as Strehlow thinks, seems to us to have little interest 
and to be without sociological importance. Whether the efficacy 
of a sacred object is represented in an abstract form in the mind 
or is attributed to some personal agent does not really matter. 
The psychological roots of both beliefs are identical : an object 

^ F. Tregear, The Maori-Polynesian Comparative Dictionary, pp. 203-205. 

206 Elementary Forms of Religious Life 

is sacred because it inspires, in one way or another, a collective 
sentiment of respect which removes it from profane touches. 
In order to explain this sentiment, men sometimes fall back on 
to a vague and imprecise cause, and sometimes on to a determined 
spiritual being endowed with a name and a history ; but these 
different interpretations are superadded to one fundamental 
phenomenon which is the same in both cases. 

This, moreover, is what explains the singular confusions, 
examples of which we have met with as we have progressed. 
The individual, the soul of the ancestor which he reincarnates 
or from which his own is an emanation, his churinga and the 
animals of the totemic species are, as we have said, partially 
equivalent and interchangeable things. This is because in certain 
connections, they all affect the collective consciousness in the 
same way. If the churinga is sacred, it is because of the collective 
sentiments of respect inspired by the totemic emblem carved 
upon its surface ; now the same sentiment attaches itself to the 
animals or plants whose outward form is reproduced by the 
totem, to the soul of the individual, for it is thought of in the 
form of the totemic being, and finally to the ancestral soul, of 
which the preceding one is only a particular aspect. So all these 
various objects, whether real or ideal, have one common element 
by which they arouse a single affective state in the mind, and 
through this, they become confused. In so far as they are 
expressed by one and the same representation, they are indistinct. 
This is how the Arunta has come to regard the churinga as the 
body common to the individual, the ancestor and even the 
totemic being. It is his way of expressing the identity of the 
sentiments of which these different things are the object. 

However, it does not follow from the fact the idea of the soul 
is derived from the idea of mana that the first has a relatively 
later origin, or that there was a period in history when men 
were acquainted with religious forces only in their impersonal 
forms. When some wish to designate by the word preanimist 
an historical period during which animism was completely un- 
known, they build up an arbitrary hypothesis ; ^ for there is no 
people among whom the ideas of the soul and of mana do not 
coexist side by side. So there is no ground for imagining that 
they were formed at two distinct times ; everything, on the 
contrary, goes to show that the two are coeval. Just as there is 
no society without individuals, so those impersonal forces which 
are disengaged from the group cannot establish themselves 

1 This is the thesis of Preuss in his articles in the Globus which we have cited 
several times. It seems that M. Lévy-Bruhl also tends towards this conception 
(see his Fonctions mentales, etc., pp. 92-93). 

The Idea of the Soul 267 

without incarnating themselves in the individual consciousnesses 
where they individuahze themselves. In reality, we do not 
have two different developments, but two different aspects of 
one and the same development. It is true that they do not 
have ati equal importance ; one is more essential than the other. 
The idea of mana does not presuppose the idea of the soul ; for 
if the mana is going to individualize itself and break itself up 
into the particular souls, it must first of all exist, and what it is 
in itself does not depend upon the forms it takes when indi- 
vidualized. But on the contrary, the idea of the soul cannot be 
understood except when taken in connection with the idea of 
mana. So on this ground, it is possible to say that it is the result 
of a secondary formation ; but we are speaking of a secondary 
formation in the logical, not the chronological, sense of the 

But how does it come that men have believed that the soul 
survives the body and is even able to do so for an indefinite 
length of time ? 

From the analysis which we have made, it is evident that 
the belief in immortality has not been established under the 
influence of moral ideas. Men have not imagined the prolongation 
of their existence beyond the tomb in order that a just retribution 
for moral acts may be assured in another life, if it fails in this 
one ; for we have seen that all considerations of this sort are 
foreign to the primitive conception of the beyond. 

Nor is the other hypothesis any better, according to which 
the other life was imagined as a means of escaping the agonizing 
prospect of annihilation. In the first place, it is not true that 
the need of personal survival was actively felt at the beginning. 
The primitive generally accepts the idea of death with a sort of 
indifference. Being trained to count his own individuality 
for little, and being accustomed to exposing his life constantly, 
he gives it up easily enough. 1 More than that, the immortality 
promised by the religions he practices is not personal. In a 
large number of cases, the soul does not continue the personality 
of the dead man, or does not continue it long, for, forgetful 
of its previous existence, it goes away, after a while, to animate 
another body and thus becomes the vivifying principle of a new 
personality. Even among the most advanced peoples, it was 
only a pale and sad existence that shades led in Sheol or Erebus, 
and could hardly attenuate the regrets occasioned by the 
memories of the life lost. 

^ On this point, see our Suicide, pp. 233 ff. 

208 Elementary Forms of Religious Life 

A more satisfactory explanation is the one attaching the 
conception of a posthumous hfe to the experiences of dreams. 
Our dead friends and relatives reappear to us in dreams : we see 
them act, we hear them speak ; it is natural to conclude that 
they continue to exist. But if these observations were able to 
confirm the idea after it had once been born, they hardly seem 
capable of creating it out of nothing. Dreams in which we see 
departed persons living again are too rare and too short and 
leave only too vague recollections of themselves, to have been 
able to suggest so important a system of beliefs to men all by 
themselves. There is a remarkable lack of proportion between 
the effect and the cause to which it is attributed. 

What makes this question embarrassing is the fact that in 
itself, the idea of the soul does not imply that of its survival, 
but rather seems to exclude it. In fact, we have seen that the 
soul, though being distinguished from the body, is believed, 
nevertheless, to be closely united to it : it ages along with 
the body, it feels a reaction from all the maladies that fall upon 
the body ; so it would seem natural that it should die with the 
body. At least, men ought to have believed that it ceased to 
exist from the moment when it definitely lost its original form, 
and when it was no longer what it had been. Yet it is at just 
this moment that a new life opens out before it. 

The myths which we have already described give the only 
possible explanation of this belief. We have seen that the 
souls of new-born children are either emanations of the ancestral 
souls, or these souls themselves reincarnated. But in order that 
they may either reincarnate themselves, or periodically give off 
new emanations, they must have survived their first holders. 
So it seems as though they admitted the survival of the dead in 
order to explain the birth of the living. The primitive does not 
have the idea of an all-powerful god who creates souls out of 
nothing. It seems to him that souls cannot be made except out 
of souls. So those who are born can only be new forms of those 
who have been ; consequently, it is necessary that these latter 
continue to exist in order that others may be born. In fine, 
the belief in the immortality of the soul is the only way in which 
men were able to explain a fact which could not fail to attract 
their attention ; this fact is the perpetuity of the life of the group. 
Individuals die, but the clan survives. So the forces which give 
it life must have the same perpetuity. Now these forces are the 
souls which animate individual bodies ; for it is in them and 
through them that the group is realized. For this reason, it is 
necessary that they endure. It is even necessary that in enduring, 
they remain always the same ; for, as the clan always keeps its 

The Idea of the Soul 269 

characteristic appearance, the spiritual substance out of which 
it is made must be thought of as quahtatively invariable. Since 
it is always the same clan with the same totemic principle, it is 
necessary that the souls be the same, for souls are only the 
totemic principle broken up and particularized. Thus there is 
something like a germinative plasm, of a mystic order, which is 
transmitted from generation to generation and which makes, 
or at least is believed to make, the spiritual unity of the clan 
through all time. And this belief, in spite of its symbolic character, 
is not without a certain objective truth. For though the group 
may not be immortal in the absolute sense of the word, still it is 
true that it endures longer than the individuals and that it is 
born and incarnated afresh in each new generation. 

A fact confirms this interpretation. We have seen that 
according to the testimony of Strehlow, the Arunta distinguish 
two sorts of souls : on the one hand are those of the ancestors 
of the Alcheringa, on the other, those of the individuals who 
actually compose the active body of the tribe at each moment 
in history. The second sort only survive the body for a relatively 
short time ; they are soon totally annihilated. Only the former 
are immortal ; as they are uncreated, so they do not perish. 
It is also to be noticed that they are the only ones whose im- 
mortality is necessary to explain the permanence of the group ; 
for it is upon them, and upon them alone, that it is incumbent to 
assure the perpetuity of the clan, for every conception is their work. 
In this connection, the others have no part to play. So souls are 
not said to be immortal except in so far as this immortality is 
useful in rendering intelligible the continuity of the collective life. 

Thus the causes leading to the first beliefs in a future life had 
no connections with the functions to be filled at a later period 
by the institutions beyond the tomb. But v/hen that had 
once appeared, they were soon utilized for other purposes besides 
those which had been their original reasons for existence. Even 
in the Australian societies, we see them beginning to organize 
themselves for this other purpose. Moreover, there was no need 
of any fundamental transformation for this. How true it is 
that the same social institution can successively fulfil different 
functions without changing its nature ! 


The idea of the soul was for a long time, and still is in part, 
the popular form of the idea of personahty.* So the genesis 

* It may be objected perhaps that unity is the characteristic of the person- 
ality, while the soul has always been conceived as multiple, and as capable of 
dividing and subdividing itself almost to mhnity. But we know to-day that the 

270 Elementary Forms of Religious Life 

of the former of these ideas should aid us in understanding how 
the second one was formed. 

From what has aheady been said, it is clear that the notion 
of person is the product of two sorts of factors. One of these is 
essentially impersonal : it is the spiritual principle serving as 
the soul of the group. In fact, it is this which constitutes the 
very substance of individual souls. Now this is not the possession 
of any one in particular : it is a part of the collective patrimony ; 
in it and through it, all consciousnesses communicate. But on 
the other hand, in order to have separate personalities, it is 
necessary that another factor intervene to break up and differ- 
entiate this principle : in other words, an individualizing factor 
is necessary. It is the body that fulfils this function. As bodies 
are distinct from each other, and as they occupy different points 
of space and time, each of them forms a special centre about 
which the collective representations reflect and colour themselves 
differently. The result is that even if all the consciousnesses in 
these bodies are directed towards the same world, to wit, the 
world of the ideas and sentiments which brings about the moral 
unity of thé group, they do not all see it from the same angle ; 
each one expresses it in its own fashion. 

Of these two equally indispensable factors, the former is 
certainly not the less important, for this is the one which furnishes 
the original matter for the idea of the soul. Perhaps some will be 
surprised to see so considerable a rôle attributed to the impersonal 
element in the genesis of the idea of personality. But the philo- 
sophical analysis of the idea of person, which has gone far ahead 
of the sociological analysis, has reached analogous results on 
this point. Among all the philosophers, Leibniz is one of those 
who have felt most vividly what a personality is ; for before 
all, the nomad is a personal and autonomous being. Yet, for 
Leibniz, the contents of all the monads is identical. In fact, 
all are consciousnesses which express one and the same object, 
the world ; and as the world itself is only a system of repre- 
sentations, each particular consciousness is really only the re- 
flection of the universal consciousness. However, each one 
expresses it from its own point of view, and in its own manner. 
We know how this difference of perspectives comes from the 

unity of the person is also made up of parts and that it, too, is capable of dividing 
and decomposing. Yet the notion of personality does not vanish because of the 
fact tliat we no longer tliink of it as a metaphysical and indivisible atom. It is 
the same with the popular conceptions of personality which fmd their expression 
in the idea of the soul. These show that men have always felt that the human 
personality does not have that absolute unity attributed to it by certain meta- 

The Idea of the Soul 271 

fact that the monads are situated differently in relation to each 
other and to the whole system which they constitute. 

Kant expresses the same sentiment, though in a different 
form. For him, the corner-stone of the personality is the will. 
Now the will is the faculty of acting in conformity with reason, 
and the reason is that which is most impersonal within us. For 
reason is not my reason ; it is human reason in general. It is 
the power which the mind has of rising above the particular, 
the contingent and the individual, to think in universal forms. 
So from this point of view, we may say that what makes a man 
a personality is that by which he is confounded with other 
men, that which makes him a man, not a certain man. The 
senses, the body and, in a word, all that individualizes, is, on 
the contrary, considered as the antagonist of the personality by 

This is because individuation is not the essential characteristic 
of the personality. A person is not merely a single subject 
distinguished from all the others. It is especially a being to 
which is attributed a relative autonomy in relation to the environ- 
ment with which it is most immediately in contact. It is repre- 
sented as capable of moving itself, to a certain degree : this is 
what Leibniz expressed in an exaggerated way when he said 
that the monad was completely closed to the outside. Now 
our analysis permits us to see how this conception was formed 
and to what it corresponds. 

In fact, the soul, a symbolic representation of the personality, 
has the same characteristic. Although closely bound to the 
body, it is believed to be profoundly distinct from it and to 
enjoy, in relation to it, a large degree of independence. During 
life, it may leave it temporarily, and it definitely withdraws at 
death. Far from being dependent upon the body, it dominates 
it from the higher dignity which is in it. It may well take from 
the body the outward form in which it individualizes itself, but 
it owes nothing essential to it. Nor is the autonomy which all 
peoples have attributed to the soul a pure illusion ; we know 
now what its objective foundation is. It is quite true that the 
elements which serve to form the idea of the soul and those 
which enter into the representation of the body come from two 
different sources that are independent of one another. One sort 
are made up of the images and impressions coming from all 
parts of the organism ; the others consist in the ideas and 
sentiments which come from and express society. So the former 
are not derived from the latter. There really is a part of our- 
selves which is not placed in immediate dependence upon the 
organic factor : this is all that which represents society in us. 

272 Elementary Forms of Religious Life 

The general ideas which reUgion or science fix in our minds, 
the mental operations which these ideas suppose, the beliefs 
and sentiments which are at the basis of our moral life, and all 
these superior forms of psychical activity which society awakens 
in us, these do not follow in the trail of our bodily states, as our 
sensations and our general bodily consciousness do. As we 
have already shown, this is because the world of representations 
in which social life passes is superimposed upon its material 
substratum, far from arising from it ; the determinism which 
reigns there is much more supple than the one whose roots 
are in the constitution of our tissues and it leaves with the 
actor a justified impression of the greatest liberty. The medium 
in which we thus move is less opaque and less resistant : we feel 
ourselves to be, and we are, more at our ease there. In a word, 
the only way we have of freeing ourselves from physical forces 
is to oppose them with collective forces. 

But whatever we receive from society, we hold in common 
with our companions. So it is not at all true that we are more 
personal as we are more individualized. The two terms are 
in no way synonymous : in one sense, they oppose more than 
they imply one another. Passion individualizes, yet it also 
enslaves. Our sensations are essentially individual ; yet we 
are more personal the more we are freed from our senses and 
able to think and act with concepts. So those who insist upon all 
the social elements of the individual do not mean by that to 
deny or debase the personality. They merely refuse to confuse 
it with the fact of individuation.^ 

' For all this, we do not deny the importance of the individual factor : this 
is explained from our point of view just as easily as its contrary. If the essential 
element of the personality is tlîe social part of us, on the other hand there can be 
no social life unless distinct individuals are associated, and this is richer the more 
numerous and different from each other they are. So the individual factor is a 
condition of the impersonal factor. And the contrary is no less true, for society 
itself is an important source of individual differences (see our Division du travail 
social, 3rd. éd., pp. 267 ff.). 



WHEN we come to the idea of the soul, we have left the 
circle of purely impersonal forces. But above the soul 
the Australian religions already recognize mythical personalities 
of a superior order : spirits, civilizing heroes and even gods who 
are properly so-called. While it will be unnecessary to enter 
into the detail of the mythologies, we mvi^^t at least seek the form 
in which these three categories of spiritual beings are presented 
in Australia, and the way in which they are connected with the 
whole religious system. 

A soul is not a spirit. In fact, it is shut up in a determined 
organism ; though it may leave it at certain moments, it is 
ordinarily a prisoner there. It definitely escapes only at death, 
and we have already seen the difficulties under which the separa- 
tion is accomplished. A spirit, on the contrary, though often 
tied by the closest bonds to some particular object, such as a 
spring, a rock, a tree, a star, etc., and though residing there by 
preference, may go away at will and lead an independent exist- 
ence in free space. So it has a more extended circle of action. 
It can act upon the individuals who approach it or whom it 
approaches. The soul, on the contrary, has almost no influence 
except over the body it animates ; it is very exceptional that 
it succeeds in influencing outside objects during the course of 
its terrestrial life. 

But if the soul does not have the distinctive characteristics of 
the spirit, it acquires them, at least in part, at death. In fact, 
when it has been disincarnated, so long as it does not descend 
into a body again, it has the same liberty of movement as a 
spirit. Of course, after the rites of mourning have been accom- 
plished, it is thought to go to the land of souls, but before this 
it remains about the tomb for a rather long time. Also, even 
after it has definitely departed, it is believed to prowl about in 


274 Elementary Forms of Religious Life 

the brush near the camp.^ It is generally represented as a 
rather beneficent being, especially for the surviving members 
of its family ; we have seen that the soul of the father comes to 
aid the growth of his children or his grandchildren. But it also 
happens sometimes that it shows signs of a veritable cruelty ; 
everything depends upon its humour and the manner in which 
it is treated by the living. ^ So it is recommended, especially to 
women and children, not to venture outside of the camp during 
the night so as not to expose oneself to dangerous encounters. ^ 

However, a ghost is not a real spirit. In the first place, it 
generally has only a limited power of action ; also, it does not 
have a definite province. It is a vagabond, upon whom no 
determined task is incumbent, for the effect of death has been 
to put it outside of all regular forms ; as regards the living, it 
is a sort of a exile. A spirit, on the other hand, always has a 
power of a certain sort and it is by this that it is defined ; it is 
set over a certain order of cosmic or social phenomena ; it has 
a more or less precise function to fulfil in the system of the 

But there are some souls which satisfy this double condition 
and which are consequently spirits, in the proper sense of the 
word. These are the souls of the mythical personages whom 
popular imagination has placed at the beginning of time, the 
Altjirangamitjina or the men of the Alcheringa among the Arunta ; 
the Mura-mura among the tribes of Lake Eyre ; the Muk-Kurnai 
among the Kurnai, etc. In one sense, they are still souls, for 
they are believed to have formerly animated bodies from which 
they separated themselves at a certain moment. But even when 
they led a terrestrial life, they already had, as we have seen, 
exceptional powers ; they had a mana superior to that of ordinary 
men, and they have kept it. Also, they are charged with definite 

In the first place, whether we a,ccept the version of Spencer 
and Gillen or that of Strehlow, it is to them that the care of 
assuring the periodical recruiting of the clan falls. They have 
charge of the phenomena of conception. 

Even when the conception has been accomplished, the task 
of the ancestor is not yet completed. It is his duty to guard 
over the new-born child. Later, when the child has become a 
man, he accompanies him in the hunt, brings game to him, 
warns him by dreams of the dangers he may run, protects him 

* Roth, Superstition, Magic, etc., §§ 65, 68 ; Spencer and Gillen, Nat. Tr., 
pp. 514, 516. 

* Spencer and Gillen, Nat. Tr., pp. 521, 515; Dawson, Austral. Aborig., 
p. 58 ; Roth, op. cit., § 67. 

' Spencer and Gillen, Nat. Tr., p. 517. 

The Idea of Spirits and Gods 275 

against his enemies, etc. On this point, Strehlow is entirely in 
accord with Spencer and Gillen.^ It is true that someone may 
ask how it is possible, according to the version of these latter, 
for the ancestor to fulfil this function ; for, since he reincarnates 
himself at the moment of conception, it seems as though he 
should be confounded with the soul of the child and should 
therefore be unable to protect it from without. But the fact is 
that he does not reincarnate himself entirely ; he merely dupli- 
cates himself. One part of him enters the body of the woman 
and fertilizes her ; another part continues to exist outside and, 
under the special name of Arumburinga, fulfils the office of 
guardian genius. ^ 

Thus we see how great a kinship there is between this ancestral 
spirit and the genius of the Latins or the ôaîfiœv of the Greeks.^ 
The identification of function is complete. In fact, at first 
the genius is the one who begets, qtii gignit ; he expresses and 
personifies the powers of generation.^ But at the same time, 
he is the protector and director of the particular individual 
to whose person he is attached.^ He is finally confused with 
the personality itself of this individual ; he represents the totality 
of the proclivities and tendencies which characterize him and 
give him a distinctive appearance among other men.* Hence 
come the well-known expressions indulgere genio, defraudere 
genium with the sense of to follow one's natural temperament. 
At bottom, the genius is another form or double of the soul 
of the individual. This is proved by the partial synonomy of 
genius and manes. "^ The manes is the genius after death ; but 
it is also all that survives of the dead man, that is to say, his 
soul. In the same way, the soul of the Arunta and the ancestral 
spirit which serves as his genius are only two different aspects 
of one and the same being. 

But it is not only in relation to persons that the ancestor 
has a definite situation ; he also has one in relation to things. 
Though he is believed to have his real residence under the ground, 
they think that he is always haunting the place where his nanja- 
tree or rock is, or the water-hole which was spontaneously 
formed at the exact spot where he disappeared into the ground, 
having terminated his first existence. As this tree or rock is 

1 Strehlow, II, p. 76 and n. i ; Spencer and Gillen, Nat. Tr., pp. 514, 516. 

* Spencer and Gillen, Nat. Tr., p. 513. 

' On this question, see Negrioli, Dei Genii presso i Romani ; the articles 
Daimon and Genius in the Diet, of Antiq. ; Preller, Romische Mythologie, II, 
pp. 193 ff- 

' Negrioli, ibid., p. 4. * Ibid., p. 8. * Ibid., p. 7. 

' Ibid., p. II. Cf. Samter, Der Ursprung der Larencultus, in Archiv f. 
Religions-wissenschaft, 1907, pp. 368-393. 

276 Elementary Forms of Religious Life 

believed to represent the body of the hero, they imagine that 
the soul itself is constantly coming back there, and lives there 
more or less permanently ; it is by the presence of this soul that 
they explain the religious respect inspired by these localities. 
No one can break the branch of a nanja-tree without a risk of 
falling sick.i " Formerly the act of breaking it down or injuring 
it was punished with death. An animal or bird taking refuge 
there could not be killed. Even the surrounding bushes had to 
be respected : the grass could not be burned, the rocks also had 
to be treated with respect. It was forbidden to remove them 
or break them," 2 As this sacred character is attributed to the 
ancestor, he appears as the spirit of this tree or rock, of this 
water-hole or spring. ^ If the spring is thought of as having some 
connection with rain,'^ he will become a spirit of rain. Thus, 
the same souls which serve as protecting geniuses for men also 
fulfil cosmic functions at the same time. It is undoubtedly in 
this sense that we must understand the text of Roth where he 
says that in northern Queensland, the spirits of nature are the 
souls of the dead who have chosen to live in the forests or caves.'' 
So we have here som.e spiritual beings that are different from 
the wandering souls with no definite powers. Strehlow calls 
them gods ; ^ but this expression is inexact, at least in the 
great majority of cases. If it were true, then in a society like 
the Arunta where each one has his protecting ancestor, there 
would be as many or more gods than there are individuals. 
It would merely introduce confusion into our terminology to 
give the name of god to a sacred being with only one worshipper. 
It may be, of course, that the figure of the ancestor grows to 
a point where it resembles a real divinity. Among the Warra- 
munga, as we have already pointed out,' the clan as a whole 
is thought to be descended from one sole and unique ancestor. 
It is easily seen how this collective ancestor might, under certain 
circumstances, become the object of a collective devotion. To 
choose a notable example, this is what has happened to the 
snake Wollunqua.^ This mythical beast, from whom the clan 
of the same name is held to be descended, continues to live, 
they believe, in water-holes which arc therefore surrounded 

' Schulzc, loc. cit., p. 237. 

