Virgin Birth

Author(s): Edmund Leach

Source: Proceedings of the Royal Anthropological Institute of Great Britain and Ireland, No.
1966 (1966), pp. 39-49
Published by: Royal Anthropological Institute of Great Britain and Ireland
Stable URL: http://www.jstor.org/stable/3031713
Accessed: 27-02-2015 12:34 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve and extend
access to Proceedings of the Royal Anthropological Institute of Great Britain and Ireland.

http://www.jstor.org

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

VIRGIN BIRTH
The Henry Myers Lecture 1966
EDMUND LEACH

King's College, Cambridge

There are three partners in every birth: God, father and
mother.-Talmud Kiddush30 b.
The so-calledprimitive ignorance of paternity is nothing else
but a very imperfectknowledgethat intercourseis a necessary
though not sufficientcondition of the womenbeing 'opened-up'
as my Trobriand friends put it.-Malinowski in AshleyMontagu (1937: 31).

THIS lecture contains three themes which follow

logically one upon another even though my presentation requires that they be slightly jumbled up. In the
first place I review the classic anthropological controversy about whether certain primitive peoples,
notably the Australian aborigines and the Trobrianders, were or were not 'ignorant of the facts of
physiological paternity' when first encountered by
early ethnographers. I conclude, as others have
concluded, that they were not. Secondly, I take note
of the fact that the anthropologist's belief in the
ignorance of his primitive contemporaries shows an
astonishing resilience in the face of adverse evidence
and I consider why anthropologists should be predisposed to think in this way. Thirdly, I suggest that
if we can once lay aside this prejudice about ignorance
and primitiveness we are left with some important
problems for investigation. Doctrines about the
possibility of conception taking place without male
insemination do not stem from innocence and
ignorance: on the contrary they are consistent with
theological argument of the greatest subtlety. If we
put the so-called primitive beliefs alongside the
sophisticated ones and treat the whole lot with equal
philosophical respect we shall see that they constitute
a set of variations around a common structural
theme, the metaphysical topography of the relationship between gods and men. Limitations of time will
prevent any elaboration of this latter theme.
The rules governing the Henry Myers Lectureship
are couched in very flexible terms but I should perhaps
make it clear that, despite my title, I am not offering a
contribution to theological debate. Professor Clifford
Geertz (1966: 35) has recently denounced my attitude
to religious matters as that of 'vulgar positivism'.
This intended insult I take as a compliment. Positivism is the view that serious scientific enquiry should
not search for ultimate causes deriving from some
outside source but must confine itself to the study of
relations existing between facts which are directly
accessible to observation. In consequence of this
limitation, positivists, whether vulgar or otherwise,
usually show signs of knowing what they are talking
about, whereas theologians, even when disguised as
Professors of Anthropology, do not. The merit of
this essay, if any, lies in its method.
I have been provoked into tackling the particular
topic of Virgin Birth by the critical comments of
another American, Professor Melford Spiro (1966:

110-112). A summary of the issue between us will

provide a useful starting-point.
One of the first detailed accounts of Australian
aborigine attitudes to sex was W. E. Roth's description of the tribes of North Central Queensland (Roth:
1903). He concluded that his informants were
ignorant of any causal connexion between copulation
and pregnancy. He described the beliefs of the
Tully River Blacks on this subject in the following
words: 'A woman begets children because (a) she has
been sitting over the fire on which she has roasted a
particular species of black bream, which must have
been given to her by the prospective father, (b) she
has purposely gone ahunting and caught a certain
kind of bullfrog, (c) some men may have told her to
be in an interesting condition, or (d) she may dream
of having the child put inside her' (p. 22 ?81).
This is a calm factual report. Ten years later
Frazer completely rephrased the above quotation so
as to ridicule the childish ignorance of the natives
concerned (Frazer 1914: 5: 102). In 1961 (Leach:
1961) I used this example to illustrate the regrettable
consequences of this habitual Frazerian practice.
I also added the following comment: '(It is not) a
legitimate inference (from this statement of Roth's)
to assert that these Australian aborigines were
ignorant of the connexion between copulation and
pregnancy. The modern interpretation of the rituals
described would be that, in this society, the relationship between the woman's child and the clansmen of
the woman's husband stems from public recognition
of the bonds of marriage, rather than from the facts of
cohabitation, which is a very normal state of affairs'.
Professor Spiro seems surprised by this formulation.
He says that he supposes it must be 'an extreme position', he declares himself to be personally persuaded
that Roth's statement must mean that the aborigines
in question were 'ignorant of physiological paternity',
and he asks 'by what evidence or from what inference
can it be concluded that . . . the statements mean
what Leach claims that they mean?' . . . and so on.
Professor Spiro is not of course a positivist. He
believes that explanation consists of postulating
causes and ultimate origins for the facts under
observation. He cannot understand the thesis that
insight comes simply from seeing how the facts fit
together. I do not expect him to agree with my
philosophy, but I am astounded by the naivete of his
anthropology. He is apparently quite unaware that
almost identical questions have beeniasked many times
before and that a huge literature of commentary
surrounds Roth's original ethnographical report.'
This suggests to me that the time has come to take
another look at this almost leigendarv controversy.

39

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

I have not much to say that is new but evidently

much that is old is now largely forgotten.
What is really at issue is the technique of anthropological comparison which depends in turn upon tlhe
kind of 'meaning' which we are prepared to attribute
to ethnographical evidence.
When an ethnographer reports that 'members of
the X tribe believe that . . .' he is giving a description
of an orthodoxy, a dogma, something which is true
of the culture as a whole. But Professor Spiro (and
all the neo-Tylorians who think like him) desperately
wants to believe that the evidence can tell us much
more than that-that dogma and ritual must somehow
correspond to the inner psychological attitudes of the
actors concerned. We need only consider the customs of our own society to see that this is not so.
For example, a high proportion of English girls go
through the rite de passage of a Church of England
marriage service. In the course of this the husband
gives the girl a ring, her veil is removed, her flowers
are thrown away, a priest lectures her on the importance of childbearing, and she has rice poured over
her head-a set of performances roughly analogous
to those reported by Roth of the Tully River Blacks.
But all this tells me absolutely nothing about the
inner psychological state of the lady in question; I
cannot infer from the ritual either what she feels or
what she knows. She may be an outright atheist.
Alternatively she may believe that a church marriage
is essential for the well-being of her future children.
Certainly her ignorance of the precise details of the
physiology of sex is likely to be quite as profound as
that of any Australian aborigine. On the other
hand, the English marriage ritual does tell the outside
observer a great deal about the formal social relations
which are being established between the various
parties concerned, and this is true of the Australian
case also.2
It would not be profitable to thrash over the details
of the Australian material again. I have given the
necessary references in footnote. I will simply make
the following summary points in justification of my
position vis-ad-visProfessor Spiro.
1. The 'vulgar positivist' interpretation which I
published in 1961 and to which Professor Spiro
takes astonished exception in 1966 was suggested
by Frazer himself in 1905 (Frazer 1905 reprinted
1910: 1: 167, 336; cf. 1910: 4: 126) and probably
derives from some still earlier source. Elaborations
of the argument have previously appeared in
Malinowski (1913), Radcliffe-Brown(1931), AshleyMontagu (1937) and elsewhere.3
2. There are two classically established reasons
for supposing that the Tully River Blacks were not
ignorant of the facts of physiological paternity in
any simple sense. These are
(a) that they freely admitted to Roth that the
cause of pregnancy in animals other than man
is copulation,