- Strehlow, 1. p. 5. Cf. Spencer and Gillen, Nat. Tr., p. 133 ; Gason, in Curr, 
II. p. 69. 

' See the case of a Mura-mura who is considered the spirit of certain hot 
springs, in Howitt, Nat. Tr., p. 482. 

* Nor. Tr., pp. 313 f. ; Mathews, Journ. of the Roy. Soc. of N.S. Wales, 
XXXVIII, p. 351. Among the Dieri there is also a Mura-mura whose function 
is to produce rain (Howitt, Nat. Tr., pp. 798 f.). 

' Roth. Superstition, etc., § 67. Cf. Dawson, p. 59. 

•^ Strehlow, 1, pp. 2 fl. ' See above, p. 249. * Nor. Tr., ch. vii. 

The Idea of Spirits and Gods 277 

with a religious respect. Thus it becomes the object of a cult 
which the clan celebrates collectively : through determined 
rites, they attempt to please him and to win his favours, and 
they address to him all sorts of prayers, etc. So we may say 
that he is like a god of the clan. But this is a very exceptional 
case, or even, according to Spencer and Gillen, a unique one 
Normally, the word " spirits " is the only one suitable for desig- 
nating these ancestral personages. 

As to the manner in which this conception has been formed, 
we may say that it is evident from what has preceded. 

As we have already shown, the existence of individual souls, 
when once admitted, cannot be understood unless one imagines 
an original supply of fundamental souls at the origin of things, 
from which all the others were derived. Now these architype 
souls had to be conceived as containing within them the source 
of all religious efficacy ; for, since the imagination does not go 
beyond them, it is from them and only from them that all sacred 
things are believed to come, both the instruments of the cult, 
the members of the clan and the animals of the totemic species. 
They incarnate all the sacredness diffused in the whole tribe 
and the whole world, and so they are attributed powers noticeably 
superior to those enjoyed by the simple souls of men. Moreover, 
time by itself increases and reinforces the sacred character of 
things. A very ancient churinga inspires much more respect 
than a new one, and is supposed to have more virtues.^ The 
sentiments of veneration of which it has been the object during 
the series of successive generations who have handled it are, 
as it were, accumulated in it. For the same reason, the personages 
who for centuries have been the subject of myths respectfully 
passed on from mouth to mouth, and periodically put into action 
by the rites, could not fail to take a very especial place in the 
popular imagination. 

But how does it happen that, instead of remaining outside 
of the organized society, they have become regular members of 

This^ is because each individual is the double of an ancestor. 
Now when two beings are related as closely as this, they are 
naturally conceived as incorporated together ; since they partici- 
pate in the same nature, it seems as though that which affects 
one ought to affect the other as well. Thus the group of mythical 
ancestors became attached to the society of thé living ; the same 
interests and the same passions were attributed to each ; they 
were regarded as associates. However, as the former had a 
higher dignity than the latter, this association takes, in the 

* Spencer and Gillen, Nor. Tr., p. 277. 

278 Elementary Forms of Religious Life 

public mind, the form of an agreement between superiors and 
inferiors, between patrons and clients, benefactors and recipients. 
Thus comes this curious idea of a protecting genius who is 
attached to each individual. 

The question of how this ancestor came to have relations 
not only with men, but also with things, may appear more 
embarrassing ; for, at the first glance, we do not see what con- 
nection there can be between a personage of this sort and a 
rock or tree. But a fact which we owe to Strehlow furnishes us 
with a solution of this problem, which is at least probable. 

These trees and rocks are not situated at any point in the 
tribal territory, but, for the most part, they are grouped around 
the sanctuaries, called ertnatulunga by Spencer and Gillen and 
arknanaua by Strehlow, where the churinga of the clan is kept.^ 
We know the respect with which these localities are enhaloed 
from the mere fact that the most precious instruments of the 
cult are there. Each of these spreads sanctity all about it. 
It is for this reason that the neighbouring trees and rocks appear 
sacred, that it is forbidden to destroy or harm them, and that all 
violence used against them is a sacrilege. This sacred character 
is really due to a simple phenomenon of psychic contagiousness ; 
but in order to explain it, the native must admit that these 
different objects have relations with the different beings in 
whom he sees the source of all religious power, that is to say, 
with the ancestors of the Alcheringa. Hence comes the system 
of myths of which we have spoken. They imagined that each 
ertnatulunga marked the spot where a group of ancestors entered 
into the ground. The mounds or trees which covered the ground 
were beheved to represent their bodies. But as the soul retains, 
in a general way, a sort of affinity for the body in which it dwelt, 
they were naturally led to believe that these ancestral souls 
continued to frequent these places where their material envelope 
remained. So they were located in the rocks, the trees or the 
water-holes. Thus each of them, though remaining attached to 
some determined individual, became transformed into a sort of 
genius loci and fulfilled its functions. ^ 

1 strehlow, I, p. 5. 

2 It is true that some nanja-trees and rocks are not situated around the 
ertnatulunga ; they are scattered over different parts of the tribal territory. It 
is said that these are places where an isolated ancestor disappeared into the 
ground, lost a member, let some blood flow, or lost a churinga which was trans- 
formed into a tree or rock. But these totemic sites have only a secondary import- 
ance ; Strehlow calls them kleinere Totempliitze (I, pp. 4-5). So it may be that 
they have taken this character only by analogy with the principal totemic 
centres, 'the trees and rocks which, for some reason or other, remind one of 
those found in the neighbourhood of an ertnatulunga, inspire analogous senti- 
ments, so the myth which was formed in regard to the latter was extended to 
the former. 

The Idea of Spirits and Gods 279 

The conceptions thus elucidated enable us to understand 
a form of totemism which we have left unexplained up to the 
present : this is individual totemism. 

An individual totem is defined, in its essence, by the two 
following characteristics : (i) it is a being in an animal or vege- 
table form whose function is to protect an individual ; (2) the 
fate of this individual and that of his patron are closely united : 
all that touches the latter is sympathetically communicated to 
the former. Now the ancestral spirits of which we have just 
been speaking answer to this same definition. They also belong, 
at least in part, to the animal or vegetable kingdoms. They, 
too, are protecting geniuses. Finally, a sympathetic bond unites 
each individual to his protecting ancestor. In fact, the nanja- 
tree, representing the mystical body of this ancestor, cannot be 
destroyed without the man's feeling himself menaced. It is 
true that this belief is losing its force to-day, but Spencer and 
Gillen have observed it, and in any case, they are of the opinion 
that formerly it was quite general. ^ 

The identity of these two conceptions is found even in their 

The ancestral souls reside in trees or rocks which are considered 
sacred. Likewise, among the Euahlayi, the spirit of the animal 
serving as individual totem is beheved to inhabit a tree or stone. ^ 
This tree or stone is sacred ; no one may touch it except the 
proprietor of the totem ; when it is a stone or rock, this inter- 
diction is still absolute.^ The result is that they are veritable 
places of refuge. 

Finally, we have seen that the individual soul is only another 
aspect of the ancestral spirit, according to Strehlow, this serves 
after a fashion, as a second self.* Likewise, following an expression 
of Mrs. Parker, the individual totem of the Euahlayi, called 
Yunbeai, is the alter ego of the individual : " The soul of a man 
is in his Yunbeai and the soul of his Yunbeai is in him." * So at 
bottom, it is one soul in two bodies. The kinship of these two 
notions is so close that they are sometimes expressed by one 
and the same word. This is the case in Melanesia and in Poly- 
nesia : atai in the island Mota, tamaniii in the island Aurora, 
and talegia in Motlaw all designate both the soul of the individual 
and his personal totem. ^ It is the same with aitu in Samoa.' 

^ Nat. Tr., p. 139. 

- Parker, The Euahlayi, p. 21. The tree serving for this use is generally one 
of those figuring among the sub-totems of the individual. As a reason for this 
choice, they say that as it is of the same family as the individual, it should be 
better disposed to giving him aid (ibid., p. 29). 

» Ibid., p. 36. * Strehlow, II, p. 81. " Parker, op. cit., p. 21. 

• Codriugton, The Melanesians, pp. 249-253. ' Turner, Samoa, p. 17. 

28o Elementary Forms of Religious Life 

This is because the individual totem is merely the outward and 
visible form of the ego or the personality, of which the soul is 
the inward and invisible form.^ 

Thus the individual totem has all the essential characteristics 
of the protecting ancestor and fills the same rôle : this is because 
it has the same origin and proceeds from the same idea. 

Each of them, in fact, consists in a duplication of the soul. 
The totem, as the ancestor, is the soul of the individual, but 
externalized and invested with powers superior to those it is 
believed to possess while within the organism. Now this duplica- 
tion is the result of a psychological necessity ; for it only expresses 
the nature of the soul which, as we have seen, is double. In one 
sense, it is ours : it expresses our personality. But at the same 
time, it is outside of us, for it is only the reaching into us of a 
religious force which is outside of us. We cannot confound 
ourselves with it completely, for we attribute to it an excellence 
and a dignity by which it rises far above us and our empirical 
individuality. So there is a whole part of ourselves which we 
tend to project into the outside. This way of thinking of our- 
selves is so well established in our nature that we cannot escape 
it, even when we attempt to regard ourselves without having 
recourse to any religious symbols. Our moral consciousness is 
like a nucleus about which the idea of the soul forms itself ; yet 
when it speaks to us, it gives the effect of an outside power, 
superior to us, which gives us our law and judges us, but which 
also aids and sustains us. When we have it on our side, we 
feel ourselves to be stronger against the trials of life, and better 
assured of triumphing over them, just as the Australian who, 
when trusting in his ancestor or his personal totem, feels himself 
more valiant against his enemies. ^ So there is something objec- 
tive at the basis of these conceptions, whether we have in mind 
the Roman genius, the individual totem, or the Alcheringa 
ancestor ; and this is why they have survived, in various forms, 
up to the present day. Everything goes just as if we really had 
two souls ; one which is within us, or rather, which is us ; the 
other which is above us, and whose function it is to control and 

^ These are the very words used by Codrington (p. 251). 

* This close connection between the soul, the guardian genius and the moral 
conscience of the individual is especially apparent among certain peoples of 
Indonesia. " One of the seven souls of the Tobabatak is buried with the placenta ; 
though preferring to live in this place, it may leave it to warn the individual or 
to manifest its approbation when he does well. So in one sense, it plays the rôle 
of a moral conscience. However, its communications are not confined to the 
domain of moral facts. It is called the younger brother of the soul, as the 
placenta is called the younger brother of the child. ... In war, it inspires the 
man with courage to march against the enemy " (Wameck, Der bataksche Ahnen 
unci Geistercult, in Allg. Missionszeitschrift, Berlin, 1904. p. 10. Cf. Kruijt, Het 
Animisme in den imlischen Archipel, p. 25). 

The Idea of Spirits and Gods 281 

assist the first one. Frazer thought that the individual totem was 
an external soul ; but he believed that this exteriority was the 
result of an artifice and a magic ruse. In reality, it is implied 
in the very constitution of the idea of the soul.^ 


The spirits of which we have just been speaking are essentially 
benefactors. Of course they punish a man if he does not treat 
them in a fitting manner ; 2 but it is not their function to work 

However, a spirit is in itself just as capable of doing evil 
as good. This is why we find a class of evil geniuses forming 
itself naturally, in opposition to these auxiliary and protecting 
spirits, which enables men to explain the permanent evils that 
they have to suffer, their nightmares ^ and illnesses,^ whirlwinds 
and tempests,^ etc. Of course this is not saying that all these 
human miseries have appeared as things too abnormal to be 
explained in any way except by supernatural forces ; but it is 
saying that these forces are thought of under a religious form. 
As it is a religious principle which is considered the source of 
life, so, all the events which disturb or destroy life ought logically 
to be traced to a principle of the same sort. 

These harmful spirits seem to have been conceived on the 

* It still remains to be investigated how it comes that after a certain moment 
in evolution, this duplication of the soul was made in the form of an individual 
totem rather than of a protecting ancestor. Perhaps this question has an 
ethnological rather than a sociological interest. However, the manner in which 
this substitution was probably effected may be represented as follows. 

The individual totem commenced by playing a merely complimentary rôle. 
Those individuals who wished to acquire powers superior to those possessed by 
everybody, did not and could not content themselves with the mere protection 
of the ancestor ; so they began to look for another assistant of the same sort. 
Thus it comes about that among the Euahlayi, the magicians are the only ones 
who have or who can procure individual totems. As each one has a collective 
totem in addition, he finds himself having many souls. But there is nothing 
surprising in this plurality of souls : it is the condition of a superior power. 

But when collective totemism once begins to lose ground, and when the 
conception of the protecting ancestor consequently begins to grow dim in the 
mind, another method must be found for representing the double nature of the 
soul, which is still felt. The resulting idea was that, outside of the individual 
soul, there was another, charged with watching over the first one. Since this 
protecting power was no longer demonstrated by the very fact of birth, men 
found it natural to employ, for its discovery, means analogous to those used by 
magicians to enter into communion with the forces of whose aid they thus 
assured themselves. 

* For example, see Strehlow, II, p. 82. 

^ Wyatt, Adelaide and Encounley Bay Tribes, in Woods, p. 168. 
^ Taplin, The Narrinyeri, pp. 62 f. ; Roth, Superstition, etc., § 116 ; Howitt, 
Nat. Tr., pp. 356, 358 ; Strehlow, pp. 11-12. 
^ Strehlow, I, pp. 13-14 ; Dawson, p. 49. 

282 Elementary Forms of Religious Life 

same model as the good spirits of which we have just been 
speaking. They are represented in an animal form, or one that 
is half-animal, half-man ; ^ but men are naturally inclined to 
give them enormous dimensions and a repulsive aspect. ^ Like 
the souls of the ancestors, they are believed to inhabit trees, 
rocks, water-holes and subterranean caverns.^ Taking the 
Arunta as a particular example, Spencer and Gillen say expressly 
that these evil geniuses, known under the name of Oruncha, 
are beings of the Alcheringa.^ Many are represented as the 
souls of persons who had led a terrestrial life.^ Among the 
personages of the fabulous epoch, there were, in fact, many 
different temperaments : some had cruel and evil instincts which 
they retained ; ® others were naturally of a bad constitution ; 
they were thin and emaciated ; so after they had entered into 
the ground, the nanja rocks to which they gave birth were con- 
sidered the homes of dangerous influences. ' 

Yet they are distinguished by special characteristics from 
their confrères, the heroes of the Alcheringa. They do not 
reincarnate themselves ; among living men, there is no one who 
represents them; they are without human posterity.^ When, 
judging from certain signs, they believe that a child is the 
result of their work, it is put to death as soon as born.^ Also, 
these belong to no determined totemic group ; they are outside 
the social organization, i** By all these traits, they are recognized 
as magic powers rather than religious ones. And in fact, it is 
especially with the magician that they have relations ; very 
frequently it is from them that he gets his powers. ^^ So we 
have now arrived at the point where the world of religion 
stops and that of magic commences ; and as this latter is 

1 Strehlow, I, pp. 11-14 ; Eylmann, pp. 182, 185 ; Spencer and Gillen, Nor. 
Jr., p. 211 ; Schurmann, The Aborig. Tr. of Fori Lincoln, in Woods, p. 239. 

* Eylmann, p. 182. 

» Mathews, Journ. of the Roy. Soc. of N.S. Wales, XXXVIII, p. 345 ; Fison 
and Hewitt, Kamilaroi and Kurnai, p. 467 ; Strehlow, I, p. 11. 

* Nat. Tr., pp. 390-391. Strehlow calls these evil spirits Erintja ; but this 
word is evidently equivalent to Oruncha. Yet there is a difference in the ways 
the two are presented to us. According to Spencer and Gillen, the Oruncha are 
malicious rather than evil ; they even say (p. 328) that the Arunta know no 
necessarily evil spirits. On the contrary, the regular business of Strehlow 's 
Erintja is to do evil. Judging from certain myths given by Spencer and Gillen 
(Nat. Tr., p. 390). they seem to have touched up the figures of the Oruncha a 
little : these were originally ogres {ibid., p. 331). 

6 Roth, Superstition, etc., § 115 ; Eylmann, p. 190. 
« Nat. Tr., pp. 390 f. ' Ibid., p. 551. 

" Ibid., pp. 326 f. 

* Strehlow, I, p. 14. When there are twins, the first one is believed to have 
been conceived in this manner. 

>o Spencer and Gillen, Nat. Tr., p. 327. 

" Howitt, Nat. Tr., pp. 358, 381, 385 ; Spencer and Gillen, Nat. Tr.. p. 334 ; 
Nor. Tr., pp. 501, 530. 

The Idea of Spirits and Gods 283 

outside the field of our research, we need not push our researches 
further. 1 


The appearance of the notion of spirits marks an important 
step in advance in the individuaUzation of rehgious forces. 

However, the spiritual beings of whom we have been speaking 
up to the present are as yet only secondary personages. They 
are either evil-working geniuses who belong to magic rather 
than religion, or else, being attached to determined individuals 
or places, they cannot make their influence felt except within a 
circle of a very limited radius. So they can only be the objects 
of private and local rites. But after the idea has once been 
established, it naturally spreads to the higher spheres of the 
religious life, and thus mythical personalities of a superior order 
are bom. 

Though the ceremonies of the different clans differ from one 
another, they all belong to the same religion, none the less ; also, 
a certain number of essential similarities exist between them. 
Since all the clans are only parts of one and the same tribe, the 
unity of the tribe cannot fail to make itself felt through this 
diversity of particular cults. In fact, there is no totemic group 
that does not have churinga and bull-roarers, and these are used 
everywhere in the same way. The organization of the tribe 
into phratries, matrimonial classes and clans, and the exogamic 
interdictions attached to them, are veritable tribal institutions. 
The initiation celebrations all include certain fundamental 
practices, the extraction of a tooth, circumcision, subincision, 
etc., which do not vary with the totems within a single tribe. 
The uniformity on this point is the more easily established 
as the initiation always takes place in the presence of the tribe, 
or at least, before an assembly to which the different clans have 
been summoned. The reason for this is that the object of the 
initiation is to introduce, the neophyte into the religious life, 
not merely of the clan into which he was born, but of the tribe 
as a whole ; so it is necessary that the various aspects of the 
tribal religion be represented before him and take place, in a 
way, under his very eyes. It is on this occasion that the moral 
and religious unity of the tribe is affirmed the best. 

Thus, in each society there are a certain number of rites 

> As the magician can either cause or cure sickness, we sometimes find, 
besides these magical spirits whose function is to do evil, others who forestall or 
neutralize the evil influence of the former. Cases of this sort will be found in 
Nor. Tr., pp. 501-502. The fact that the latter are magic just as much as the 
former is well shown by the fact that the two have the same name, among the 
Arunta. So they are different aspects of a single magic power. 

284 Elementary Forms of Religious Life 

which are distinguished from all the others by their homogeneity 
and their generality. So noticeably a harmony seemed to be 
explicable only by a unity of origin. So they imagined that 
each group of similar rites had been founded by one and the same 
ancestor, who came to reveal them to the tribe as a whole. Thus, 
among the Arunta, it was an ancestor of the Wild Cat clan, 
named Putiaputia,^ who is thought to have taught men the way 
of making churinga and using it ritually ; among the Warra- 
munga, it was Murtu-murtu ; ^ among the Urabunna, Witurna ; ^ 
it was Atnatu among the Kaitish ^ and Tendun among the 
Kumai.5 Likewise, the practice of circumcision is attributed 
by the eastern Dieri and many other tribes ^ to two special 
Muramura, and by the Arunta to a hero of the Alcheringa, of 
the Lizard totem, named Mangarkunjerkunja.' To this same 
personage are ascribed the foundation of the matrimonial in- 
stitutions and the social organization they imply, the discovery 
of fire, the invention of the spear, the buckler, the boomerang, 
etc. It also happens very frequently that the inventor of the 
bull-roarer is also considered the founder of the rites of initiation.^ 
These special ancestors cannot be put in the same rank as 
the others. On the one hand, the sentiments of veneration 
which they inspire are not limited to one clan, but are common 
to the whole tribe. On the other hand, it is to them that men 
ascribe all that is most esteemed in the tribal civilization. For 
this double reason, they became the object of a special con- 
sideration. For example, they say of Atnatu that he was born 
in heaven at an epoch even prior to the times of the Alcheringa, 
that he made himself and that he gave himself the name he 
bears. The stars are his wives and daughters. Beyond the 
heaven where he lives, there is another one with another sun. 
His name is sacred, and should never be pronounced before 
women or non-initiated persons.* 

* Strehlow, I, p. g. Putiaputia is not the only personage of this sort of whom 
the Arunta myths speak : certain portions of the tribe give a different name to 
the hero to whom the same invention is ascribed. We must not forget that the 
extent of the territory occupied by the Arunta prevents their mythology from 
being completely homogeneous. 

^ Spencer and Gillen, Nor. Tr., p. ^03. ' Ibid., p. 498. 

* Ibid., pp. 498 f. * Howitt, Nat. Tr.. p. 135. 

* Ibid., pp. 476 ff. 

' Strehlow, I, pp. 6-8. The work of Mangarkunjerkunja must be taken up 
again later among other heroes ; for, according to a belief that is not confined to 
the Arunta, a time came when men forgot the teaching of their first initiators and 
became corrupt. 

* This is the case, for example, of Atnatu (Spencer and Gillen, Nor. Tr., 
p. 153) and the Witurna {Nor. Tr., p. 498), If Tendun did not establish these 
rites, it is he who is charged with the direction of their celebration (Howitt, Nat. 
Tr., p. 670). 

* Nor. Tr., p. 499. 

The Idea of Spirits and Gods 285 

Yet, howsoever great the prestige enjoyed by these personages 
may be, there was no occasion for founding special rites in their 
honour ; for they themselves are only rites personified. They 
have no other reason for existence than to explain existing 
practices ; they are only another aspect of these. The churinga 
and the ancestor who invented it are only one ; sometimes, 
both have the same name.^ When someone makes the bull- 
roarer resound, they say that it is the voice of the ancestor 
making himself heard. 2 But, for the very reason that each 
of these heroes is confounded with the cult he is believed to 
have founded, they believe that he is attentive to the way in 
which it is celebrated. He is not satisfied unless the worshippers 
fulfil their duties exactly ; he punishes those who are negligent. ^ 
So he is thought of as the guardian of the rite, as well as its 
founder, and for this reason, he becomes invested with a veritable 
moral rôle.^ 


However, this mythological formation is not the highest which 
is to be found among the Australians. There are at least a 
certain number of tribes who have arrived at a conception 
of a god who, if not unique, is at least supreme, and to whom 
is attributed a pre-eminent position among all the other religious 

The existence of this belief was pointed out long ago by 
different observers ; ^ but it is Howitt who has contributed the 
most to establishing its relative generality. In fact, he has 
verified it over a very extended geographical area embracing 
the State of Victoria and New South Wales and even extending 
up to Queensland.* In all this entire region, a considerable 
number of tribes believe in the existence of a veritable tribal 
divinity, who has different names, according to the district. The 
ones most frequently employed are Bunjil or Punjil,' Daramulun^ 

1 Howitt, Nat. Tr., p. 493 ; Kamilaroi and Kurnai, pp. 197 and 247 ; Spencer 
and Gillen, Nor. Tr.. p. 492. 

- For example, see Nor. Tr., p. 499. 

' AW. Tr., pp. 338, 347, 49g. 

* It is true that Spencer and Gillen maintain that these mythical beings play- 
no moral rôle {Nor. Tr., p. 493) ; but this is because they give too narrow a 
meaning to the word. Religious duties are duties : so the fact of looking after 
the manner in which these are observed concerns morals, especially because all 
morals have a religious character at this period. 

'■' The fact was observed as early as 1845 by Eyre, Journals, etc., II, p. 362, 
and, before Eyre, by Henderson, Observations on the Colonies of N.S. Wales and 
Van Diemen's Land, p. 147. 

« Nat. Tr.. pp. 488-508. 

' Among the Kulin, Wotjobaluk and Woëworung (Victoria). 

" Among the Yuin, Ngarrigo and Wolgal (New South Wales). 

286 Elementary Forms of Religious Life 

and Baiame.i But we also find Nuralie or Nurelle,^ Kohin ^ 
and Mangan-ngaua.^ The same conception is found again 
farther west, among the Narrinyeri, where the great god is called 
Nurunderi or Ngurrunderi.^ Among the Dieri, it is probable 
that there is one of the Mura-mura, or ordinary ancestors, who 
enjoys a sort of supremacy over the others.^ Finally, in oppo- 
sition to the affirmations of Spencer and Gillen, who declare 
that they have observed no belief in a real divinity among the 
Arunta,' Strehlow assures us that this people, as well as the 
Loritja, recognize, under the name Alt j ira, a veritable " good 
god." 8 

The essential characteristics of this personage are the same 
everywhere. It is an immortal, and even an eternal being, 
for it was not derived from any other. After having lived on 
earth for a certain length of time, he ascended to heaven, or 
else was taken up there, ^ and continues to live there, surrounded 
by his family, for generally he is said to have one or several wives, 
children and brothers, ^^ who sometimes assist him in his functions. 
Under the pretext of a visit he is said to have made to them, 
he and his family are frequently identified with certain stars. ^^ 

^ Among the Kamilaroi and Euahlayi (northern part of New South Wales) ; 
and more to the centre, in the same province, among the Wonghibon and the 

^ Among the Wiimbaio and the tribes on the lower Murra3- (Ridley, Kamilaroi, 
p. 137 ; Brough Smyth, I, pp. 423, n., 431). 

^ Among the tribes on the Herbert River (Howitt, Nat. Tr., p. 49S). 

* Among the Kurnai. * Taplin, p. 55 ; Eylmann, p. 182. 

* It is undoubtedly to this supreme Mura-mura that Gason makes allusion 
in the passage already cited (Curr, II, p. 55). 

' Nat. Tr., p. 24O. 

* Between Baiame, Bunjil and Daramulun on the one hand, and Altjira on 
the other, there is the difference that the latter is completely foreign to all that 
concerns humanity ; he did not make man and does not concern himâelf with 
what they do. The Arunta have neither love nor fear for him. But when this 
conception is carefully observed and analysed, it is hard to admit that it is 
primitive ; for if the Altjira plays no rôle, explains nothing, serves for nothing, 
what made the Arunta imagine him ? Perhaps it is necessary to consider him 
as a sort of Baiame who has lost his former prestige, as an ancient god whose 
memory is fading away. Perhaps, also, Strehlow has badly interpreted the 
testimony he has gathered. According to Eylmann, who, it is to be admitted, is 
neither a very competent nor a very sure observer, Altjira made men (op. cit., 
p. 134). Moreover, among the Loritja, the corresponding personage, iukura, is 
believed to celebrate the initiation ceremonies himself. 

* For Bunjil, see Brough Smyth, 1, p. 417 ; for Baiame, see Ridley, Kami- 
laroi, p. 136 ; for Daramulun, see Howitt, Nat. Tr., p. 495. 

*" On the composition of Bunjil's family, for example, see Howitt, Nat. Tr., 
pp. 128, 129, 489, 491 ; Brough Smyth, I, pp. 417, 423 ; for Baiame's, see 
L. Parker, The Euahlayi, pp. 7, 66, 103 ; Howitt, Nat. Tr., pp. 502, 585, 407 ; for 
Nurunderi's, Taplin, The Narrinyeri, pp. 57 f. Of course, there are all sorts of 
variations in the ways in which the families of these great gods are conceived. 
The personage who is a brother here, is a son there. The number and names of 
tlie wives vary with the locality. 

" Howitt, Nat. Tr., p. 128. 

The Idea of Spirits and Gods 287 

Moreover, they attribute to him a power over stars. It is he 
who regulates the journey of the sun and moon ; ^ he gives them 
orders. 2 It is he who makes the hghtning leap from the clouds 
and who throws the thunder-bolts.^ Since he is the thunder, 
he is also connected with the rain : * it is to him that men address 
themselves when there is a scarcity of water, or when too much 
falls. 6 

They speak of him as a sort of creator : he is called the father 
of men and they say that he made them. According to a legend 
current around Melbourne, Bunjil made the first man in the 
following manner. He made a little statue out of white clay ; 
then, after he had danced all around it several times and had 
breathed into its nostrils, the statue became animated and 
commenced to walk about.* According to another myth, he 
lighted the sun ; thus the earth became heated and men came 
out of it.' At the same time that he made men,^ this divine 
personage made the animals and trees ; ^ it is to him that 
men owe all the arts of life, arms, language and tribal rites. ^•^ 
He is the benefactor of humanity. Even yet, he plays the rôle 
of a sort of providence for them. It is he who supplies his wor- 
shippers with all that is necessary for their existence. ^^ He is 
in communication with them, either directly or through inter- 
mediaries. ^^ But being at the same time guardian of the morals 
of the tribe, he treats them severely when these are violated. ^^ 
If we are to believe certain observers, he will even fulfil the 
office of judge, after this life ; he will separate the good from 
the bad, and will not reward the ones like the others. ^^ In any 
case, they are often represented as ruling the land of the dead,^^ 
and as gathenng the souls together when they arrive in the 

^ Brough Smyth, I, pp. 430, 431. * Ibid., I, p. 432, n. 

' Howitt, Nat. Tr., pp. 498, 53S ; Mathews, Jour, of the Roy. Sac. of N.S. 
Wales. XXXVIII, p. 343 ; Ridley, p. 136. 

* Howitt, Nat. Tr., p. 538 ; Taplin, The Narrinyeri, pp. 57-58. 
' L. Parker, The Euahlayi, p. 8. 

* Brough Smyth, I, p. 424. 
' Howitt, Nat. Tr., p. 492. 

* According to certain myths, he made men but not women ; this is related 
of Bunjil. But then, the origin of women is attributed to his son-brother, 
Pallyan (Brough Smyth, I, pp. 417 and 423). 

* Howitt, Nat. Tr., pp. 4S9, 492 ; Mathews, Journ. of the Roy. Soc. of N.S. 
Wales, XXXVIII, p. 340. 

^* L. Parker, The Euahlayi, p. 7 ; Howitt, Nat. Tr., p. 630. 

*^ Ridley, Kamilaroi, p. 136 ; L. Parker, The Euahlayi, p. 114. 

** L. Parker, More Austr. Leg. Tales, pp. 84-89, 90-91. 

'* Howitt, Nat. Tr., pp. 495, 498, 543, 563, 564 ; Brough Smyth, I, p. 429 ; 
L. Parker, The Euahlayi, pp. 79. 

** Ridley, p. 137. 

1' L. Parker, The Euahlayi, pp. 90-91. 

*" Howitt, Nat. Tr., p. 495 ; Taplin, The Narrinyeri, p. 58. 

288 Elementary Forms of Religious Life 

As the initiation is the principal form of the tribal cult, it is 
to the rites of initiation that he is attached especially ; he is 
their centre. He is very frequently represented by an image 
cut on a piece of bark or soaked into the ground. They dance 
around it ; they sing in its honour ; they even address real 
prayers to it.^ They explain to the young men who the personage 
is whom this image represents ; they tell them his secret name, 
which the women and the uninitiated cannot know ; they relate 
to them his history and the part attributed to him in the life of 
the tribe. At other times they raise their hands towards the 
heaven where he is thought to dwell, or else they point their 
arms or the ritual instruments they have in hand in this direction ;- 
this is a way of entering into communication with him. They 
feel his presence everywhere. He watches over the neophyte 
when he has withdrawn into the forest.^ He is attentive to the 
manner in which the ceremonies are celebrated. The initiation 
is his cult. So he gives special attention to seeing that these 
are carried out exactly : if there are any faults or negligences, 
he punishes them in a terrible manner.^ 

Moreover, the authority of each of these supreme gods is not 
limited to a single tribe ; it is recognized equally by a number 
of neighbouring tribes. Bunjil is adored in nearly all of Victoria, 
Baiamc in a considerable portion of New South Wales, etc. ; 
this is why there are so few gods for a relatively extended 
geographical area. So the cults of which they are the object have 
an international character. It even happens sometimes that 
mythologies intermingle, combine and make mutual borrowings. 
Thus the majority of the tribes who believe in Baiame also admit 
the existence of Daramulun ; however, they accord him a slighter 
dignity. They make him a son or brother of Baiame, and sub- 
ordinate to this latter.^ Thus the faith in Daramulun has spread 
in diverse forms, into all of New South Wales. So it is far from 
true that religious internationalism is a peculiarity of the most 
recent and advanced religions. From the dawn of history, 
religious beliefs have manifested a tendency to overflow out of 
one strictly limited political society ; it is as though they had 
a natural aptitude for crossing frontiers, and for diffusing and 
internationalizing themselves. Of course there have been 

1 Howitt, Nat. Tr., pp. 538, 543, 553, 555, 556; Mathews, loc. cit., p. 318 ; 
L. Parker, The Euahlayi, pp. 6, 79, 80. 

"^ Howitt, Nat. Tr., pp. 498, 528. 

^ Howitt, ibid., p. 493 ; L. Parker, The Euahlayi, p. 76. 

* L. Parker, The Euahlayi, p. 76 ; Howitt, Nat. Tr., pp. 493, G12. 

'' Ridley, Kamilaroi, p. 153 ; L. Parker, The Euahlayi, p. 67 ; Howitt, Nat 
Tr., p. 585 ; Mathews, loc. cit., p. 343. In opposition to Baiame, Daramuhin is 
sometimes presented as a necessarily evil spirit (L. Parker, loc. cit. ; Ridley, in 
Brough Smyth, II, p. 285). 

The Idea of Spirits and Gods 28g 

peoples and times when this spontaneous aptitude has been held 
in check by opposed social necessities ; but that does not keep 
it from being real and, as we see, very primitive. 

To Tylor this conception has appeared to be a part of so 
elevated a theology that he refuses to see in it anything but the 
product of a European importation : he would have it be a more 
or less denatured Christian idea.^ Andrew Lang, on the contrary, 
considers them autochthonous ; ^ but as he also admits that it 
is contrasted with all the other Australian beliefs and rests on 
completely different principles, he concludes that the religions 
of Australia are made up of two heterogeneous systems, super- 
imposed one upon the other, and consequently derived from a 
double origin. On the one hand, there were ideas relative to 
totems and spirits, which had been suggested to men by the 
sight of certain natural phenomena. But at the same time, by 
a sort of intuition as to the nature of which he refuses to make 
himself clear, ^ the human intelligence succeeded at the first 
onset in conceiving a unique god, creator of the world and 
legislator of the moral order. Lang even estimates that this 
idea was purer of foreign elements at the beginning, and especially 
in Australia, than in the civilizations which immediately followed. 
With time, it was covered over and obscured little by little by the 
ever-growing mass of animistic and totemic superstitions. Thus 
it underwent a sort of progressive degeneration up to the day 
when, as the effect of a privileged culture, it succeeded in coming 
into its own and restated itself again with more force and clarity 
than it had in the first place.'' 

But the facts allow neither the sceptical hypothesis of Tylor 
nor the theological interpretation of Lang. 

In the first place, it is certain to-day that the ideas relative 
to the great tribal god are of indigenous origin. They were 

» J.A.I., XXI, pp. 292 (Ï. 

* The Making of Religion, pp. 187-293. 

* Lang, ibid., p. 331. The author confines himself to stating that the hypo- 
thesis of St. Paul does not appear to him " the most unsatisfactory." 

* The thesis of Lang has been taken up again by Father Schmidt in the 
Anthropos (1908-1909). Replying to Sydney Hartland, who had criticized Lang's 
theory in an article entitled The " High Gods " of Australia, in Folk-Lore (Vol. IX, 
pp. 290 £f.), Father Schmidt undertook to show that Baiame, Bunjil, etc., are 
eternal gods, creators, omnipotent, omniscient and guardians of the moral order. 
We are not going to enter into this discussion, which seems to have neither 
interest nor importance. If these different adjectives are given a relative sense, 
in harmony with the Australian mind, we are quite ready to accept them, and 
have even used them ourselves. From this point of view, omnipotent means 
having more power than the other sacred beings ; omniscient, seeing things that 
escape the vulgar and even the greatest magicians ; guardian of the moral order, 
one causing the rules of Australian morality to be respected, howsoever much 
these may differ from our own. But if they want to give these words meanings 
which only a spiritualistic Christian could attach to them, it seems useless to 
discuss an opinion so contrary to the prmciples of the historical method. 

290 Elementary Forms of Religious Life 

observed before the influence of the missionaries had as yet 
had time to make itself felt.^ But it does not follow that it is 
necessary to attribute them to a mysterious revelation. Far 
from being derived from a different source than the regular 
totemic beliefs, they are, on the contrary, only the logical working- 
out of these beliefs and their highest form. 

We have already seen how the notion of mythical ancestors 
is implied in the very principles upon which totemism rests, 
for each of them is a totemic being. Now, though the great gods 
are certainly superior to these, still, there are only differences 
of degree between them ; we pass from the first to the second 
with no break of continuity. In fact, a great god is himself an 
ancestor of especial importance. They frequently speak to us 
about him as though he were a man, endowed, to be sure, with 
more than human powers, but one who lived a human life upon 
the earth. 2 He is pictured as a great hunter,^ a powerful magician,^ 
or the founder of the tribe. ^ He was the first man.' One legend 
even represents him in the form of a worn-out old man who 
could hardly move about.' If a supreme god named Mura-mura 
has existed among the Dieri, the very word is significant, for it 
serves to designate the class of the ancestors. Likewise, Nuralie, 
the name of a great god among the tribes on the Murray River, 
is sometimes used as a collective expression which is applied to 
the group of mythical beings whom tradition places at the origin 
of things.^ They are personages wholly comparable to those 
of the Alcheringa.* In Queensland, we have already met with 
a god Anjea or Anjir, who made men but who seems, nevertheless, 
to be only the first man.^** 

A fact that has aided Australian thought to pass from the 
numerous ancestral geniuses to the idea of the tribal god is that 
between the two extremes a middle term has been inserted, 
which has served as a transition : these are the civilizing heroes. 
The fabulous beings whom we call by this name are really simple 

* On this question, see N. W. Thomas, Baiame and Bell-hird — A Note on 
Australian Religion, in Man, 1905, No. 28. Cf. Lang, Magic and Religion, p. 25. 
Waitz had already upheld the original character of this conception in his A nthro- 
pologie d. Naturvolker, pp. 796—798. 

' Dawson, p. 49 ; Meyer, Encounter Bay Tribe, in Woods, pp. 205, 206 ; 
Howitt, Nat. Tr., pp. 481, 491, 492, 494 ; Ridley, Kamilaroi, p. 136. 
' Taplin, The Narrinyeri, pp. 55—56. 

* L. Parker, More Austr. Leg. Tales, p. 94. 

* Brough Smyth, I, pp. 425-427. • Taplin, ibid., p. 60. 
' Taplin, ibid., p. 61. 

* " The world was created by beings called Nuralie ; these beings, who had 
already long existed, had the forms of crows or of eagle-hawks " (Brough Smyth, 
I, pp. 423-424). 

» " Bayamee," says Mrs. Parker, " is for the Euahlayi what the Alcheringa is 
for the Arunta " {The Euahlayi, p. 6). 
'" See above, pp. 257 f. 

The Idea of Spirits and Gods 291 

ancestors to whom mythology has attributed an eminent place 
in the history of the tribe, and whom it has, for this reason, set 
above the others. We have even seen that they ordinarily form 
a part of the totemic organization : Mangarkunjerkunja belongs 
to the Lizard totem and Putiaputia to the Wild Cat totem. 
But on the other hand, the functions which they are believed 
to fulfil, or to have fulfilled, are closely similar to those incumbent 
upon a great god. He, too, is believed to have introduced the 
arts of civilization among men, to have been the founder of the 
principal social institutions and the revealer of the great religious 
ceremonies which still remain under his control. If he is the 
father of men, it is because he manufactured them rather than 
begat them : but Mangarkunjerkunja also made them. Before 
his time, there were no men, but only unformed masses of flesh, 
in which the different members and even the different individuals 
were not yet separated from one another. It was he who cut 
up this original matter and made real human beings out of it.^ 
Between this mode of fabrication and the one the myth we 
have spoken of attributes to Bun j il, there are only shades of 
difference. Moreover, the bonds uniting these two sorts of 
figures to each other are well shown by the fact that a relationship 
of descent is sometimes established between them. Among 
the Kumai, the hero of the bull-roarer, Tundun, is the son of 
the great god Mungan-ngaua.^ Likewise, among the Euahlayi, 
Daramulun, the son or brother of Baiame, is identical with 
Gayandi who is the equivalent of the Timdun of the Kurnai.^ 
Of course it is not necessary to conclude from these facts that 
the great god is nothing more than a civilizing hero. There are 
cases where these two personages are carefully differentiated. 
But if they are not confounded, they are at least relatives. 
So it sometimes happens that we find it hard to distinguish 
them ; there are some who could be classified equally well in 
one category or the other. Thus, we have spoken of Atnatu 
as a civilizing hero ; but he comes very near to being a great 

The notion of a supreme god even depends so closely upon 
the entire system of the totemic beliefs that it still bears their 
mark. Tundun is a divine hero, as we have just seen, who is 
very close to the tribal divinity ; now among the Kumai, the 

^ In another myth, reported by Spencer and Gillen, a wholly analogous rôle 
is filled by two personages living in heaven, named Ungambikula [Nat. Tr., 
pp. 388 ff.). 

* Howitt, Nat. Tr., p. 493. 

' Parker, The Euahlayi, pp. 62-66, 67. This is because the great god is 
connected with the bull-roarer, which is identified with the thunder ; for the 
roaring of this ritual instrument is connected with the rolling of thunder. 

292 Elementary Forms of Religious Life 

same word means totem. 1 Similarly, among the Arunta, Altjira 
is the name of a great god ; it is also the name of the maternal 
totem. 2 But there is more to be said than this ; many great 
gods have an obviously totemic aspect. Daramulun is an eagle- 
hawk ; ^ his mother, an emu.* It is also under the features of 
an emu that Baiame is represented.^ The Altjira of the Arunta 
has the legs of an emu.® Before being the name of a great god, 
Nuralie designated, as we just saw, the ancestor-founders of 
the tribe ; now some of these were crows, the others hawks.' 
According to Howitt,^ Bun j il is always represented in a human 
form ; however, the same word serves to designate the totem 
of a phratry, the eagle-hawk. At least one of his sons is among 
the totems included in the phratry to which he has given, or from 
which he has taken his name.^ His brother is Pallyan, the 
bat ; now this latter serves as sexual totem for the men in 
many tribes in Victoria. ^° 

We can even go farther and state more definitely the con- 
nection which these great gods have with the totemic system. 
We have just seen that Bimjil is the totem of a phratry. Dara- 
mulun, like Bunjil, is an eagle-hawk, and we know that this bird 
is the totem of phratries in a large number of south-eastern 
tribes. ^^ We have already pointed out that Nuralie seems to 
have originally been a collective term designating indistinctly 
either eagle-hawks or crows ; now in the tribes where this myth 
has been observed, the crow is the totem of one of the two phra- 
tries, the eagle-hawk, that of the other. ^^ Also, the legendary 
history of the great gods resembles that of the totems of the 
phratries very closely. The myths, and sometimes the rites, 
commemorate the struggles which each of these divinities fought 
against a carnivorous bird, over which it triumphed only with 
the greatest difficulty. Bunjil, the first man, after making the 
second man, Karween, entered into a confiict with him, and in 
the course of a sort of duel, he wounded him severely and changed 

1 Howitt, Nat. Tr., p. 135. The word meaning totem is written thundung by 

^ Strehlow, I, pp. 1-2 and II, p. 59. It will be remembered that, among the 
Arunta, the maternal totem was quite probably the real totem at first. 

^ Howitt, Nat. Tr., p. 555. 

* Ibid., pp. 546, 560. 

'•• Ridley, Kamilaroi, pp. 136, 156. He is represented in this form during the 
initiation rites of the Kamilaroi. According to another legend, he is a black 
swan (L. Parker, More Aust. Leg. Tales, p. 94). 

* Strehlow, I, p. i. ' Brough Smyth, I, pp. 423-424. 
8 Nat. Tr., p. 492. » Howitt, Nat. Tr., p. 128. 

1" Brough Smyth, I, pp. 417-423. ^^ See above, p. 108. 

1* There are phratries bearing the names Kilpara (crow) and Mukwara. This 
is the explanation of the myth itself, which is reported by Brough Smyth (I, 
PP- 423-4-4). 

The Idea of Spirits and Gods 293 

him into a crow.* The two species of NurtaUe are represented 
as two hostile groups which were originally in a constant state 
of war." Baiame, on his side, had to fight against Mullian, the 
cannibal eagle-hawk, who, by the way, is identical with Dara- 
mulun.^ Now, as we have seen, there is also a sort of constitu- 
tional hostility between the totems of the phratries. This 
parallelism completes the proof that the mythology of the great 
gods and that of these totems are closely related. This relation- 
ship will appear still more evident if we notice that the rival of 
the god is regularly either a crow or an eagle-hawk, and that 
these are quite generally the totems of the phratries.* 

So Baiame, Daramulun, Nuralie and Bunjil seem to be phratry- 
totems who have been deified ; and we may imagine that this 
apotheosis took place as follows. It is obviously in the assemblies 
which take place in regard to the initiation that the conception 
was elaborated, for the great gods do not play a rôle of any 
importance except in these rites, and are strangers to the other 
religious ceremonies. Moreover, as the initiation is the principal 
form of the tribal cult, it is only on this occasion that a tribal 
mythology could arise. We have already seen how the rituals 
of circumcision and subincision spontaneously tend to personify 
themselves under the form of civilizing heroes. However, these 
heroes exercised no supremacy ; they were on the same footing 
as the other legendary benefactors of society. But wherever 
the tribe acquired a livelier sentiment of itself, this sentiment 
naturally incarnated itself in some personage, who became 
its symbol. In order to account for the bonds uniting them to 
one another, no matter what clan they belonged to, men imagined 
that they were all descended from the same stock and that they 
were all descended from a single father, to whom they owe their 
existence, though he owed his to no one. The god of the initiation 
was predestined to this rôle, for, according to an expression 
frequently coming to the lips of the natives, the object of the 
initiation is to make or manufacture men. So they attributed 
a creative power to this god, and for all these reasons, he found 
himself invested with a prestige setting him well above the 
other heroes of the mythology. These others became his auxili- 
aries, subordinate to him ; they were made his sons or younger 
brothers, as was the case with Tundun, Gayandi, Karween, 

1 Brough Smyth, I, pp. 425-427. Cf. Howitt, Nat. Tr.. p. 486. In this case. 
Karween is identified with the blue heron. 

^ Brough Smyth, I, p. 423. 

' Ridley, Kamilaroi, p. 136 ; Howitt, Nat. Tr., p 585; Mathews, /. of R. S. 
of N.S. Wales, XXVIII (1894), p. iii. 

* See above, p. 145. Cf. Father Schmidt, The Origin of the Idea of God, in 
Anthropos, 1909. 

294 Elementary Forms of Religious Life 

Pally an, etc. But other sacred beings already existed, who 
occupied an equally eminent place in the religious system of 
the clan : these were the totems of the phratries. Wherever 
these are maintained, they are believed to keep the totems 
of the clans dependent upon them. Thus they had all that was 
necessary for becoming tribal divinities themselves. So it was 
only natural that a partial confusion should arise between these 
two sorts of mythical beings ; it is thus that one of the two 
fundamental totems of the tribe gave his traits to the great god. 
But as it was necessary to explain why only one of them was 
called to this dignity and the other excluded, they supposed 
that this latter, in the course of a fight against his rival, was 
vanquished and that his exclusion was the consequence of his 
defeat. This theory was the more readily admitted because it 
was in accord with the rest of the mythology, where the totems 
of the phratries are generally considered enemies of one another. 

A myth observed by Mrs. Parker among the Euahlayi ^ may 
serve to confirm this explanation, for it merely translates it 
into figurative language. It is related that in this tribe, the 
totems were only the names given to the different parts of 
Baiame's body at first. So the clans were, in a sense, the frag- 
ments of the divine body. Now is this not just another way of 
saying that the great god is the synthesis of all the totems and 
consequently the personification of the tribal unity ? 

But at the same time, it takes an international character. 
In fact, the members of the tribe to which the young initiates 
belong are not the only ones who assist at the ceremonies of 
initiation ; representatives from the neighbouring tribes are 
specially summoned to these celebrations, which thus become 
sorts of international fairs, at once religious and laical. ^ Beliefs 
elaborated in social environments thus constituted could not 
remain the exclusive patrimony of any special nationality. 
The stranger to whom they are revealed carries them back to 
his own tribe when he returns home ; and as, sooner or later, 
he is forced to invite his former hosts, there is a continual exchange 
of ideas from tribe to tribe. It is thus that an international 
mythology was established, of which the great god was quite 
naturally the essential element, for it had its origin in the rites 

^ op. cit., p. 7. Among these same people, the principal wife of Baiame is 
also represented as the mother of all the totems, without belonging to any totem 
herself {ibid., pp. 7, 79). 

* See Howitt, Nat. Tr., pp. 511 f., 513. G02 ff. ; Mathews, /. 0/ R.S. of N.S. 
Wales, XXXVIII, p. 270. They invite to these feasts not only the tribes with 
whom a regular connnbium is established, but also those with whom there are 
quarrels to be arranged ; the vendetta, half-ceremonial and half-serious, take 
place on these occasions. 

The Idea of Spirits and Gods 295 

of initiation which it is his function to personify. So his name 
passed from one language to another, along with the representa- 
tions which were attached to it. The fact that the names of the 
phratries are generally the same in very different tribes could 
not fail to facilitate this diffusion. The internationalism of the 
totems opened the way for that of the great god. 

We thus reach the highest conception to which totemism 
has arrived. This is the point where it touches and prepares 
the religions which are to follow, and aids us in understanding 
them. But at the same time, we are able to see that this cul- 
minating idea is united without any interruption to the crudest 
beliefs which we analysed to start with. 

In fact, the great tribal god is only an ancestral spirit who ' 
finally won a pre-eminent place. The ancestral spirits are only 
entities forged in the image of the individual souls whose origin 
they are destined to explain. The souls, in their turn, are only 
the form taken by the impersonal forces which we found at the 
basis of totemism, as they individualize themselves in the human 
body. The unity of the system is as great as its complexity. _ j 

In this work of elaboration, the idea of the soul has undoubtedly 
played an important part : it is through it that the idea of 
personality has been introduced into the domain of religion. 
But it is not true that, as the theorists of animism maintain, 
it contains the germ of the whole religion. First of all, it pre- 
supposes the notion of tnana or the totemic principle of which 
it is only a special form. Then, if the spirits and gods could 
not be conceived before the soul, they are, nevertheless, more 
than mere human souls, liberated by death ; else whence would 
come their supernatural powers ? The idea of the soul has 
merely served to direct the mythological imagination in a new 
way and to suggest to it constructions of a new sort. But the 
matter for these conceptions has been taken, not from the 
representation of the soul, but from this reservoir of the anony- 
mous and diffused forces which constitute the original foundation 
of religions. The creation of mythical personalities has only 
been another way of thinking of these essential forces. 

As for the notion of the great god, it is due entirely to the 
sentiment whose action we have already observed in the genesis 
of the most specifically totemic beliefs : this is the tribal senti- 
ment. In fact, we have seen that totemism was not the work of 
isolated clans, but that it was always elaborated in the body of 
a tribe which was to some degree conscious of its unity. It is 

296 Eleme?itary Forms of Religious Life 

for this reason that the different cults peculiar to each clan 
mutually touch and complete each other in such a way as to 
form a unified whole. ^ Now it is this same sentiment of a tribal 
unity which is expressed in the conception of a supreme god, 
common to the tribe as a whole. So they are quite the same 
causes which are active at the bottom and at the top of this 
religious system. 

However, up to the present, we have considered the religious 
representations as if they were self-sufficient and could be ex- 
plained by themselves. But in reality, they are inseparable 
from the rites, not only because they manifest themselves there, 
but also because they, in their turn, feel the influence of these. 
Of course the cult depends upon the beliefs, but it also reacts 
upon them. So in order to understand them better, it is im- 
portant to understand it better. The moment has come for 
undertaking its study. 

^ See above, p. 155. 




WE do not have the intention of attempting a complete 
description of the primitive cult in what is to follow. 
Being preoccupied especially with reaching that which is most 
elementary and most fundamental in the religious life, we shall 
not attempt to reconstruct in detail the frequently confused 
multiplicity of all the ritual forms. But out of the midst of this 
extreme diversity of practices we should like to touch upon the 
most characteristic attitudes which the primitive observes in the 
celebration of his cult, to classify the most general forms of his 
rites, and to determine their origins and significance, in order 
that we may control and, if there is occasion, make more definite 
the results to which the analysis of the beliefs has led us.^ 

Every cult presents a double aspect, one negative, the other 
positive. In reality, of course, the two sorts of rites which we 
denominate thus are closely associated ; we shall see that they 
suppose one another. But still, they are different and, if it is 
only to understand their connection, it is necessary to distinguish 


By definition, sacred beings are separated beings. That which^ 
characterizes them is that there is a break of continuity between 
them and the profane beings. Normally, the first are outside 
the others. A whole group of rites has the object of realizing 
this state of separation which is essential. Since their function 
is to prevent undue mixings and to keep one of these two domains 
from encroaching upon the other, they are only able to impose 
abstentions or negative acts. Therefore, we propose to give the 
name negative cult to the system formed by these special rites. 
They do not prescribe certain acts to the faithful, but confine 
themselves to forbidding certain ways of acting ; so they all 

1 There is one form of ritual especially which we leave completely aside ; this 
is the oral ritual which must be studied in a special volume of the Collection de 
l'Année Sociologique. 


300 Elementary Forms of Religious Life 

take the form of interdictions, or as is commonly said by ethno- 
\graphers, of taboos. This latter word is the one used in the 
Polynesian languages to designate the institution in virtue of 
which certain things are withdrawn from common use^ ; it is 
also an adjective expressing the distinctive characteristic of 
these kinds of things. We have already had occasion to show 
how hard it is to translate a strictly local and dialectical expres- 
sion like this into a generic term. There is no religion where there 
are no interdictions and where they do not play a considerable 
part ; so it is regrettable that the consecrated terminology should 
seem to make so universal an institution into a peculiarity of 
Polynesia. 2 The expression interdicts or interdictions seems to 
us to be much more preferable. However, the word taboo, like 
the word totem, is so customary that it would show an excess of 
purism to prohibit it systematically ; also, the inconveniences 
it may have are attenuated when its real meaning and importance 
have once been definitely stated. 

But there are interdictions of different sorts which it is im- 
portant to distinguish ; for we shall not have to treat all kinds 
of interdictions in this chapter. 

First of all, beside those coming from religion, there are others 
which are due to magic. The two have this in common, that 
they declare certain things incompatible, and prescribe the 
separation of the things whose incompatibility is thus proclaimed. 
But there are also very grave differences between them. In the 
first place, the sanctions are not the same in the two cases. Of 
course the violation of the religious interdicts is frequently 
believed, as we shall presently see, to bring about material dis- 
orders mechanically, from which the guilty man will suffer, and 
which are regarded as a judgment on his act. But even if these 
really come about this spontaneous and automatic judgment is 
not the only one ; it is always completed by another one, sup- 
posing human intervention. A real punishment is added to 
this, if it does not anticipate it, and this one is deliberately 
inflicted by men ; or at least there is a blame and public repro- 
bation. Even when the sacrilege has been punished, as it were, 
by the sickness or natural death of its author, it is also defamed ; 
it offends opinion, which reacts against it ; it puts the man who 
did it in fault. On the contrary, the magic interdiction is judged 
only by the material consequences which the forbidden act is 

* See the article Taboo in the Encyclopœdia Britannica, written by Frazer. 

* Facts prove the reality of this inconvenience. There is no lack of writers 
who, putting their trust in the word, have believed that the institution thus 
designated was peculiar to primitive peoples in general, or even to the Polynesians 
(see Réville, Religion des peuples primitifs, II, p. 55 ; Richard, La Femme dans 
l'histoire, p. 435). 

The Negative Cult and its Functions 301 

believed to produce, with a sort of physical necessity. In dis- 
obeying, a man runs risks similar to those to which an invalid 
exposes himself in not following the advice of his physician ; 
but in this case disobedience is not a fault ; it creates no indig- 
nation. There is no sin in magic. Moreover, this difference in 
sanction is due to a profound difference in the nature of the 
interdictions. The religious interdiction necessarily implies the^ 
notion of sacredness ; it comes from the respect inspired by the 
sacred object, and its purpose is to keep this respect from failing, j 
On the other hand, the interdictions of magic suppose only a 
wholly lay notion of property. The things which the magician 
recommends to be kept separate are those which, by reason of 
their characteristic properties, cannot be brought together and 
confused without danger. Even if he happens to ask his clients 
to keep at a distance from certain sacred things, it is not through 
respect for them and fear that they may be profaned, for, as we 
know, magic lives on profanations ; ^ it is merely for reasons of 
temporal utility. In a word, religious interdictions are categorical 
imperatives ; others are useful maxims, the first form of hygienic 
and medical interdictions. We cannot study two orders of facts 
as different as these simultaneously, or even under the same 
name, without confusion. We are only concerned with the 
religious interdictions here.'^ 

But a new distinction is necessary between these latter. 

There are religious interdictions whose object is to separate 
two sacred things of different species from each other. For 
example, it will be remembered that among the Wakelbura the 
scaffold upon which the corpse is exposed must be made exclu- 
sively of materials belonging to the phratry of the dead man ; 
this is as much as to say that all contact between the corpse, 
which is sacred, and the things of the other phratry, which are 
also sacred, but differently, is forbidden. Elsewhere, the arms 
which one uses to hunt an animal with cannot be made out of a 
kind of wood that is classed in the same social group as the animal 
itself.^ But the most important of these interdictions are the 
ones which we shall study in the next chapter ; they are intended 
to prevent all communication between the purely sacred and 
the impurely sacred, between the sacredly auspicious and the 
sacredly inauspicious. All these interdictions have one common 

* See above, p. 43. 

* This is not saying that there is a radical break of continuity between the 
religious and the magic interdictions : on the contrary, it is one whose true 
nature is not decided. There are interdicts of folk-lore of which it is hard to say 
whether they are religious or magic. But their distinction is necessary, for we 
believe that the magic interdicts cannot be understood except as a function of the 
religious ones. 

^ See above, p. 149. 

302 Elementary Forms of Religious Life 

characteristic; they come, not from the fact that some things 
are sacred while others are not, but from the fact that there 
are inequalities and incompatibilities between sacred things. So 
they do not touch what is essential in the idea of sacredness. 
The observance of these prohibitions can give place only to 
isolated rites which are particular and almost exceptional ; but 
it could not make a real cult, for before all, a cult is made by 
regular relations between the profane and the sacred as such, 

f" But there is another system of religious interdictions which is 
much more extended and important ; this is the one which 
separates, not different species of sacred things, but all that is 
sacred from all that is profane. So it is derived immediately 
from the notion of sacredness itself, and it limits itself to express- 
ing and realizing this. Thus it furnishes the material for a verit- 
able cult, and even of a cult which is at the basis of all the others ; 
for the attitude which it prescribes is one from which the wor- 
shipper must never depart in all his relations with the sacred. 

^It is what we call the negative cult. We may say that its inter- 
dicts are the religious interdicts par excellence.'^ It is only these 
that we shall discuss in the following pages. 

But they take multiple forms. Here are the principal ones 
which we observe in Australia. 

Before all are the interdictions of contact ; these are the original 
taboos, of which the others are scarcely more than particular 
varieties. They rest upon the principle that the profane should 
never touch the sacred. We have seen already that the unin- 
itiated may not touch the churinga or the bull-roarers under any 
circumstances. If adults are allowed the free use of them, it is 
because initiation has conferred a sacred character upon them. 
Blood, and especially that which flows during the initiation, has 
a religious virtue ;2 it is under the same interdict. ^ It is the same 

^ Many of the interdictions between sacred things can be traced back, we 
think, to those between the sacred and the profane. Tliis is the case with the 
interdicts of age or rank. For example, in Australia, there are sacred foods 
which are reserved for the initiated. But these foods are not all sacred to the 
same degree ; there is a hierarchy among them. Nor are the initiated all equal. 
They do not enjoy all their religious rights from the first, but only enter step by 
step into the domain of religious things. They must pass through a whole series 
of ranks which are conferred upon them one after another, after special trials and 
ceremonies ; it requires months and sometimes even years to reach the highest 
rank. Now special foods are assigned to each of these ranks ; the men of the 
lower ranks may not touch the foods which rightfully belong to the men of the 
superior ones (see Mathews, Ethnol. Notes, etc., loc. cit. pp. 262 If. ; Parker, The 
Euahlayi, p. 23 ; Spencer and Gillen, Nov. Tr., pp. 611 ff. ; Nat. Tr., pp. 470 ff.). 
So the more sacred repels the less sacred ; but this is because the second is 
profane in relation to the first. In fine, all the interdictions arrange themselves 
in two classes : the interdictions between the sacred and the profane and the 
purely sacred and the impurely sacred. 

'' See above, p. 137. ' Spencer and Gillen, Nat. Tr., p. 463. 

The Negative Cult and its Functions 303 

\vith the hair.i A dead man is sacred because the soul which 
animated the body stays with the corpse ; for this reason it is 
sometimes forbidden to carry the bones of a dead man about 
unless they are wrapped up in a piece of bark.^ Even the place 
where the death took place should be avoided, for they believe 
that the soul of the dead man continues to haunt the spot. That 
is why they break camp and move some distance away ; ^ in 
certain cases they destroy it along with everything it contains,* 
and a certain time must elapse before they can come back to the 
same place. ^ Thus it comes about that a dying man creates an 
empty space about him ; they abandon him after they have 
installed him as comfortably as possible.* 

An exceptionally intimate contact is the one resulting from 
the absorption of food. Hence comes the interdiction against 
eating the sacred animals or vegetables, and especially those 
serving as totems.' Such an act appears so very sacrilegeous 
that the prohibition covers even adults, or at least, the majority 
of them ; only the old men attain a sufficient religious dignity 
to escape this interdict sometimes. This prohibition has some- 
times been explained by the mythical kinship uniting the man 
to the animals whose name he bears ; they are protected by the 
sentiment of sympathy which they inspire by their position as 
kin. 8 But the fact that the consumption of the forbidden flesh 
is believed to cause sickness or death automatically shows that 
this interdiction does not have its origin in the simple revolt of 
the feeling of domestic relationship. Forces of another sort are 
in action which are analogous to those in all religions and which 
are believed to react against sacrileges. 

Moreover, if certain foods are forbidden to the profane because 
they are sacred, certain others, on the contrary, are forbidden 
to persons of a sacred character, because they are profane. Thus 
it frequently happens that certain animals are specially desig- 
nated as the food of women ; for this reason, they believe that 
they partake of a feminine nature and that they are consequently 

1 Nat. Tv., p. 538 ; Nor. Tr., p. 640. ^ i^^ir. Tr., p. 531. 

* Nor. Tr., pp. 518 f. ; Howitt, Nat. Tr., p. 449. 

* Spencer and Gillen, Nat. Tr., p. 498 ; Schulze, loc. cit., p. 231. 

^ Spencer and Gillen, Nat. Tr., p. 499. * Howitt, Nat. Tr., p. 451. 

' If the alimentary interdictions which concern the totemic plant or vegetable 
are the most important, they are far from being the only ones. We have seen 
that there are foods which are forbidden to the non-initiated because they are 
sacred ; now very different causes may confer this character. For example, as 
we shall presently see, the birds which are seen on the tops of trees are reputed 
to be sacred, because they are neighbours to the great god who lives in heaven. 
Thus, it is possible that for different reasons the flesh of certain animals has been 
specially reserved for the old men and that consequently it has seemed to 
partake of the sacred character recognized in these latter. 

" Sec Frazer, Toteniism, p. 7. 

304 Elementary Forms of Religious Life 

profane. On the other hand, the young initiate is submitted to 
a series of rites of particular severity ; to give him the virtues 
which will enable him to enter into the world of sacred things, 
from which he had up till then been excluded, they centre an 
exceptionally powerful group of religious forces upon him. Thus 
he enters into a state of sanctity which keeps all that is profane 
at a distance. Then he is not allowed to eat the game which is 
regarded as the special food of women. ^ 

But contact may be established by other means than the touch. 
One comes into relations with a thing by merely regarding it : a 
look is a means of contact. This is why the sight of sacred 
things is forbidden to the profane in certain cases. A woman 
should never see the instruments of the cult ; the most that is 
permitted her is to catch a glimpse of them from afar.- It is 
the same with the totemic paintings executed on the bodies of 
the officiants in the exceptionally important ceremonies.^ The 
exceptional solemnity of the rites of initiation prevents the women 
in certain tribes from seeing the place where they were celebrated* 
or even the neophyte himself.^ The sacred character which is 
imminent in the ceremony as a whole is naturally found in the 
persons of those who directed it or took some part in it ; the 
result of this is that the novice may not raise his eyes to them, 
and this interdiction continues even after the rite is accomplished.* 
A dead man is also removed from view sometimes : his face is 
covered over in such a way that it cannot be seen.' 

The word is another way of entering into relations with persons 
or things. The breath expired establishes a communication ; 
this is a part of us which spreads outwards. Thus it is forbidden 
to the profane to address the sacred beings or simply to speak 
in their presence. Just as the neophyte must not regard either 
the operators or the assistants, so it is forbidden to him to con- 
verse with them except by signs ; and this interdiction keeps 
the place to which it has been raised, by means of a special rite.^ 

^ Howitt, Nat. Tr., p. 674. — There is one interdiction of contact of which we 
say nothing because it is very hard to determine its exact nature : this is sexual 
contact. 'Ihere are religious periods when a man cannot have commerce with a 
woman {Nor. Tr., pp. 293, 295 ; Nat. Tr., p. 397). Is this because the woman 
is profane or because the sexual act is dreaded ? This question cannot be decided 
in passing. We set it aside along with all that concerns conjugal and sexual rites. 
It is too closely connected with the problems of marriage and the family to be 
separated from them. 

^ Nat. Tr., p. 134 ; Howitt, Nat. Tr., p. 354. 

^ Spencer and Gillen, Nat. Tr., p. 624. 

* Howitt, Nat. Tr., p. 572. " Ibid., p. 661. 

« Nat. Tr., p. 386 ; Howitt, Nat. Tr., pp. 655, 665. 

' Among the Wiimbaio (Howitt, ibid., p. 451). 

" Howitt, ibid., pp. 624, 661, 663, 667 ; Spencer and Gillen, Nat. Tr., pp. 221, 
382 ff. ; Nor. Tr., pp. 335, 344, 353, 369. 

The Negative Cult and its Functions 305 

In a general way, there are, among the Arunta, moments in the 
course of the great ceremonies when silence is obligatory. ^ As 
soon as the churinga are exposed, every one keeps still, or if 
someone talks, he does so in a low voice or with his lips only.^ 

Besides the sacred things, there are words and sounds which 
have the same character ; they should not pass the lips of the 
profane or enter their ears. There are ritual songs which women 
must not hear under pain of death. ^ They may hear the noise of 
the bull-roarers, but only from a distance. Every proper name is 
considered an essential element Gf the person who bears it ; being 
closely associated in the mind to the idea of this person, it partici- 
pates in the sentiments which this latter inspires. So if the one is 
sacred, the other is. Therefore, it may not be pronounced in the 
course of the profane life. Among the Warramunga there is one 
totem which is particularly venerated, this is the snake called 
WoUunqua ; its name is taboo. ^ It is the same with Baiame, 
Daramulun and Bun j il ; the esoteric form of their name must 
not be revealed to the uninitiate.^ During mourning, the name 
of the dead man must not be mentioned, at least by his parents, 
except when there is an absolute necessity, and even in this case 
it must be whispered.* This interdiction is frequently perpetual 
for the widow and certain relatives.' Among certain peoples, 
this even extends beyond the family ; all the individuals whose 
name is the same as that of the dead man must change theirs 
temporarily.^ But there is more than this : the relatives and 
intimate friends sometimes abstain from certain words in the 
usual language, undoubtedly because they were employed by 
the dead man ; these gaps are filled in by means of periphrases 
or words taken from some foreign dialects.® In addition to their 
public and everyday names all men have another which is kept 
a secret : the women and children do not know it ; it is never 
used in the ordinary life. This is because it has a religious 
character.^" There are even ceremonies during which it is neces- 
sary to speak a special language which must not be used for 
profane purposes. It is the beginning of a sacred language. ^^ 

Not only are the sacred beings separated from the profane, l 
but also nothing which either directly or indirectly concerns the 

* Spencer and Gillcn, Nat. Tr., pp. 221, 262, 288, 303, 378, 380. 
- Ibid., p. 302. » Howitt, Nat. Tr., p. 581. 

* Noy. Tr., p. 227. * See above, p. 288. 

« Spencer and Gillen, Nat. Tr., p. 498 ; Nor. Tr., p. 526 ; Taplin, Narrin- 
yeri, p. 19. 

' Howitt, Nat. Tr., pp. 466, 469 ff. 

* Wyatt, Adelaide and Encounter Bay Tribes, in Woods, p. 165. 
' Howitt, Nat. Tr., p. 470. 

*" Ibid., p. 657 ; Spencer and Gillen, Nat. Tr., p. 139 ; ^or. Tr., pp. 580 flf. 
*' Howitt, Nat. Tr., p. 537. 

3o6 Elementary Forms of Religious Life 

profane life should be confused with the religious life. Complete 
nudity is frequently demanded of the native as a prerequisite 
to being admitted to participation in the rites ;i he is required 
to strip himself of all his habitual ornaments, even those to which 
he is the most attached, and from which he separates himself the 
least willingly because of the protecting virtues he attributes 
to them.'-^ If he is obliged to decorate himself to play his part 
in the ritual, this decoration has to be made specially for the 
occasion ; it is a ceremonial costume, a gala dress. ^ As these 
ornaments are sacred, owing to the use made of them, he is 
forbidden to use them in profane affairs ; when the ceremony 
is finished, they are buried or burnt ;'* the men must even wash 
themselves in such a way as to carry away with them no trace 

F of the decorations with which they were adorned.^ 
In general, all acts characteristic of the ordinary life are for- 
bidden while those of the religious life are taking place. The 
act of eating is, of itself, profane ; for it takes place every day, it 
satisfies essentially utilitarian and material needs and it is a part 
of our ordinary existence. ^ This is why it is prohibited in rehgious 
J times. When one totemic group has loaned its churinga to a 
foreign clan, it is an exceptionally solemn moment when they are 
brought back and put into the ertnatulunga ; all those who take 
part in the ceremony must fast as long as it lasts, and it lasts a 
long time.' The same rule is observed during the rites, ^ of which 
we shall speak in the next chapter, as well as at certain moments 
of the initiation.^ 

For this same reason, all temporal occupations are suspended 
while the great religious solemnities are taking place. According 
to a remark of Spencer and Gillen,^'' which we have already had 
occasion to cite, the life of the Australian is divided into two 
very distinct parts : the one is devoted to hunting, fishing and 
warfare ; the other is consecrated to the cult, and these two forms 

^ Howitt, Nat. Tr., pp. 544, 597, 614, G20. 

* For example, the hair belt which he ordinarily wears (Spencer and Gillen, 
Nat. Tr., p. 171). 

« Ibid., p. 624 ff. * Howitt, Nat. Tr., p. 556. * Ibid., p. 5S7. 

* This act takes on a sacred character, it is true, when the elements eaten are 
sacred. But in itself, the act is so very profane that eating a sacred food always 
constitutes a profanation. The profanation may be permitted or even ordered, 
but, as we shall see below, only on condition that rites attenuating or expiating 
it precede or accompany it. The existence of these rites shows that, by itself, the 
sacred thing should not be eaten. 

^ Nor. Tr., p. 263. « Spencer and Gillen, Nat. Tr., p. 171. 

» Howitt, Nat. Tr., p. 674. Perhaps the rule against talking during the great 
religious solemnities is due to the same cause. Men speak, and especially in a 
high voice, during ordinary life ; then, in the religious life they ought to keep 
still or talk in a low voice. This same consideration is not foreign to the 
alimentary interdictions (see above, p. 128/. 

*" A'or. Tr., p. 33. 

The Negative Cult and its Functions 307 

of activity mutually exclude and repel one another. It is on 
this principle that the universal institution of religious days of. 
rest reposes. The distinctive character of the feast-days in all 
known religions is the cessation of work and the suspension of 
public and private life, in so far as it does not have a religious 
objective. This repose is not merely a sort of temporary relaxa- 
tion which men have given themselves in order to give themselves 
up more freely to the sentiments of joy ordinarily awakened by 
the feast-days ; for they are sad feasts, consecrated to mourning 
and repentance, and during which this cessation is no less obliga- 
tory. This is because work is an eminent form of profane activity : 
it has no other apparent end than to provide for the temporal 
necessities of life ; it puts us in relations with ordinary things 
only. On feast days, on the contrary, the religious life attains 
an exceptional degree of intensity. So the contrast between the 
two forms of existence is especially marked at this moment ; 
consequently, they cannot remain near to each other. A man 
cannot approach his god intimately while he still bears on him 
marks of his profane life ; inversely, he cannot return to his usual 
occupations when a rite has just sanctified him. So the ritual 
day of rest is only one particular case of the general incompati- 
bility separating the sacred from th» profane ; it is the result 
of an interdiction. z^ 

It would be impossible to enumerate here all the different 1 
interdictions which have been observed, even in the Australian 
religions alone. Like the notion of sacredness upon which it 
rests, the system of interdicts extends into the most diverse 
relations ; it is even used deliberately for utilitarian ends. ^ 

^ Since there is a sacred principle, the sonl, within each man, from the very 
first, the individual is surrounded by interdicts, the original form of the moral 
interdicts which isolate and protect the human person to-day. Thus the corpse 
of his victim is considered dangerous for a murderer (Spencer and Gillen, Nat. Tr., 
p. 492), and is taboo for him. Now the interdicts having this origin are frequently 
used by individuals as a means of withdrawing certain things from common use 
and thus establishing a property right over them. " When a man goes away 
from the camp, leaving his arms and food there," says Roth, speaking of the 
tribes on the Palmer River (North Queensland), " if he urinates near the objects 
he leaves, they become tami (equivalent to taboo) and he may be sure of finding 
them intact on his return " {North Queensland Ethnography, in Records of the 
Australian Museum, Vol. VII, No. 2, p. 75). Tliis is because the urine, like the 
blood, is believed to contain some of the sacred force vvhich is personal to the 
individual. So it keeps strangers at a distance. For the same reasons, the spoken 
word may also serve as a vehicle for these same influences ; that is how it becomes 
possible to prevent access to an object by a mere verbal declaration. This power 
of making interdicts varies with different individuals ; it is greater as their 
character is more sacred. Men have this privilege almost to the exclusion of 
women (Roth cites one single case of a taboo imposed by women) ; it is at its 
ma.ximum with the chiefs and old men, who use it to monopolize whatever things 
they tind it convenient to (Roth, ibid., p. 77) . Thus the religious interdict becomes 
a right of property and an administrative rule. 

3o8 Elementary Forms of Religious Life 

|But howsoever complex it may be, it finally rests upon two 
fundamental interdictions, which summarize it and dominate it. 
In the first place, the religious life and the profane life cannot 
coexist in the same place. If the former is to develop, a special 
spot must be placed at its disposition, from which the second is 
excluded. Hence comes the founding of temples and sanctuaries : 
these are the spots awarded to sacred beings and things and serve 
them as residences, for they cannot establish themselves in any 
place except on the condition of entirely appropriating to them- 
selves all within a certain distance. Such arrangements are so 
indispensable to all religious life that even the most inferior 

Lj-eligions cannot do without them. The ertnatulunga, the spot 
where the churinga are deposited, is a veritable sanctuary. So 
the uninitiated are not allowed to approach it. It is even for- 
bidden to carry on any profane occupation whatsoever there. 
As we shall presently see, there are other holy places where 
important ceremonies are celebrated.^ 

r^ Likewise, the religious life and the profane life cannot coexist 
in the same unit of time. It is necessary to assign determined 
days or periods to the first, from which all profane occupations 
are excluded. Thus feast days are born. There is no religion, 
and, consequently, no society which has not known and practised 
this division of time into two distinct parts, alternating with 
one another according to a law varying with the peoples and the 
civilizations ; as we have already pointed out, it was probably 
the necessity of this alternation which led men to introduce into 
the continuity and homogeneity of duration, certain distinctions 
and differentiations which it does not naturally have.^ Of 
course, it is almost impossible that the religious life should 
ever succeed in concentrating itself hermetically in the places 
and times which are thus attributed to it ; it is inevitable that 
a little of it should filter out. There are always some sacred 
things outside the sanctuaries ; there are some rites that can 
be celebrated on work-days. But these are sacred things of the 
second rank and rites of a lesser importance. Concentration 
remains the dominating characteristic of this organization. 
Generally this concentration is complete for all that concerns 
the public cult,, which cannot be celebrated except in common. 
The individual, private cult is the only one which comes very 
near to the temporal life. Thus the contrast between these 
two successive phases of human life attains its maximum of 
intensity in the inferior societies ; for it is there that the in- 
l dividual cult is the most rudimentary.^ 

^ See below, this book, ch. ii. * See above, p. lo. 

^ See above, p. 219. 

The Negative Cult and its Functions 309 


Up to the present, the negative cult has been presented to us 
only as a system of abstentions. So it seems to serve only to 
inhibit activity, and not to stimulate it or to modify it. And 
yet, as an unexpected reaction to this inhibitive effect, it is 
found to exercise a positive action of the highest importance 
over the religious and moral nature of the individual. 

In fact, owing to the barrier which separates the sacred ' 
from the profane, a man cannot enter into intimate relations 
with sacred things except after ridding himself of all that is 
profane in him. He cannot lead a religious life of even a slight 
intensity unless he commences by withdrawing more or less 
completely from the temporal life. So the negative cult is in one 
sense a means in view of an end : it is a condition of access to the 
positive cult. It does not confine itself to protecting sacred 
beings from vulgar contact ; it acts upon the worshipper himself 
and modifies his condition positively. The man who has sub- 
mitted himself to its prescribed interdictions is not the same 
afterwards as he was before. Before, he was an ordinary being 
who, for this reason, had to keep at a distance from the religious 
forces. Afterwards, he is on a more equal footing with them ; 
he has approached the sacred by the very act of leaving the 
profane ; he has purified and sanctified himself by the very 
act of detaching himself from the base and trivial matters that 
debased his nature. So the negative rites confer efficient powers 
just as well as the positive ones ; the first, like the second, can 
serve to elevate the religious tone of the individual. x\ccording to 
a very true remark which has been made, no one can engage in 
a religious ceremony of any importance without first submitting 
himself to a sort of preliminary initiation which introduces 
him progressively into the sacred world. ^ Unctions, lustrations, 
benedictions or any essentially positive operation may be used 
for this purpose ; but the same result may be attained by means 
of fasts and vigils or retreat and silence, that is to say, by ritual 
abstinences, which are nothing more than certain interdictions , 
put into practice. ^ 

When there are only particular and isolated negative rites, 
their positive action is generally too slight to be easily perceptible. 
But there are circumstances when a whole system of interdictions 
is concentrated on one man ; in these cases, their effects accumu- 
late, and thus become more manifest. This takes place in Australia 
at the time of the initiation. The neophyte is submitted to a 

^ See Hubert and Mauss, Essai sur la nature et la fonction du sacrifice, in 
Mélanges d'histoire des religions, pp. 22 û. 

3IO Elementary Forms of Religious Life 

great variety of negative rites. He must withdraw from the 
society in which his existence has been passed up till then, and 
from almost all human society. Not only is it forbidden for him 
to see women and uninitiated persons, ^ but he also goes to live 
in the brush, far from his fellows, under the direction of some 
old men who serve him as godfathers. ^ So very true is it that 
the forest is considered his natural environment, that in a certain 
number of tribes, the word with which the initiation is designated 
signifies that which is from the forest.^ For this same reason, he 
is frequently decorated with leaves during the ceremonies at 
which he assists.^ In this way he passes long months,^ inter- 
spersed from time to time with rites in which he must take a 
part. This time is a period of all sorts of abstinences for him. 
A multitude of foods are forbidden him ; he is allowed only 
that quantity of food which is absolutely indispensable for the 
maintenance of life ; ® he is even sometimes bound to a rigorous 
fast,' or must eat impure foods. ^ When he eats, he must not 
touch the food with his hands ; his godfathers put it into his 
mouth for him.^ In some cases, he must go to beg his food.^^ 
Likewise, he sleeps only as much as is indispensable. ^^ He 
must abstain from talking, to the extent of not uttering a word ; 
it is by signs that he makes known his needs. ^^ He must not 
wash ; ^^ sometimes he must not move. He remains stretched 
out upon the earth, immobile ^^ and without clothing of any 
sort.^^ Now the result of the numerous interdictions is to bring 
about a radical change of condition in the initiate. Before the 
initiation, he lived with the women ; he was excluded from the 
cult. After it, he is admitted to the society of men ; he takes 
part in the rites, and has acquired a sacred character. The meta- 
morphosis is so complete that it is sometimes represented as a 
second birth. They imagine that the profane person, who was 
the young man up till then, has died, that he has been killed and 
carried away by the god of the initiation, Bun j il, Baiame or 

1 Howitt, Nat. Tr., pp. 560, 657, 659, 661. Even the shadow of a woman must 
not fall upon him (ibid., p. 633). Whatever he has touched must not be touched 
by a woman (ibid., p. 621). 

* Ibid., pp. 561, 563, 670 f. ; Spencer and Gillen, Nat. Tr., p. 223 ; Nov. Tr., 

PP- 340. 342- 

' The word Jeraeil, for example, among the Kumai, or Kuringal among the 
Yuin and Wolgal (Howitt, Nat. Tr., pp. 518, 617). 

* Spencer and Gillen, Nat. Tr., p. 348. * Howitt, p. 561. 

* Howitt, pp. 633, 538, 560. ' Ibid., p. 674 ; Parker, Euahlayi, p. 75. 
^ Ridley, Kamilaroi, p. 154. • Howitt, p. 563. 

1° Ibid., p. 611. ^1 Ibid., pp. 549, 674. 

** Howitt, Nat. Tr., pp. 580, 596, 604, 6t)8, 670 ; Spencer and Gillen, Nat. Tr., 
pp. 223. 351. 

19 Howitt, p. 557. 

1* Ibid., p. 604 ; Spencer and Gillen, Nat. Tr., p. 351. 

** Howitt, p. 611. 

The Negative Cult and its Functions 311 

Daramulun, and that quite another individual has taken the 
place of the one that no longer is.^ So here we ônd the very- 
heart of the positive effects of which negative rites are capable. 
Of course we do not mean to say that these latter produced this 
great transformation all by themselves ; but they certainly 
contributed to it, and largely. 

In the light of these facts, we are able to understand what 
asceticism is, what place it occupies in the religious life and 
whence come the virtues which have generally been attributed 
to it. In fact, there is no interdict, the observance of which 
does not have an ascetic character to a certain degree. Abstaining 
from something which may be useful or from a form of activity 
which, since it is usual, should answer to some human need, 
is, of necessity, imposing constraints and renunciations. So in 
order to have real asceticism, it is sufficient for these practices 
to develop in such a way as to become the basis of a veritable 
scheme of life. Normally, the negative cult serves only as an 
introduction and preparation for the positive cult. But it 
sometimes happens that it frees itself from this subordination 
and passes to the first place, and that the system of interdicts 
swells and exaggerates itself to the point of usurping the entire 
existence. Thus a systematic asceticism is born which is con- 
sequently nothing more than a hypertrophy of the negative cult. 
The special virtues which it is believed to confer are only an 
amphfied form of those conferred, to a lesser degree, by the 
practice of any interdiction. They have the same origin ; for 
they both rest on the principle that a man sanctifies himself only 
by efforts made to separate himself from the profane. The 
pure ascetic is a man who raises himself above men and acquires 
a special sanctity by fasts and vigils, by retreat and silence, or in 
a word, by privations, rather than by acts of positive piety 
(offerings, sacrifices, prayers, etc.). History shows to what 
a high religious prestige one may attain by this method : the 
Buddhist saint is essentially an ascetic, and he is equal or superior 
to the gods. 

It follows that asceticism is not a rare, exceptional and nearly 
abnormal fruit of the religious life, as some have supposed it to 
be ; on the contrary, it is one of its essential elements. Every 
religion contains it, at least in germ, for there are none in which 
a system of interdicts is not found. Their only difference in this 
regard which there may be between cults is that this germ is 
more or less developed in different ones. It should also be added 
that there probably is not a single one in which this development 
does not take, at least temporarily, the characteristic traits of 

* Howitt, p. 589. 

31^ Elementary Forms of Religious Life 

real asceticism This is what generally takes place at certain 
critical periods when, for a relatively short time, it is necessary 
to bring about a grave change of condition in a subject. Then, 
in order to introduce him more rapidly into the circle of sacred 
things with which he must be put in contact, he is separated 
violently from the profane world ; but this does not come 
without many abstinences and an exceptional recrudescence 
of the system of interdicts. Now this is just what happens in 
Australia at the moment of initiation. In order to transform 
youths into men, it is necessary to make them live the life of a 
veritable ascetic. Mrs, Parker very justly calls them the monks 
of Baiame.^ 

But abstinences and privations do not come without suffering. 
We hold to the profane world by all the fibres of our flesh ; 
our senses attach us to it ; our life depends upon it. It is not 
merely the natural theatre of our activity ; it penetrates us 
from every side ; it is a part of ourselves. So we cannot detach 
ourselves from it without doing violence to our nature and 
without painfully wounding our instincts. In other words, the 
negative cult cannot develop without causing suffering. Pain is one 
of its necessary conditions. Some have been led to think of it as 
constituting a sort of rite in itself ; they have seen in it a state 
of grace which is to be sought and aroused, even artificially, 
because of the powers and privileges which it confers in the same 
way as these systems of interdicts, of which it is the natural 
accompaniment. So far as we know, Preuss is the first who has 
realized the religious rôle ^ which is attributed to suffering in the 
inferior societies. He cites the case of the Arapahs who inflict 
veritable torments upon themselves in order to become immune 

^ One may compare these ascetic practices with those used at the initiation of 
a magician. Just like the young neophyte, the apprentice magician is submitted 
to a multitude of interdictions, the observation of which contributes to his 
acquisition of his specific powers (see L'Origine des pouvoirs magiques, in Hubert 
and Mauss, Mélanges d'histoire des religions, pp. 171, 173, 176). The same is true 
for the husband and wife on the day before and the day after the wedding (taboos 
of the betrothed and newly married) ; this is because marriage also implies a 
grave change of condition. We limit ourselves to mentioning these facts 
summarily, without stopping over them ; for the first concern magic, which 
is not our subject, and the second have to do with that system of juridico- 
rcligious rules which relates to the commerce of the sexes, the study of which will 
be possible only in conjunction with the other precepts of primitive conjugal 

* It is true that Preuss interprets these facts by saying that suffering is a way 
of increasing a man's magic force {die menschliche Zauberkraft) ; from this 
expression, one might believe that suffering is a magic rite, not a religious one. 
But as we have already pointed out, Preuss gives the name magic, without great 
precision, to all anonymous and impersonal forces, whether they belong to magic 
or religion. Of course, there are tortures which are used to make magicians ; but 
many of those which we have described are a part of the real religious ceremonies, 
and, consequently, it is the religious state of the individuals which they modify. 

The Negative Cult and its Functions 313 

from the dangers of battle ; of the Big Belly Indians who submit 
to actual tortures on the eve of military expeditions ; of the 
Hupa who swim in icy rivers and then remain stretched out on 
the bank as long as possible, in order to assure themselves of 
success in their enterprises ; of the Karaya who from time to 
time draw blood from their arms and legs by means of scratches 
made out of the teeth of fish, in order to strengthen their muscles ; 
of the men of Dallmannhafen (Emperor William's Land in New 
Guinea) who combat the sterility of their women by making 
bloody incisions in the upper part of their thighs. ^ 

But similar facts may be found without leaving Australia, 
especially in the course of the initiation ceremonies. Many 
of the rites practised on this occasion consist in systematically 
inflicting certain pains on the neophyte in order to modify his 
condition and to make him acquire the qualities characteristic 
of a man. Thus, among the Larakia, while the young men are 
in retreat in the forest, their godfathers and guardians give 
them violent blows at any instant, without warning and without 
cause. 2 Among the Urabunna, at a certain time, the novice is 
stretched out on the ground, his face against the earth. All the 
men present beat him rudely ; then they make four or eight 
gashes on his back, arranged on each side of the dorsal spine 
and one on the meridial line of the nape of his neck.^ Among 
the Arunta, the first rite of the initiation consists in tossing the 
subject in a blanket ; the men throw him into the air and catch 
him when he comes down, to throw him up again.* In the same 
tribe, at the close of this long series of ceremonies, the young 
man lies down on a bed of leaves under which they have placed 

* Preuss, Der Ursprung der Religion und Kunst, in Globus, LXXXVIII, 
pp. 309-400. Under this same rubric Preuss classes a great number of incon- 
gruous rites, for example, effusions of blood which act in virtue of the positive 
qualities attributed to blood and not because of the sufiering which they imply. 
We retain only those in which suffering is an essential element of the rite and the 
cause of its efficacy. 

' Nor. Tr., pp. 331 f. 

' Ibid., p. 335. A similar practice will be found among the Dieri (Howitt, 
Nat. Tr., pp. 658 fi.). 

* Spencer and Gillen, Nat. Tr., pp. 214 fi. — From this example we see that the 
rites of initiation sometimes have all the characteristics of hazing. In fact, 
hazing is a real social institution which arises spontaneously every time that 
two groups, inequal in their moral and social situation, come into intimate 
contact. In this case, the one considering itself superior to the other resists the 
mtrusion of the new-comers ; it reacts against them is such a way as to make 
them aware of the superiority it feels. This reaction, which is produced auto- 
matically and which takes the form of more or less grave cruelties quite naturally, 
is also destined to shape the individuals for their new existence and assimilate 
them into their new environment. So it is a sort of initiation. Thus it is 
explained how the initiation, on its side, takes the form of hazing. It is because 
the group of old men is superior in religious and moral dignity to that of the 
young men, and yet the first must assimilate the second. So all the conditions for 
hazing are given. 

314 Elementary Forms of Religious Life 

live coals ; he remains there, immobile in the midst of the heat 
and suffocating smoke. ^ A similar rite is observed among the 
Urabunna ; but in addition, while the patient is in this painful 
situation, they beat him on the back.^ In a general way, all the 
exercises to which he is submitted have this same character to 
such an extent that when he is allowed to re-enter the ordinary 
life, he has a pitiful aspect and appears half stupefied.* It is 
true that all these practices are frequently represented as ordeals 
destined to prove the value of the neophyte and to show whether 
he is worthy of being admitted into the religious society or not.^ 
But in reality, the probational function of the rite is only another 
aspect of its efficacy. For the fact that it has been undergone 
is proved by its producing its effect, that is to say, by its con- 
ferring the qualities which are the original reason for its existence. 

In other cases, these ritual cruelties are executed, not on the 
organism as a whole, but on a particular organ or tissue, whose 
vitality it is their object to stimulate. Thus, among the Arunta, 
the Warramunga and many other tribes,^ at a certain moment 
in the initiation, certain persons are charged with biting the 
novice severely in the scalp. This operation is so painful that 
the patient can hardly support it without uttering cries. Its 
object is to make the hair grow.' The same treatment is applied 
to make the beard grow. The rite of pulling out hairs, which 
Howitt mentions in other tribes, seems to have the same reason 
for existence.' According to Eylmann, the men and women of 
the Arunta and the Kaitish make small wounds on their arms 
with sticks red with fire, in order to become skilful in making 
fire or to acquire the strength necessary for carrying heavy loads 
of wood.^ According to this same observer, the Warramunga 
girls amputate the second and third joints of the index finger 
on one hand, thinking that the finger thus becomes better fitted 
for finding yams.^ 

It is not impossible that the extraction of teeth was sometimes 
destined to produce effects of this sort. In any case, it is certain 
that the cruel rites of circumcision and subincision have the 
object of conferring particular-powers on the genital organs. 
In fact, the young man is not allowed to marry until after he has 
undergone them ; so he owes them special virtues. What makes 

* Spencer and Gillen, Nat. Tr., p. 372. 

* Ibid., p. 335. s Howitt, Nat. Tr.. p. 675. 

* Howitt, Nat. Tr., pp. 569, 604. 

* Spencer and Gillen, Nat. Tr., p. 251 ; Nor. Tr., 341, 352. 

* Among the Warramunga, the operation must be made by persons favoured 
with beautiful hair. 

' Howitt, Nat. Tr., p. 675 ; this concerns the tribes on the lower Darling. 

* Eylmann, op. cit., p. 212. » Ibid. 

The Negative Cult and its Functions ÇSt 

this initiation sui generis indispensable is that in all inferior 
societies, the union of the sexes is marked with a religious charac- 
ter. It is believed to put redoubtable forces into play which 
a man cannot approach without danger, until after he has acquired 
the necessary immunity, by ritual processes : ^ for this, a whole 
series of positive and negative practices is used, of which circum- 
cision and subincision are the forerunners. By painfully mutilating 
an organ, a sacred character is given to it, since by that act, it is 
put into shape for resisting the equally sacred forces which it 
could not meet otherwise. 

At the beginning of this work, we said that all the essential 
elements of religious thought and life ought to be found, at least 
in germ, in the most primitive religions : the preceding facts 
confirm this assertion. If there is any one belief which is believed 
to be peculiar to the most recent and idealistic religions, it is the 
one attributing a sanctifying power to sorrow. Now this same 
belief is at the basis of the rites which have just been observed. 
Of course, it is understood differently at the different moments 
of history when it is studied. For the Christian, it acts especially 
upon the soul : it purges it, ennobles it, spiritualizes it. For the 
Australian, it is the body over which it is efficient : it increases 
its vital energies ; it makes its beard and hair grow ; it toughens 
its members. But in both cases the principle is the same. In 
both it is admitted that suffering creates exceptional strength. 
And this belief is not without foundation. In fact, it is by the 
way in which he braves suffering that the greatness of a man 
is best manifested. He never rises above himself ^vith more 
brilliancy than when he subdues his own nature to the point of 
making it follow a way contrary to the one it would spontaneously 
take. By this, he distinguishes himself from all the other creatures 
who follow blindly wherever pleasure calls them ; by this, he 
makes a place apart for himself in the world. Suffering is the 
sign that certain of the bonds attaching him to his profane 
environment are broken ; so it testifies that he is partially 
freed from this environment, and, consequently, it is justly 
considered the instrument of deliverance. So he who is thus 
delivered is not the victim of a pure illusion when he believes 
himself invested with a sort of mastery over things : he really 
has raised himself above them, by the very act of renouncing 
them ; he is stronger than nature, because he makes it 

Moreover, it is by no means true that this virtue has only an 

• References on this question will be found in our memoir on La Prohibition 
de I'incest et ses origines (Année Social., I, pp. i If.), and Crawley, The Mystic Rose, 
pp. 37 fi. 

3i6 Elementary Forms of Religious Life 

aesthetic value : the whole religious life supposes it. Sacrifices 
and privations do not come without privations which cost the 
worshipper dear. Even if the rites do not demand material 
gifts from him, they require his time and his strength. In order 
to serve his gods, he must forget himself ; to make for them a 
fitting place in his own life, he must sacrifice his profane interests. 
The positive cult is possible only when a man is trained to 
renouncement, to abnegation, to detachment from self, and 
consequently to suffering. It is necessary that he have no dread 
of them : he cannot even fulfil his duties joyfully unless he loves 
them to some extent. But for that, it is necessary that he train 
himself, and it is to this that the ascetic practices tend. So the 
suffering which they impose is not arbitrary and sterile cruelty ; 
it is a necessary school, where men form and temper themselves, 
and acquire the qualities of disinterestedness and endurance 
without which there would be no religion. If this result is to 
be obtained, it is even a good thing that the ascetic ideal be 
incarnated eminently in certain persons, whose speciality, so 
to speak, it is to represent, almost with excess, this aspect of 
the ritual life ; for they are like so many living models, inciting 
to effort. Such is the historic rôle of the great ascetics. When 
their deeds and acts are analysed in detail, one asks himself 
what useful end they can have. He is struck by the fact that 
there is something excessive in the disdain they profess for all 
that ordinarily impassions men. But these exaggerations are 
necessary to sustain among the believers a sufficient disgust for 
an easy life and common pleasures. It is necessary that an elite 
put the end too high, if the crowd is not to put it too low. It is 
necessary that some exaggerate, if the average is to remain at 
a fitting level. 

But asceticism does not serve religious ends only. Here, as 
elsewhere, religious interests are only the symbolic form of 
social and moral interests. The ideal beings to whom the cults 
are addressed are not the only ones who demand of their followers 
a certain disdain for suffering : society itself is possible only 
at this price. Though exalting the strength of man, it is frequently 
rude to individuals ; it necessarily demands perpetual sacrifices 
from them ; it is constantly doing violence to our natural appe- 
tites, just because it raises us above ourselves. If we are going 
to fulfil our duties towards it, then we must be prepared to do 
violence to our instincts sometimes and to ascend the decline 
of nature when it is necessary. So there is an asceticism which, 
being inherent in all social life, is destined to survive all the 
mythologies and all the dogmas ; it is an integral part of all 
human culture. At bottom, this is the asceticism which is the 

The Negative Cult and its Functions 317 

reason for the existence of and the justification of that which 
has been taught by the rehgions of all times. 


Having determined what the system of interdicts consists in 
and v/hat its positive and negative functions are, we must now 
seek the causes which have given it birth. 

In one sense, it is logically implied in the very notion of sacred- 
ness. All that is sacred is the object of respect, and every 
sentiment of respect is translated, in him who feels it, by move- 
ments of inhibition. In fact, a respected being is always expressed 
in the consciousness by a representation which, owing to the 
emotion it inspires, is charged with a high mental energy ; con- 
sequently, it is armed in such a way as to reject to a distance 
every other representation which denies it in whole or in part. 
Now the sacred world and the profane world are antagonistic to 
each other. They correspond to two forms of life which mutually 
exclude one another, or which at least cannot be lived at the 
same time with the same intensity. We cannot give ourselves 
up entirely to the ideal beings to whom the cult is addressed 
and also to ourselves and our own interests at the same time ; 
we cannot devote ourselves entirely to the group and entirely 
to our own egoism at once. Here there are two systems of 
conscious states which are directed and which direct our conduct 
towards opposite poles. So the one having the greater power of 
action should tend to exclude the other from the consciousness. 
When we think of holy things, the idea of a profane object cannot 
enter the mind without encountering grave resistance ; something 
within us opposes itself to its installation. This is because the 
representation of a sacred thing does not tolerate neighbours. 
But this psychic antagonism and this mutual exclusion of ideas 
should naturally result in the exclusion of the corresponding 
things. If the ideas are not to coexist, the things must not 
touch each other or have any sort of relations. This is the very 
principle of the interdict. -J 

Moreover, the world of sacred things is, by definition, a world « 
apart. Since it is opposed to the profane world by all the charac- 
teristics we have mentioned, it must be treated in its own peculiar 
way : it would be a misunderstanding of its nature and a con- 
fusion of it with something that it is not, to make use of the 
gestures, language and attitudes which we employ in our re- 
lations with ordinary things, when we have to do with the things 
that compose it. We may handle the former freely ; we speak 
freely to vulgar beings ; so we do not touch the sacred beings. 

3i8 Elementary Forms of Religious Life 

or we touch them only with reserve ; we do not speak in their 
presence, or we do not speak the common language there. All 
that is used in our commerce with the one must be excluded 

l__from our commerce with the other. 

r But if this explanation is not inexact, it is, nevertheless, 
insufficient. In fact, there are many beings which are the 
objects of respect without being protected by systems of rigorous 
interdictions such as those we have just described. Of course 
there is a general tendency of the mind to localize different things 
in different places, especially when they are incompatible with 
each other. But the profane environment and the sacred one 
are not merely distinct, but they are also closed to one another ; 
between them there is an abyss. So there ought to be some 
particular reason in the nature of sacred things, which causes 
this exceptional isolation and mutual exclusion. And, in fact, 
by a sort of contradiction, the sacred world is inclined, as it 
were, to spread itself into this same profane world which it 
excludes elsewhere : at the same time that it repels it, it tends to 
flow into it as soon as it approaches. This is why it is necessary 
to keep them at a distance from one another and to create a 

^sort of vacuum between them. 

r What makes these precautions necessary is the extraordinary 
contagiousness of a sacred character. Far from being attached 
to the things which are marked with it, it is endowed with a 
sort of elusiveness. Even the most superficial or roundabout 
contact is sufficient to enable it to spread from one object to 
another. Religious forces are represented in the mind in such 
a way that they always seem ready to escape from the points 
where they reside and to enter everything passing within their 

grange. The nanja tree where the spirit of an ancestor lives is 
sacred for the individual who considers himself the reincarnation 
of this ancestor. But every bird which alights upon this tree 
participates in this same nature : it is also forbidden to touch 
it.i We have already had occasion to show how simple contact 
with a churinga is enough to sanctify men and things ; ^ it is 
also upon this principle of the contagiousness of sacredness 
that all the rites of consecration repose. The sanctity of the 
churinga is so great that its action is even felt at a distance. 
It will be remembered how this extends not only to the cave 
where they are kept, but also to the whole surrounding district, 
to the animals who take refuge there, whom it is forbidden to 
kill, and to the plants which grow there, which must not be 
touched. 3 A snake totem has its centre at a place where there 

^ Spencer and Gillen, Nat. Tr., p. 133. ^ See above, p. 121. 

^ Spencer and Gillen, Nat. Tr.. pp. 134 f. ; Strehlow, I. p. 78. 

The Negative Cult and its Functions 319 

is a water-hole. The sacred character of the totem is communi- 
cated to this place, to the water-hole and even to the water 
itself, which is forbidden to all the members of the totemic group. ^ 
The initiate lives in an atmosphere charged with religiousness, 
and it is as though he were impregnated with it himself. ^ Con- 
sequently all that he possesses and all that he touches is forbidden 
to the women, and withdrawn from their contact, even down to 
the bird he has struck with his stick, the kangaroo he has pierced 
with his lance or the fish which has bit on his hook.^ But, on 
the other hand, the rites to which he is submitted and the things 
which have a part in them have a sanctity superior to his own : 
this sanctity is contagiously transmitted to everything which 
evokes the idea of one or the other. The tooth v/hich has been 
knocked out of him is considered very holy.^ For this reason, 
he may not eat animals with prominent teeth, because they 
make him think of his own lost tooth. The ceremonies of the 
Kuringal terminate with a ritual washing ; ^ a<(quatic birds 
are forbidden to the neophyte because they make him think 
of this rite. Animals that climb to the tops of trees are equally 
sacred for him, because they are too near to Daramulun, the 
god of the initiation, who lives in heaven.* The soul of a dead 
man is a sacred thing : we have already seen how this same 
property passes to the corpse in which the soul resided, to the 
spot where this is buried, to the camp in which he lived when 
alive, and which is either destroyed or quitted, to the name he 
bore, to his wife and to his relations.' They, too, are invested, 
as it were, with a sacred character ; consequently, men keep 
at a distance from them ; they do not treat them as mere profane 
beings. In the societies observed by Dawson, their names, 
like that of the dead man, cannot be pronounced during the 
period of mourning. 8 Certain animals which he ate may also be 
This contagiousness of sacredness is too well known £p 

* Spencer and Gillen, Nor. Tr., pp. 167, 299. 

* In addition to the ascetic rites of which we have spoken, there are some 
positive ones whose object is to charge, or, as Howitt says, to saturate the 
initiate with religiousness (Howitt, Nat. Tr., p. 535). It is true that instead of 
religiousness, Howitt speaks of magic powers, but as we know, for the majority 
of the ethnologists, this word merely signifies religious virtues of an impersonal 

^ Howitt, ibid., pp. 674 f. 

* Spencer and Gillen, Nat. Tr.. p. 454. Cf. Howitt, Nat. Tr., p. 561. 

* Howitt, Nat. Tr.. p. 557. « Ibid., p. 560. 

' See above, pp. 303, 306. Cf. Spencer and Gillen, Nat. Tr., p. 498 ; Nor. 
Tr., pp. 506, 507, 518 f., 526 ; Howitt, Nat. Tr., p. 449, 461, 469 ; Mathews, in 
J. of R.S. of N.S. Wales, XXXVIII, p. 274 ; Schulze, loc. cit., p. 231 ; Wyatt. 
Adelaide and Encounter Bay Tribes, in Woods, pp. 165, 198. 

^ Australian Aborigines, p. 42. " Howitt, Nat. Tr., pp. 470-471. 

320 Elementary Forms of Religious Life 

phenomenon^ to require any proof of its existence from numerous 
examples ; we only wish to show that it is as true in totemism 
as in the more advanced religions. When once established, it 
quickly explains the extreme rigour of the interdicts separating 
the sacred from the profane. Since, in virtue of this extraordinary 
power of expansion, the slightest contact, the least proximity, 
either material or simply moral, suffices to draw religious forces 
out of their domain, and since, on the other hand, they cannot 
leave it without contradicting their nature, a whole system of 
measures is indispensable for maintaining the two worlds at a 
respectful distance from one another. This is why it is forbidden 
to the profane, not only to touch, but even to see or hear that 
which is sacred, and why these two sorts of life cannot be mixed 
in their consciousnesses. Precautions are necessary to keep them 
apart because, though opposing one another, they tend to 

\ confuse themselves into one another. 

I When we understand the multiplicity of these interdicts we 
also understand the way in which they operate and the sanctions 
which are attached to them. Owing to the contagiousness 
inherent in all that is sacred, a profane being cannot violate 
an interdict without having the religious force, to which he has 
unduly approached, extend itself over him and establish its 
empire over him. But as there is an antagonism between them, 
he becomes dependent upon a hostile power, whose hostility 
cannot fail to manifest itself in the form of violent reactions 
which tend to destroy him. This is why sickness or death are 
considered the natural consequences of every transgression of 
this sort ; and they are consequences which are believed to 
come by themselves, with a sort of physical necessity. The guilty 
man feels himself attacked by a force which dominates Jiim and 
Lagainst which he is powerless. Has he eaten the totemic animal ? 
Then he feels it penetrating him and gnawing at his vitals ; he lies 
down on the ground and awaits death. ^ Every profanation implies 
a consecration, but one which is dreadful, both for the subject con- 
secrated and for those who approach him. It is the consequences 
of this consecration which sanction, in part, the interdict.^ 
It should be noticed that this explanation of the interdicts 

^ On this question, see Robertson Smith, Religion of the Semites, pp. 152 IT., 
446, 481 ; Frazer, art. Taboo in Encyc. Brit., Jevons, Introduction to the History 
of Religions, pp. 59 f. ; Crawley, Mystic Rose, ch. ii-ix ; Van Gennep, Taboti et 
Totémisme à Madagascar, ch. iii. 

* See references above, p. 128, n. i. Cf. Nor. Tr., pp. 323, 324 ; Nat. Tr., 
p. 168 ; Taplin, The Narrinyeri, p. 16 ; Roth, North Queensland Ethnography. 
Bull. 10, Records of Austral. Museum, VII, p. 76. 

^ It is to be remembered that when it is a religious interdict that has been 
violated, these sanctions are not the only ones ; there is also a real punishment 
or a stigma of opinion. 

The Negative Cult and its Functions 321 

does not depend upon the variable symbols by the aid of which 
religious forces are conceived. It matters little whether these 
are conceived as anonymous and impersonal energies or figured 
as personalities endowed with consciousness and feeling. In the 
former case, of course, they are believed to react against pro- 
faning transgressions in an automatic and unconscious manner, 
while in the latter case, they are thought to obey passionate 
movements determined by the offence resented. But at bottom, 
these two conceptions, which, moreover, have the same practical 
effect, only express one and the same psychic mechanism in two 
different languages. The basis of both is the antagonism of the 
sacred and the profane, combined with the remarkable aptitude 
of the former for spreading over to the latter ; now this antagonism 
and this contagiousness act in the same way, whether the sacred 
character is attributed to blind forces or to conscious ones. Thus^ \ 
so far is it from being true that the real religious life commences 
only where there are mythical personalities, that we see that in 
this case the rite remains the same, whether the religious beings 
are personified or not. This is a statement which we shall have 
occasion to repeat in each of the chapters which follow. 


But if this contagiousness of sacredness helps to explain the 
system of interdicts, how is it to be explained itself ? 

Some have tried to explain it with the well-known laws of 
the association of ideas. The sentiments inspired in us by a 
person or a thing spread contagiously from the idea of this thing 
or person to the representations associated with it, and thence 
to the objects which these representations express. So the 
respect which we have for a sacred being is communicated to 
everything touching this being, or resembling it, or recalling it. 
Of course a cultivated man is not deceived by these associations ; 
he knows that these derived emotions are due to mere plays of the 
images and to entirely mental combinations, so he does not give 
way to the superstitions which these illusions tend to bring about. 
But they say that the primitive naively objectifies his impressions, 
without criticising them. Does something inspire a reverential fear 
in him ? He concludes that an august and redoubtable force really 
resides in it ; so he keeps at a distance from this thing and treats it 
as though it were sacred, even though it has no right to this title. ^ 

' See Jevons, Introduction to the History of Religions, pp. 67-68. We say 
nothing of the recent, and slightly explicit, theory of Crawley {Mystic Rose, 
ch. iv-vii), according to which the contagiousness of taboos is due to a false 
interpretation of the phenomena of contagion. It is arbitrary. As Jevons very 
truly says in the passage to which we refer, the contagious character of sacred- 
ness is alTirmed a priori, and not on a faith in badly interpreted experiences. 

322 Elementary Forms of Religious Life 

But whoever says this forgets that the most primitive religions 
are not the only ones which have attributed this power of propa- 
gation to the sacred character. Even in the most recent cults, 
there is a group of rites which repose upon this principle. Does 
not every consecration by means of anointing or washing consist 
in transferring into a profane object the sanctifying virtues of 
a sacred one ? Yet it is difficult to regard an enlightened Catholic 
of to-day as a sort of retarded savage who continues to be deceived 
by his associations of ideas, while nothing in the nature of things 
explains or justifies these ways of thinking. Moreover, it is 
quite arbitrarily that they attribute to the primitive this tendency 
to objectify blindly all his emotions. In his ordinary life, and in 
the details of his lay occupations, he does not impute the properties 
of one thing to its neighbours, or vice versa. If he is less careful 
than we are about clarity and distinction, still it is far from true 
that he has some vague, deplorable aptitude for jumbling and 
confusing everything. Religious thought alone has a marked 
leaning towards these sorts of confusions. So it is in something 
special to the nature of religious things, and not in the general 
laws of the human intelligence, that the origin of these pre- 
dispositions is to be sought. 

When a force or property seems to be an integral part or 
constitu'^nt element of the subject in which it resides, we cannot 
easily imagine its detaching itself and going elsewhere. A body 
is defined by its mass and its atomic composition ; so we do 
not think that it could communicate any of these distinctive 
characteristics by means of contact. But, on the other hand, 
if we are dealing with a force which has penetrated the body 
from without, since nothing attaches it there and since it is 
foreign to the body, there is nothing inconceivable in its escaping 
again. Thus the heat or electricity which a body has received 
from some external source may be transmitted to the surrounding 
medium, and the mind readily accepts the possibility of this 

Ptransmission. So the extreme facility with which religious 
forces spread out and diffuse themselves has nothing surprising 
about it, if they are generally thought of as outside of the beings 
in which they reside. Now this is just what the theory we have 

^proposed implies. 

r In fact, they are only collective forces hypostatized, that is to 
say, moral forces ; they are made up of the ideas and sentiments 
awakened in us by the spectacle of society, and not of sensations 
coming from the physical world. So they are not homogeneous 
with the visible things among which we place them. They may 
well take from these things the outward and material forms 

L_in which they are represented, but they owe none of their efficacy 

The Negative Cult and its Functions 323 

to them. They are not united by external bonds to the different 
supports upon which they aUght ; they have no roots there ; 
according to an expression we have already used ^ and which 
serves best for characterizing them, they are added to them. So 
there are no objects which are predestined to receive them, to 
the exclusion of all others ; even the most insignificant and 
vulgar may do so ; accidental circumstances decide which are 
the chosen ones. The terms in which Codrington speaks of the 
mana should be borne in mind : it is a force, he says, which 
" is not fixed in anything and can be conveyed in almost anything." ^ 
Likewise, the Dakota of Miss Fletcher represented the wakan 
as a sort of surrounding force which is always coming and going 
through the world, alighting here and there, but definitely fixing 
itself nowhere.' Even the religious character inherent in men 
does not have a different character. There is certainly no other 
being in the world of experience which is closer to the very source 
of all religious life ; none participates in it more directly, for it is 
in human consciousnesses that it is elaborated. Yet we know 
that the religious principle animating men, to wit, the soul, is 
partially external. 

But if religious forces have a place of their own nowhere, 
their mobility is easily explained. Since nothing attaches 
them to the things in which we localize them, it is natural that 
they should escape on the slightest contact, in spite of themselves, 
so to speak, and that they should spread afar. Their intensity 
incites them to this spreading, which everything favours. This 
is why the soul itself, though holding to the body by very personal 
bonds, is constantly threatening to leave it : all the apertures 
and pores of the body are just so many ways by which it tends 
to spread and diffuse itself into the outside.^ 

But we shall account for this phenomenon which we are 
trying to understand, still better if, instead of considering the 
notion of religious forces as it is when completely formulated, 
we go back to the mental process from which it results. 

We have seen, in fact, that the sacred character of a being 
does not rest in any of its intrinsic attributes. It is not because 
the totemic animal has a certain aspect or property that it inspires 
religious sentiments ; these result from causes wholly foreign to 
the nature of the object upon which they fix themselves. What 
constitutes them are the impressions of comfort and dependence 
which the action of the society provokes in the mind. Of 
themselves, these emotions are not attached to the idea of any 

' See above, p. 229. * See above, p. 194. 

* See above, p. 190. 

* Tliis has been well demonstrated by Preuss in his articles in the Globus. 

324 Elementary Forms of Religious Life 

particular object ; but as these emotions exist and are especially 
intense, they are also eminently contagious. So they make a 
stain of oil ; they extend to all the other mental states which 
occupy the mind ; they penetrate and contaminate those repre- 
sentations especially in which are expressed the various objects 
which the man had in his hands or before his eyes at the moment : 
the totemic designs covering his body, the bull-roarers which he 
j^was making roar, the rocks surrounding him, the ground under 
I his feet, etc. It is thus that the objects themselves get a religious 
value which is really not inherent in them but is conferred from 
without. So the contagion is not a sort of secondary process 
by which sacredness is propagated, after it has once been acquired ; 
it is the very process by which it is acquired. It is by contagion 
that it establishes itself : we should not be surprised, therefore, 
if it transmits itself contagiously. What makes its reality is a 
special emotion ; if it attaches itself to some object, it is because 
this emotion has found this object in its way. So it is natural 
that from this one it should spread to all those which it finds in 
its neighbourhood, that is to say, to all those which any reason 
whatsoever, either material contiguity or mere similarity, has 
fmentally connected with the first. 

Thus, the contagiousness of sacredness finds its explanation 
in the theory which we have proposed of religious forces, and by 
this very fact, it serves to confirm our theory,^ And, at the 
same time, it aids us in understanding a trait of primitive 
mentality to which we have already called the attention. 
P We have seen ^ the facility with which the primitive confuses 
kingdoms and identifies the most heterogeneous things, men, 
animals, plants, stars, etc. Now we see one of the causes which 
has contributed the most to facilitating these confusions. Since 
religious forces are eminently contagious, it is constantly happen- 
ing that the same principle animates very different objects 
equally ; it passes from some into others as the result of either 
a simple material proximity or of even a superficial similarity. 
It is thus that men, animals, plants and rocks come to have the 
same totem : the men because they bear the name of the animal : 
the animals because they bring the totemic emblem to mind ; 
the plants because they nourish these animals ; the rocks because 
they mark the place where the ceremonies are celebrated. Now 
religious forces are therefore considered the source of all efficacy ; 

^ It is true that this contagiousness is not peculiar to religious forces ; those 
belonging to magic have the same property ; yet it is evident that they do not 
correspond to objectified social sentiments. It is because magic forces have been 
conceived on the model of religious forces. We shall come back to this point 
again (see p. 361). 

* See above, p. 235. 

The Negative Cult and its Functions 325 

so beings having one single religious principle ought to pass as 
having the same essence, and as differing from one another only 
in secondary characteristics. This is why it seemed quite natural 
to arrange them in a single category and to regard them as mere 
varieties of the same class, transmutable into one another. 1 

When this relation has been established, it makes the » 
phenomena of contagion appear under a new aspect. Taken 
by themselves, they seem to be quite foreign to the logical life. 
Is their effect not to mix and confuse beings, in spite of their 
natural differences ? But we have seen that these confusions 
and participation have played a rôle of the highest utility in 
logic ; they have served to bind together things which sensation 
leaves apart ftom one another. So it is far from true that con- 
tagion, the source of these connections and confusions, is marked 
with that fundamental irrationality that one is inclined to 
attribute it at first. It has opened the way for the scientifiicj 
explanations of the future. 



I. — The Elements of the Sacrifice 

WHATEVER the importance of the negative cult may be, 
and though it may indirectly have positive effects, it 
does not contain its reason for existence in itself ; it introduces 
one to the religious life, but it supposes this more than it con- 
stitutes it. If it orders the worshipper to flee from the profane 

Pworld, it is to bring him nearer to the sacred world. Men have 
never thought that their duties towards religious forces might 
be reduced to a simple abstinence from all commerce ; they 
have always believed that they upheld positive and bilateral 

relations with them, whose regulation and organization is the 

function of a group of ritual practices. To this special system 
of rites we give the name of positive cult. 

For some time we almost completely ignored the positive cult 
of the totemic religion and what it consists in. We knew almost 
nothing more than the initiation rites, and we do not know those 
sufficiently well even now. But the observations of Spencer and 
Gillen, prepared for by those of Schulze and confirmed by those 
of Strehlow, on the tribes of central Australia, have partially 
filled this gap in our information. There is one ceremony especially 
which these explorers have taken particular pains to describe 
to us and which, moreover, seems to dominate the whole totemic 
cult : this is the one that the Arunta, according to Spencer and 
Gillen, call the Intichiunia. It is true that Strehlow contests 
the meaning of this word. According to him, intichiuma (or, 
as he writes it, intijiuma) means " to instruct " and designates 
the ceremonies performed before the young man to teach him 
the traditions of the tribe. The feast which we are going to 
describe bears, he says, the name mbatjalkatiuma, which means 
" to fecundate " or " to put into a good condition." ^ But we 
shall not try to settle this question of vocabulary, which touches 
the real problem but slightly, as the rites in question are all 

^ Strehlow, I, p. 4. 

The Elements of the Sacrifice 327 

celebrated in the course of the initiation. On the other hand, 
as the word Intichiuma now belongs to the current language of 
ethnography, and has almost become a common noun, it seems 
useless to replace it with another. * 

The date on which the Intichiuma takes place depends largely 
upon the season. There are two sharply separated seasons in 
Australia : one is dry and lasts for a long time ; the other is 
rainy and is, on the contrary, very short and frequently irregular. 
As soon as the rains arrive, vegetation springs up from the ground 
as though by enchantment and animals multiply, so that the 
country which had recently been only a sterile desert is rapidly 
filled with a luxurious flora and fauna. It is just at the moment 
when the good season seems to be close at hand that the Inti- 
chiuma is celebrated. But as the rainy season is extremely 
variable, the date of the ceremonies cannot be fixed once for all. 
It varies with the climatic circumstances, which only the chief 
of the totemic group, the Alatunja, is qualified to judge : on 
a day which he considers suitable, he informs his companions 
that the moment has arrived. ^ 

Each totemic group has its own Intichiuma. Even if this 
rite is general in the societies of the centre, it is not the same 
everywhere ; among the Warramunga, it is not what it is among 
the Arunta ; it varies, not only among the tribes, but also within 
the tribe, among the clans. But it is obvious that the different 
mechanisms in use are too closely related to each other to be 
dissociated completely. There is no ceremony, perhaps, which 
is not made up of several, though these are very unequally 
developed : what exists only as a germ in one, occupies the 
most important place in another, and inversely. Yet they must 
be carefully distinguished, for they constitute just so many 
different ritual types to be described and explained separately, 
but afterwards we must seek some common source from which 
they were derived. 

Let us commence with those observed among the Arunta. 


The celebration includes two successive phases. The object of 
the rites which take place in the first is to assure the prosperity 
of the animal or vegetable species serving the clan as totem. 
The means employed for this end may be reduced to two principal 
types. I 

^ Of course the word designating these celebrations changes with the tribes. 
The Urabunna call them Pitjinta (Nor. Tr., p. 284) ; the Warramunga Thala- 
minta (ibid., p. 297), etc. 

* Schulze, he cit., p. 243 ; Spencer and Gillen, Nat. Tr., pp. i6q f. 

328 Elementary Forms of Religious Life 

It will be remembered that the fabulous ancestors from whom 
each clan is supposed to be descended, formerly lived on earth 
and left traces of their passage there. These traces consist 
especially in stones and rocks which they deposited at certain 
places, or which were formed at the spots where they entered 
into the ground. These rocks and stones are considered the 
bodies or parts of the bodies of the ancestors, whose memory 
they keep alive ; they represent them. Consequently, they 
also represent the animals and plants which served these same 
ancestors as totems, for an individual and his totem are only 
one. The same reality and the same properties are attributed 
to them as to the actually living plants or animals of the same 
species. But they have this advantage over these latter, that 
they are imperishable, knowing neither sickness nor death. 
So they are like a permanent immutable and ever-available 
reserve of animal and vegetable life. Also, in a certain number 
of cases, it is this reserve that they annually draw upon to assure 
the reproduction of the species. 

Here, for example, is how the Witchetty grub clan, at Alice 
Springs, proceeds at its Intichiuma.^ 

On the day fixed by the chief, all the members of the totemic 
group assemble in the principal camp. The men of the other 
totems retire to a distance ; ^ for among the Arunta, they are 
not allowed to be present at the celebration of the rite, which 
has all the characteristics of a secret ceremony. An individual 
of a different totem, but of the same phratry, may be invited 
to be present, as a favour ; but this is only as a witness. In no 
case can he take an active part. 

After the men of the totem have assembled, they leave the 
camp, leaving only two or three of their number behind. They 
advance in a profound silence, one behind another, all naked, 
without arms and without any of their habitual ornaments. 
Their attitude and their pace are marked with a religious 
gravity : this is because the act in which they are taking part 
has an exceptional importance in their eyes. Also, until the 
end of the ceremony they are required to observe a rigorous 

The country which they traverse is all filled with souvenirs 
left by the glorious ancestors. Thus they arrive at a spot where 
a huge block of quartz is found, with small round stones all 
around it. This block represents the witchetty grub as an adult. 
The Alatunja strikes it with a sort of wooden tray called aptnara,^ 

*■ Nat. Tr., pp. 170 ff. 

* Of course the women are under the same obligation. 

' The apmara is the only thing which he brought from the camp. 

The Elements of the Sacrifice 329 

and at the same time he intones a chant, whose object is to 
invite the animal to lay eggs. He proceeds in the same fashion 
with the stones which are regarded as the eggs of the animal 
and with one of which he rubs the stomach of each assistant. 
This done, they all descend a little lower, to the foot of a cliff 
also celebrated in the myths of the Alcheringa, at the base of 
which is another stone, also representing the witchetty grub. 
The Alatunja strikes it with his apmara ; the men accompanying 
him do so as well, with branches of a gum-tree which they have 
gathered on the way, all of which goes on in the midst of chants 
renewing the invitation previously addressed to the animal. 
About ten different spots are visited in turn, some of which are 
a mile or more from the others. At each of them there is a 
stone at the bottom of a cave or hole, which is believed to repre- 
sent the witchetty grub in one of his aspects or at one of the 
phases of his existence, and upon each of these stones, the same 
ceremonies are repeated. 

The meaning of the rite is evident. When the Alatunja 
strikes the sacred stones, it is to detach some dust. The grains 
of this very holy dust are regarded as so many germs of life ; 
each of them contains a spiritual principle which will give birth 
to a new being, when introduced into an organism of the same 
species. The branches with which the assistants are provided 
serve to scatter this precious dust in all directions ; it is scattered 
everywhere, to accomplish its fecundating work. By this means, 
they assure, in their own minds, an abundant reproduction of the 
animal species over which the clans guard, so to speak, and upon 
which it depends. 

The natives themselves give the rite this interpretation. 
Thus, in the clan of the ilpirla (a kind of "manna"), they 
proceed in the following manner. When the day of the Inti- 
chiuma arrives, the group assembles near a huge rock, about 
fifty feet high ; on top of this rock is another, very similar to 
the first in aspect and surrounded by other smaller ones. Both 
represent masses of manna. The Alatunja digs up the ground 
at the foot of this rock and uncovers a churinga which is believed 
to have been buried there in Alcheringa times, and which is, 
as it were, the quintessence of the manna. Then he climbs up 
to the summit of the higher rock and rubs it, first with the 
churinga and then with the smaller stones which surround it. 
Finally, he brushes away the dust which has thus been collected 
on the surface of the rock, with the branches of a tree ; each of 
the assistants does the same in his turn. Now Spencer and 
Gillen say that the idea of the natives is that the dust thus 
scattered will " settle upon the mulga trees and so produce 

330 Elementary Forms of Religious Life 

manna." In fact, these operations are accompanied by a hymn 
sung by those present, in which this idea is expressed. ^ 

With variations, this same rite is found in other societies. 
Among the Urabunna, there is a rock representing an ancestor 
of the Lizard clan ; bits are detached from it which they throw 
in every direction, in order to secure an abundant production 
of Uzards.2 In this same tribe, there is a sand-bank which 
mythological souvenirs closely associate with the louse totem. 
At the same spot are two trees, one of which is called the ordinary 
louse tree, the other, the crab-louse tree. They take some of 
this sand, rub it on these trees, throw it about on every side 
and become convinced that, as a result of this, lice will be born in 
large numbers. ^ The Mara perform the Intichiuma of the bees 
by scattering dust detached from sacred rocks. ^ For the 
kangaroo of the plains, a slightly different method is used. 
They take some kangaroo-dung and wrap it up in a certain 
herb of which the animal is very fond, and which belongs to the 
kangaroo totem for this reason. Then they put the dung, thus 
enveloped, on the ground between two bunches of this herb 
and set the whole thing on fire. With the flame thus made, 
they light the branches of trees and then whirl them about in 
such a way that sparks fly in every direction. These sparks 
play the same rôle as the dust in the preceding cases. ^ 

In a certain number of clans,® men mix something of their 
own substance with that of the stone, in order to make the 
rite more efficacious. Young men open their veins and let streams 
of blood flow on to the rock. This is the case, for example, in 
the Intichiuma of the Hakea flower among the Arunta. The 
ceremony takes place in a sacred place around an equally sacred 
rock which, in the eyes of the natives, represents Hakea flowers. 
After certain preliminary operations, " the old leader asks one 
of the young men to open a vein in his arm, which he does, 
and allows the blood to sprinkle freely, while the other men 
continue the singing. The blood flows until the stone is com- 
pletely covered." ' The object of this practice is to revivify 
the virtues of the stone, after a fashion, and to reinforce its 
efficacy. It should not be forgotten that the men of the clan are 
relatives of the plant or animal whose name they bear ; the 
same principle of life is in them, and especially in their blood. 
So it is only natural that one should use this blood and the mystic 
germs which it carries to assure the regular reproduction of the 

1 Nat. Tr.. pp. 185-186. « Nor. Tr., p. 288. » Ibid. 

* Nor. Tr., p. 312. * Ibid. 

* We shall see below that these clans are much more numerous than Spencer 
and Gillen say. 

' Nat. Tr., pp. 184-185. 

The Elements of the Sacrifice 331 

totemic species. It frequently happens among the Arunta 
that when a man is sick or tired, one of his young companions 
opens his veins and sprinkles him with his blood in order to re- 
animate him.i If blood is able to reawaken hfe in a man in this 
way, it is not surprising that it should also be able to awaken it 
in the animal or vegetable species with which the men of the clan 
are confounded. 

The same process is employed in the Intichiuma of the Undiara 
kangaroo among the Arunta. The theatre of the ceremony is a 
water-hole vaulted over by a peaked rock. This rock represents 
an animal-kangaroo of the Alcheringa which was killed and 
deposited there by a man-kangaroo of the same epoch ; many 
kangaroo spirits are also believed to reside there. After a certain 
number of sacred stones have been rubbed against each other 
in the way we have described, several of the assistants climb up 
on the rock upon which they let their blood flow.^ " The purpose 
of the ceremony at the present day, so say the natives, is by 
means of pouring out the blood of kangaroo men upon the rock, 
to drive out in all directions the spirits of the kangaroo animals 
and so to increase the number of the animals." ^ 

There is even one case among the Arunta where the blood 
seems to be the active principle in the rite. In the Emu group, 
they do not use sacred stones or anything resembling them. 
The Alatunja and some of his assistants sprinkle the ground 
with their blood ; on the ground thus soaked, they trace lines 
in various colours, representing the difterent parts of the body 
of an emu. They kneel down around this design and chant a 
monotonous hymn. From the fictitious emu to which this chant 
is addressed, and, consequently, from the blood which has served 
to make it, they believe that vivifying principles go forth, which 
animate the embryos of the new generation, and thus prevent 
the species from disappearing.'* 

Among the Wonkgongaru,^ there is one clan whose totem 
is a certain kind of fish ; in the Intichiuma of this totem also, 
it is the blood that plays the principal part. The chief of the 

* Nat. Tr., pp. 438, 461, 464 ; Nor. Tr., pp. 596 ff. 

* Nat. Tr., p. 201. 

^ Ibid., p. 206. We use the words of Spencer and Gillen, and with them, we 
say that " spirits or spirit parts of kangaroo " are disengaged from the rocks. 
Strehlow (III, p. 7) contests the exactness of this expression. According to him, 
the rite makes real kangaroos, with living bodies, appear. But this dispute is 
without interest, just as the one about the notion of the ratapa was (see above, 
p. 252). The kangaroo germs thus escaping from the rock are not visible, so they 
are not made out of the same substance as the kangaroos which wc see. This is 
all that Spencer and Gillen mean to say. It is quite certain, moreover, that they 
are not pure spirits such as a Christian might conceive. Like human souls, they 
have a material form. 

* Nat. Tr., p. 181. * A tribe on the east of Lake Eyre. 

332 Elementary Forms of Religious Life 

group, after being ceremoniously painted, goes into a pool of 
water and sits down there. Then he pierces his scrotum and the 
skin around his navel with small pointed bones. " The blood 
from the wounds goes into the water and gives rise to fish." ^ 

By a wholly similar process, the Dieri think that they assure 
the reproduction of two of their totems, the carpet snake and the 
woma snake (the ordinary snake). A Mura-mura named Minkani 
is thought to liv.e under a dune. His body is represented by some 
fossil bones of animals or reptiles, such as the deltas of the rivers 
flowing into Lake Eyre contain, according to Howitt. When 
the day of the ceremony arrives, the men assemble and go to the 
home of the Minkani, There they dig until they come to a 
layer of damp earth which they call " the excrement of Minkani." 
From now on, they continue to turn up the soil with great care 
until they uncover " the elbow of Minkani." Then two young 
men open their veins and let their blood flow on to the sacred 
rock. They chant the hymn of Minkani while the assistants, 
carried away in a veritable frenzy, beat each other with their 
arms. The battle continues until they get back to the camp, 
which is about a mile away. Here, the women intervene and 
put an end to the combat. They collect the blood which has 
flown from the wounds, mix it with the " excrement of Minkani," 
and scatter the resulting mixture over the dune. When this rite 
has been accomplished, they are convinced that carpet snakes 
will be born in abundance. ^ 

In certain cases, they use the very substance which they wish 
to produce as the vivifying principle. Thus among the Kaitish, 
in the course of a ceremony whose object is to create rain, they 
sprinkle water over a sacred rock which represents the mythical 
heroes of the Water clan. It is evident that they believe that 
by this means they augment the productive virtues of the rock 
just as well as with blood, and for the same reasons.^ Among 
the Mara, the actor takes water from a sacred hole, puts it in his 
mouth and spits it out in every direction.^ Among the Worgaia, 
when the yams begin to sprout, the chief of the Yam clan sends 
men of the phratry of which he is not a rriember himself to gather 
some of these plants ; these bring some to him, and ask him to 
intervene, in order that the species may develop well. He takes 
one, chews it, and throws the bits in every direction.^ Among 
the Kaitish when, after various rites which we shall not describe, 
the grain of a certain grass called Erlipinna has reached its full 

* Nor. Tr., pp. 287 f. 

* Howitt, Nat. Tr., p. 798. Cf. Howitt, Legends of the Dieri and Kindred 
Tribes of Central Australia, in J. A. I., XXIV, pp. 124 ft. Howitt believes that the 
ceremony is performed by the men of the totem, but is not prepared to say so 
definitely. * Nor. Tr., p. 295. * Ibid., p. 314. * Ibid., pp. 296 f. 

The Elements of the Sacrifice 333 

development, the chief of the totem brings a little of it to camp 
and grinds it between two stones ; the dust thus obtained is 
piously gathered up, and a few grains are placed on the lips 
of the chief, who scatters them by blowing. This contact with 
the mouth of the chief, which has a very special sacramental 
virtue, undoubtedly has the object of stimulating the vitaUty 
of the germs which these grains contain and which, being blown 
to all the quarters of the horizon, go to communicate these 
fecundating virtues which they possess to the plants.^ 

The efficacy of these rites is never doubted by the native : 
he is convinced that they must produce the results he expects, 
with a sort of necessity. If events deceive his hopes, he merely 
concludes that they were counteracted by the sorcery of some 
hostile group. In any case, it never enters his mind that a favour- 
able result could be obtained by any other means. If by chance 
the vegetation grows or the animals produce before he has per- 
formed his Intichiuma, he supposes that another Intichiuma 
has been celebrated under the ground by the ancestors and that 
the living reap the benefits of this subterranean ceremony. ^ 


This is the first act of the celebration. 

During the period immediately following, there are no regular 

* Nat. Tr., p. 170. 

^ Ibid., p. 519. — The analysis of the rites which have just been studied is 
based solely on the observations of Spencer and Gillen. Since this chapter was 
written, Strehlow has published the third fascicule of his work, which deals 
with the positive cult and especially the Intichiuma, or, as he says, the rites of 
the mbatjalkatiuma. But we have found nothing in this publication which 
obliges us to modify the preceding description or even to complete it with 
important additions. The most interesting thing taught by Strehlow on this 
subject is that the effusions and oblations of blood are much more frequent than 
one would suspect from the account of Spencer and Gillen (see Strehlow, III, 
pp. 13, 14, 19, 29, 39, 43, 46, 56, 67, 80, 89). 

Moreover, the information given by Strehlow in regard to the cult must be 
taken carefully, for he was not a witness of the rites he describes ; he confined 
himself to collecting oral testimony, which is generally rather summary (see 
fasc. Ill, Preface of Leonhardi, p. v). It may even be asked if he has not con- 
fused the totemic ceremonies of initiation with those wliich he calls mbatjal- 
katiuma, to an excessive degree. Of course, he has made a praiseworthy attempt 
to distinguish them and has made two of their distinctive characteristics very 
evident. In the first place, the Intichiuma always takes place at a sacred spot 
to which the souvenir of some ancestor is attached, while the initiation ceremonies 
may be celebrated anywhere. Secondly, the oblations of blood are special to the 
Intichiuma, which proves that they are close to the heart of the ritual (III, p. 7). 
But in the description which he gives us of the rites, we find facts belonging 
indifferently to each species of ceremony. In fact, in what he describes under 
the name mbatjalkatiuma, the young men generally take an important part (for 
example, see pp. 11, 13, etc.), which is characteristic of the initiation. Also, it 
seems as though the place of the rite is arbitrary, for the actors construct their 
scene artificially. They dig a hole into which they go ; he seldom makes any 
allusion to sacred trees or rocks and their ritual rôle. 

334 Elementary Forms of Religious Life 

ceremonies. However, the religious life remains intense : this is 
manifested especially by an aggravation of the system of interdicts. 
It is as though the sacred character of the totem were reinforced : 
they do not even dare to touch it. In ordinary times, the Arunta 
may eat the animal or plant which serves as totem, provided 
they do so with moderation, but on the morrow of the Intichiuma 
this right is suspended ; the alimentary interdiction is strict and 
without exceptions. They believe that any violation of this 
interdict would result in neutralizing the good effects of the rite 
and in preventing the increase of the species. It is true that the 
men of other totems who happen to be in the same locality are 
not submitted to the same prohibition. However, their liberty 
is less than ordinary at this time. They may not consume the 
totemic animal wherever they place, in the brush, for example ; 
they must bring it to camp, and it is there only that it may be 

A final ceremony terminates this period of extraordinary 
interdictions and definitely closes this long series of rites. It 
varies somewhat in different clans, but the essential elements 
are the same everywhere. Here are the two principal forms 
which it takes among the Arunta. One of these is in connection 
with the witchetty grub, the other with the kangaroo. 

When the grubs have attained full maturity and appear in 
abundance, the men of the totem, as well as others, collect as 
many of them as possible ; then they all bring those they have 
found back to camp and cook them until they become hard and 
brittle. They are then preserved in wooden vessels called pitchi. 
The harvest of grubs is possible only during a very short time, 
for they appear only after the rain. When they begin to be less 
numerous, the Alatunja summons everybody to the camp ; on 
his invitation, each one brings his supply. The others place 
theirs before the men of that totem. The Alatunja takes one 
of these pitchi and, with the aid of his companions, he grinds 
its contents between two stones ; after this, he eats a little of 
the powder thus obtained, his assistants do the same, and what 
remains is given to the men of the other clans, who may now 
dispose of it freely. They proceed in exactly the same manner 
with the supply provided by the Alatunja. From now on, 
the men and women of the totem may eat it, but only a little at 
a time ; if they went beyond the limits allowed, they would lose 
the powers necessary to celebrate the Intichiuma and the species 
would not reproduce. Yet, if they did not eat any at all, and 
especially if the Alatunja ate none in the circumstances we have 
just described, they would be overtaken by the same incapacity. 

^ Nat. Tr., p. 203. Cf. Meyer, The Encounter Bay Tribe, in Woods, p. 187. 

The Elements of the Sacrifice 335 

In the totemic group of the Kangaroo, which has its centre ât 
Undiara, certain characteristics of the ceremony are more clearly 
marked. After the rites which we have described have been 
accomplished on the sacred rock, the young men go and hunt 
the kangaroo, bringing their game back to the camp. Here, 
the old men, with the Alatnnja in their midst, eat a little of the 
flesh of the animal, and anoint the bodies of those who took 
part in the Intichiuma with its fat. The rest is divided up among 
the men assembled. Next, the men of the totem decorate 
themselves with totemic designs and the night is passed in 
songs commemorating the exploits accomplished by men and 
animal kangaroos in the times of the Alcheringa. The next day, 
the young men go hunting again in the forest and bring back 
a larger number of kangaroos than the first time, and the cere- 
monies of the day before recommence.^ 

With variations of detail, the same rite is found in other 
Arunta clans, ^ among the Urabunna,^ the Kaitish,* the Un- 
matjera,^ and in the Encounter Bay Tribe.* Everywhere, it is 
made up of the same essential elements. A few specimens of the 
totemic animal or plant are presented to the chief of the clan, 
who solemnly eats them and who must eat them. If he did not 
fulfil this duty, he would lose the power of celebrating the Inti- 
chiuma efficaciously, that is to say, so as to recreate the species 
annually. Sometimes the ritual consumption is followed by 
an unction made with the fat of the animal or certain parts of 
the plant.' This rite is generally repeated by the men of the 
totem, or at least by the old men, and after it has been accom- 
plished, the exceptional interdictions are raised. 

In the tribes located farther north, among the Warramunga 
and neighbouring societies,^ this ceremony is no longer found. 
However, traces are found which seem to indicate that there 
was a time when it was known. It is true that the chief of the 
clan never eats the totem ritually and obligatorily. But in certain 
cases, men who are not of the totem whose Intichiuma has just 
been celebrated, must bring the animal or plant to camp and 
offer it to the chief, asking him if he wants to eat it. He refuses 
and adds, " I have made this for you ; you may eat it freely." ^ 
So the custom of the presentation remains and the question asked 

^ Spencer and Gillen, Nat. Tr., p. 204. 

* Nat. Tr., pp. 205-207. 

^ Nor. Tr., pp. 286 f. * Ibid., p. 294. 

^ Ibid., p. 296. 

• Meyer, in Woods, p. 187. 

' We have already cited one case ; others will be found in Spencer and 
Gillen, Nat. Tr., p. 208 ; Nor. Tr., p. 286. 

* The Walpari, Wulniala, Tjingilli, Umbaia. 

• Nor. Tr., p. 318. 

336 Elementary Forms of Religious Life 

of the chief seems to date back to an epoch when the ritual con- 
sumption was practised.^ 


The interest of the system of rites which has just been described 
Hes in the fact that in them we find, in the most elementary 
form that is actually known, all the essential principles of a 
great religious institution which was destined to become one of 
the foundation stones of the positive cult in the superior religions : 
this is the institution of sacrifice. 

We know what a revolution the work of Robertson Smith 
brought about in the traditional theory of-sacdfice.^ Before 
him, sacrifice was regarded as a sort of tribute or homage, either 
obligatory or optional, analogous to that which subjects owe to 
their princes. Robertson Smith was the first to remark that 
this classic explanation did not account for two essential charac- 
teristics of the rite. In the first place, it is a repast : its substance 

4s food. Secondly, it is a repast in which the worshippers who 
offer it take part, along with the god to whom it is offered. Certain 

^tarts of the victim are reserved for the divinitjT;^ others are 

^ For the second part of the ceremony as for the first, we have followed 
Spencer and Gillen. On this subject, the recent fascicule of Strehlow only con- 
firms the observations of his predecessors, at least on all essential points. He 
recognizes that after the first ceremony (two months afterwards, he says, p. 13), 
the chief of the clan eats the totemic animal or plant ritually and that after this 
he raises the interdicts ; he calls this operation die Freigabe des Totems zum 
allgemeinen Gebrauch (III, p. 7). He even tells us that this operation is im- 
portant enough to have a special word for it in the Arunta language. He adds, 
it is true, that this ritual consummation is not the only one, but that the chiefs 
and old men sometimes eat the sacred plant or animal before the first ceremony 
and that the performer of the rite does so after the celebration. The fact is not 
improbable ; these consummations are means employed by the officiants or 
assistants to acquire virtues which they acquire ; it is not surprising if they are 
numerous. It does not invalidate the account of Spencer and Gillen at all, for 
the rite upon which they insist, and not without reason, is the Freigabe des 

On only two points does Strehlow contest the allegations of Spencer and 
Gillen. In the first place, he declares that the ritual consumption does not take 
place in every case. This cannot be doubted, for there are some animals and 
plants which are not edible. But still, the rite is very frequent ; Strehlow himself 
cites numerous examples (pp. 13, 14, 19, 23, 33, 36, 50, 59, 67, 68, 71, 75, 80, 84, 
89, 93). Secondly, we have seen that according to Spencer and Gillen, if the 
chief does not eat the totemic animal or plant, he will lose his powers. Strehlow 
assures us that the testimony of natives does not confirm this assertion. But 
this question seems to us to be quite secondary. The assured fact is that the 
ritual consumption is required, so it must be thought useful or necessary. Now, 
like every communion, it can only serve to confer needed virtues upon the 
person communicating. It does not follow from the fact that the natives, or 
some of them, have forgotten this function of the rite, that it is not real. Is it 
necessary to repeat that worshippers are generally ignorant of the real reasons 
for their practices ? 

^ See The Religion of the Semites, Lectures vi-xi, and the article Sacrifice in 
the Encyclopcedia Britannica (Ninth Edition). 

The Elements of the Sacrifice 337 

attributed to the sacrificers, who consume them ; this is why 
the Bible often speaks of the sacrifice as a repast in the presence 
of Jahveh. Now in a multitude of societies, meals taken in 
common are believed to create a bond of artificial kinship 
between those who assist at them. In fact, relatives are people 
who are naturally made of the same flesh and blood. But food 
is constantly remaking the substance of the organism. So a 
common food may produce the same effects as a common origin. 
According to Smith, sacrificial banquets have the object of 
making the worshipper and his god communicate in the same 
flesh, in order to form a bond of kinship between them. From 
this point of view, sacrifice takes on a wholly new aspect. Its 
essential element is no longer the act of renouncement which 
the word sacrifice ordinarily expresses ; before all, it is an act 
of alimentary communion. 

Of course there are some reservations to be made in the details 
of this way of explaining the efficacy of sacrificial banquets. 
This does not result exclusively from the act of eating together. 
A man does not sanctify himself merely by sitting down, in 
some way, at the same table with a god, but especially by eating 
food at this ritual repast which has a sacred character. It has 
been shown how a whole series of preliminary operations, lustra- 
tions, unctions, prayers, etc., transform the animal to be immo- 
lated into a sacred thing, whose sacredness is subsequently 
transferred to the worshipper who eats it.^ But it is true, none 
the less, that the alimentary communion is one of the essential 
elements of the sacrifice. Now when we turn to the rite which 
terminates the ceremonies of the Intichiuma, we find that it, 
too, consists in an act of this sort. After the totemic animal 
has been killed, the Alatunja and the old men solemnly eat it. 
So they communicate with the sacred principle residing in it 
and they assimilate it. The only difference we find here is that 
the animal is naturally sacred while it ordinarily acquires this 
character artificially in the course of the sacrifice. 

Moreover, the object of this communion is manifest. Every 
member of a totemic clan contains a mystic substance within 
him which is the pre-eminent part of his being, for his soul 
is made out of it. From it come whatever powers he has and 
his social position, for it is this which makes him a person. So 
he has a vital interest in maintaining it intact and in keeping it, 
as far as is possible, in a state of perpetual youth. Unfortunately 
all forces, even the most spiritual, are used up in the course of 
time if nothing comes to return to them the energy they lose 

* See Hubert and Mauss, Essai sur la nature et la fonction du sacrifice, in 
Mélanges d'histoire des religions, pp. 40 ff. 

338 Elementary Forms of Religious Life 

through the normal working of things ; there is a necessity of 
the first importance here which, as we shall see, is the real reason 
for the positive cult. Therefore the men of a totem cannot 
retain their position unless they periodically revivify the totemic 
principle which is in them ; and as they represent this principle 
in the form of a vegetable or animal, it is to the corresponding 
animal or vegetable species that they go to demand the sup- 
plementary forces needed to renew this and to rejuvenate it. 
A man of the Kangaroo clan believes himself and feels himself 
a kangaroo ; it is by this quality that he defines himself ; it is 
this which marks his place in the society. In order to keep it, he 
takes a little of the flesh of this same animal into his own body 
from time to time. A small bit is enough, owing to the rule : 
the part is equal to the whole.^ 

If this operation is to produce all the desired effects, it may 
not take place at no matter what moment. The most opportune 
time is when the new generation has just reached its complete 
development, for this is also the moment when the forces 
animating the totemic species attain their maximum intensity. 
They have just been drawn with great difficulty from those rich 
reservoirs of life, the sacred trees and rocks. Moreover, all 
sorts of means have been employed to increase their intensity 
still more ; this is the use of the rites performed during the first 
part of the Intichiuma. Also, by their very aspect, the firstfruits 
of the harvest manifest the energy which they contain : here the 
totemic god acclaims himself in all the glory of his youth. This 
is why the firstfruits have always been regarded as a very 
sacred fruit, reserved for very holy beings. So it is natural that 
the Australian uses it to regenerate himself spiritually. Thus 
both the date and the circumstances of the ceremonies are 
I Perhaps some will be surprised that so sacred^a^Jood may 
be eaten by ordinary profane persons. But in the first place, 
there is no positive cult which does not face this contradiction. 
Every sacred being is removed from profane touch by this very 
character with which it is endowed ; but, on the other hand, 
they would serve for nothing and have no reason whatsoever 
for their existence if they could not come in contact with these 
same worshippers who, on another ground, must remain respect- 
fully distant from them. At bottom, there is no positive rite 
which does not constitute a veritable sacrilege, for a man cannot 
hold commerce with the sacred beings without crossing the 
barrier which should ordinarily keep them separate. But the 
important thing is that the sacrilege should be accompanied 

• See the explanation of this rule, above, p. 229. 

The Elements of the Sacrifice 339 

with precautions which attenuate it. Among those employed, 
the most usual one consists in arranging the transition so as 
to introduce the worshipper slowly and gradually into the 
circle of sacred things. When it has been broken and diluted 
in this fashion, the sacrilege does not offend the religious con- 
science so violently ; it is not regarded as a sacrilege and so 
vanishes. This is what happens in the case now before us. \ 
The effect of the whole series of rites which has preceded the 
moment when the totem is solemnly eaten has been to sanctify 
those who took an active part in them. They constitute an 
essentially religious period, through which no one could go with- 
out a transformation of his religious state. The fasts, the contact 
with sacred rocks, the churinga,i the totemic decorations, etc., 
have gradually conferred upon him a character which he did not 
have before and which enables him to approach, without a shock- 
ing and dangerous profanation, this desirable and redoubtable 
food which is forbidden him in ordinary times. ^ 

If the act by which a sacred being is first im.molated and then 
eaten by those who adore it may be called a sacrifice, the rite 
of which we have just been speaking has a right to this same 
name. Moreover, its significance is well shown by the striking 
analogies it presents with so many practices met with in a large 
number of agrarian cults. It is a very general rule that even 
among peoples who have attained a high degree of civilization, 
the firstfruits of the harvest are used in the ritual repasts, of 
which the pascal feast is the best known example.^ On the 
other hand, as the agrarian rites are at the very basis of the 
most advanced forms of the cult, we sec that the Intichiuma 
of the Australian societies is closer to us than one might imagine 
from its apparent crudeness. 

By an intuition of genius. Smith had an intuition of all this, 
though he was not acquainted with the facts. By a series of 
ingenious deductions — which need not be reproduced here, for 
their interest is now only historical* — he thought that he could 
establish the fact that at the beginning the animal immolated in 
the sacrifice must have been regarded as quasi-divine and as a 
close relative of those who immolated it : now these charac- 
teristics are just the ones with which the totemic species is defined. 
Smith even went so far as to suppose that totemism must have 
known and practised a rite wholly similar to the one we have 
been studying ; he was even inclined to see the original source 

^ See Strehlow, III, p. 3. 

* We must not forget that among the Arunta it is not completely forbidden 
to eat the totemic animal. 

* See other facts in Frazer, Golden Bough, pp. 348 fl. 

* The Religion of the Semites, pp. 275 ft. 

340 Elementary Forms of Religious Life 

of the whole sacrificial institution in a sacrifice of this sort.^ 
Sacrifice was not founded to create a bond of artificial kinship 
between a man and his gods, but to maintain and renew the 
natural kinship which primitively united them. Here, as else- 
where, the artifice was born only to imitate nature. But in the 
book of Smith this hypothesis was presented as scarcely more 
than a theory which the then known facts supported very 
imperfectly. The rare cases of totemic sacrifice which he cites 
in support of his theory do not have the significance he attributed 
to them ; the animals which figure in them are not real totems. ^ 
But to-day we are able to state that on at least one point the 
demonstration is made : in fact, we have just seen that in an 
important number of societies the totemic sacrifice, such as 
Smith conceived it, is or has been practised. Of course, we have 
no proof that this practice is necessarily inherent to totemism 
or that it is the germ out of which all the other types of sacrifices 
have developed. But if the universality of the rite is hypothetical, 
its existence is no longer to be contested. Hereafter it is to be 
regarded as established that the most mystical form of the 
alimentary communion is found even in the most rudimentary 
cults known to-day. 


But on another point the new facts at our disposal invalidate 
the theories of Smith. 

According to him, the communion was not only an essential 
element of the sacrifice, but at the beginning, at least, it was the 
unique element. Not only is one mistaken when he reduces 
-sacrifice to nothing more than a tribute <3r-efi^eFing,-b u L- the very 
idea of an offering was originally absent from it ; this intervened 
only at a late period and under the influence of external circum- 
stances ; so instead of being able to aid us in understanding it, 
it has rather masked the real nature of the ritual mechanism. 
In fact, Smith claimed to find in the very notion of oblation an 
absurdity so revolting that it could never have been the funda- 
mental reason for so great an institution. One of the most 
important functions incumbent upon the divinity is to assure to 
men that food which is necessary for life ; so it seems impossible 
that the sacrifice, in its turn, should consist in a presentation of 
food to the divinity. It even seems self-contradictory that the 
gods should expect their food from a man, when it is from them 
that he gets his. Why should they have need of his aid in order 
to deduct beforehand their just share of the things which he 

* The Religion of the Semites, pp. 31S-319. 

^ On this point, see Hubert and Mauss, Mélanges d'histoire des religions, 
preface, p. v tï. 

The Elements of the Sacrifice 341 

receives from their hands ? From these considerations Smith 
concluded that the idea of a sacrifice-offering could have been 
born only in the great religions, where the gods, removed from 
the things with which they were primitively confused] were 
thought of as sorts oi kings and the eminent proprietors of the 
-^affh arid its products. From this moment onwards; the sacrifice 
was a.ssôciaféd with the tribute which subjects paid to their 
prince, as a price of the rights which were conceded to them. 
But this new interpretation was really an alteration and even 
a corruption of the primitive conception. For " the idea of 
property materializes all that it touches " ; by introducing 
itself into the sacrifice, it denatured it and made it into a sort of 
bargain between the man jand the divinity. ^ 

But the facts which we have described overthrow this argu- 
mentation. These rites are certainly among the most primitive 
that have ever been observed. No determined mythical per- 
sonality appears in them ; there is no question of gods or spirits 
that are properly so called ; it is only vaguely anonymous and 
impersonal forces which they put into action. Yet the reasoning 
which they suppose is exactly the one that Smith declared 
impossible because of its absurdity. 

Let us return to the first act of the Intichiuma, to the rites 
destined to assure the fecundity of the animal or vegetable 
species which serves the clan as totem. This species is the pre- 
eminently sacred thing ; in it is incarnated that which we have 
been able to call, by metaphor, the totemic divinity. Yet we 
have seen that to perpetuate itself it has need of the aid of men. 
It is they who dispense the life of the new generation each year ; 
without them, it would never be born. If they stopped cele- 
brating the Intichiuma, the sacred beings would disappear from 
the face of the earth. So in one sense, it is from men that they 
get their existence ; yet in another way, it is from them that 
men get theirs ; for after they have once arrived at maturity, 
it is from them that men acquire the force needed to support and 
repair their spiritual beings. Thus we are able to say that men 
make their gods, or, at least, make them live ; but at the same 
time, it is from them that they live themselves. So they are 
regularly guilty of the circle which, according to Smith, is implied 
in the very idea of a sacrificial tribute : they give to the sacred 
beings a little of what they receive from them, and they receive 
from them all that they give. 

But there is still more to be said : the oblations which he is 
thus forced to make every year do not differ in nature from 
those which are made later in the rites properly called sacrifices. 

* The Religion of the Semites, pp. 390 ff. 

342 Elementary Forms of Religious Life 

If the sacrificer immolates an animal, it is in order that the living 
principles within it may be disengaged from the organism and go 
to nourish the divinity. Likewise, the grains of dust which the 
Australian detaches from the sacred rock are so many sacred 
principles which he scatters into space, so that they may go to 
animate the tôternîc species and assure its renewal. The gesture 
with which this scattering is made is also that which normally 
accompanies offerings. In certain cases, the resemblance between 
the two rites may be followed even -to the details of the move- 
ments effected. We have seen that in order to have rain the 
Kaitish pour water over the sacred stone ; among certain 
peoples, the priest pours water over the altar, with the same end 
in view.^ The effusions of blood which are usual in a certain 
number of Intichiuma are veritable oblations. Just as the 
Arunta or Dieri sprinkle the sacred rock or the totemic design 
with blood, so it frequently happens that in the more advanced 
cults, the blood of the sacrificed victim or of the worshipper 
himself is spilt before or upon the altar. ^ In these cases, it is 
given to the gods, of whom it is the preferred food ; in Australia, 
it is given to the sacred species. So we have no ground for 
saying that the idea of oblation is a late product of civilization. 

A document which we owe to Strehlow puts this kinship of 
the Intichiuma and the sacrifice clearly into evidence. This is 
a hymn which accompanies the Intichiuma of the Kangaroo ; 
the ceremony is described at the same time that its expected 
effects are announced. A morsel of kangaroo fat has been placed 
by the chief upon a support made of branches. The text says 
that this fat makes the fat of the kangaroos increase.^ This time, 
they do not confine themselves to sprinkling sacred dust or 
human blood about ; the animal itself is immolated, or sacrificed 
as one might say, placed upon a sort of altar, and offered to the 
species, whose life it should maintain. 

Now we see the sense in which we may say that the Intichiuma 
contains the germs of the sacrificial system. In the form which 
it takes when fully constituted, a sacrifice is composed of two 
essential elements : an act of communion and an act of oblation. 
The worshipper communes with his god by taking in a sacred 
food, and at the same time he makes an offering to- this god. 
We find these two acts in the Intichiuma, as we have described 
it. The only difference is that in the ordinary sacrifice* they are 

* Smith cites some cases himself in The Rel. of the Semites, p. 231. 

2 For example, see Exodus xxix. 10-14 ; Leviticus ix. 8-n ; it is their own 
blood which the priests of Baal pour over the altar (i Kings xviii. 28). 
' Strehlow, III, p. 12, verse 7. 

* At least when it is complete : in certain cases, it may be reduced to one of 
its elements. 

The Elements of the Sacrifice 343 

made simultaneously or else follow one another immediately, 
while in the Australian ceremony they are separated. In the 
former case, they are parts of one undivided rite ; here, they 
take place at difterent times, and may even be separated by a 
rather long interval. But, at bottom, the mechanism is the same. 
Taken as a whole, the Intichiuma is a sacrifice, but one whose 
parts are not yet articulated and organized. 

The relating of these two ceremonies has the double advantage 
of enabling us to understand better the nature of the Intichiuma 
and that of sacrifice. 

We understand the Intichiuma better. In fact, the conception 
of Frazer, which made it a simple magic operation^ with no 
religious character at all, is now seen to be unsupportable. One 
cannot dream of excluding from religion a rite which is the 
forerunner of so great a religious institution. 

But we also understand what the sacrifice itself is better. 
In the first place, the equal importance of the two elements 
entering into it is now established. If the Australian makes 
offerings to his sacred beings, there is no reason for supposing 
that the idea of oblation was foreign to the primitive organization 
of the sacrificial institution and later upset its natural arrange- 
ment. The theory of Smith must be revised on this point. ^ 
Of course the sacrifice is partially a communion ; but it is also, 
and no less essentially, a gift and an act of renouncement. It 
always presupposes that the worshipper gives some of his sub- 
stance or his goods to his gods. Every attempt to deduce one 
of these elements from the other is hopeless. Perhaps the 
oblation is even more permanent than the communion.^ 

In the second place, it ordinarily seems as though the sacrifice, 
and especially the sacrificial oblation, could only be addressed 
to personal beings. But the oblations which we have met with 
in Australia imply no notion of this sort. In other words, the 
sacrifice is independent of the varying forms in which the religious 
forces are conceived ; it is founded upon more profound reasons, 
which we shall seek presently. 

In any case, it is clear that the act of offering naturally arouses 
in the mind the idea of a moral subject, whom this offering is 
destined to please. The ritual acts which we have described 

* Strehlow says that the natives " regard these ceremonies as a sort of divine 
service, just as a Christian regards the exercises of his religion " (III, p. 9). 

* It should be asked, for example, whether the effusions of blood and the 
offerings of hair which Smith regards as acts of communion are not real oblations 
(see Smith, op. cit., pp. 320 fi.). 

* The expiatory rites, of which we shall speak more fully in the fifth chapter 
of this same book, are almost exclusively oblations. They are communions only 

344 Elementary Forms of Religious Life 

become more intelligible when it is believed that they are 
addressed to persons. So the practices of the Intichiuma, while 
actually putting only impersonal forces into play, prepare the 
way for a different conception. ^ Of course they were not sufficient 
to form the idea of mythical personalities by themselves, but 
when this idea had once been formed, the very nature of these 
rites made it enter into the cult ; thus, taking a more direct 
interest in action and life, it also acquired a greater reality. 
So we are even able to believe that the cult favoured, in a 
secondary manner, no doubt, but nevertheless one which is 
worthy of attention, the personification of the religious forces. 


But we still have to explain the contradiction in which 
Robertson Smith saw an inadmissible logical scandal. 

If the sacred beings always manifested their powers in a 
perfectly equal manner, it would appear inconceivable that men 
should dream of offering them services, for we cannot see what 
need they could have of them. But in the first place, in so far 
as they are confused with things, and in so far as they are regarded 
as principles of the cosmic life, they are themselves submitted to 
the rhythm of this life. Now this goes in oscillations in contrary 
directions, which succeed one another according to a determined 
law. Sometimes it is affirmed in all its glory ; sometimes it 
weakens to such an extent that one may ask himself whether it 
is not going to fade away. Vegetation dies every year ; will it 
be reborn ? Animal species tend to become extinguished by the 
effect of natural and violent death ; will they be renewed at 
such a time and in such a way as is proper ? Above all, the rain 
is capricious ; there are long periods during which it seems to 
have disappeared for ever. These periodical variations of nature 
bear witness to the fact that at the corresponding periods, the 
sacred beings upon whom the plants, animals, rain, etc., depend 
are themselves passing through grave crises ; so they, too, have 
their periods of giving way. But men could not regard these 
spectacles as indifferent spectators. If he is to live, the universal 
life must continue, and consequently the gods must not die. 
So he seeks to sustain and aid them ; for this, he puts at their 
service whatever forces he has at his disposition, and mobilizes 
them for this purpose. The blood flowing in his veins has 
fecundating virtues ; he pours it forth. From the sacred rocks 

* 1 his is why wc frequently speak of the ceremonies as if they were addressed 
to hving personalities (see, for example, texts by Krichauff and Kemp, in 
Eylmann, p. 202). 

The Elements of the Sacrifice 345 

possessed by his clan he takes those germs of hfe which lie 
dormant there, and scatters them into space. In a word, he 
makes oblations. ^.., 

The external and physical crises, moreover, duplicate internal 
and mental crises which tend toward the same result. Sacred 
beings exist only when they are represented as such in the 
mind. When we cease to believe in them, it is as though they 
did not exist. Even those which have a material form and are 
given by sensible experience, depend upon the thought of the 
worshippers who adore them ; for the sacred character which 
makes them objects of the cult is not given by their natural 
constitution ; it is added to them by belief. The kangaroo is 
only an animal like all others ; yet, for the men of the Kangaroo, 
it contains within it a principle which puts it outside the company 
of others, and this principle exists only in the minds of those 
who believe in it.^ If these sacred beings, when once conceived, 
are to have no need of men to continue, it would be necessary 
that the representations expressing them always remain the 
same. But this stability is impossible. In fact, it is in the com- 
munal life that they arc formed, and this communal life is 
essentially intermittent. So they necessarily partake of this 
same intermittency. They attain their greatest intensity at 
the moment when the men are assembled together and are in 
immediate relations with one another, when they all partake of 
the same idea and the same sentiment. But when the assembly 
has broken up and each man has returned to his own peculiar 
life, they progressively lose their original energy. Being covered 
over little by little by the rising flood of daily experiences, they 
would soon fall into the unconscious, if we did not find some 
means of calling them back into consciousness and revivifying 
them. If we think of them less forcefully, they amount to less 
for us and we count less upon them ; they exist to a lesser degree. 
So here we have another point of view, from which the services 
of men are necessary to them. This second reason for their 
existence is even more important than the first, for it exists all 
the time. The intermittency of the physical life can affect 
religious beliefs only when religions are not yet detached from 
their cosmic basis. The intermittency of the social life, on the 
other hand, is inevitable ; even the most idealistic religions 
cannot escape it. ^ 

Moreover, it is owing to this state of dependency upon the""^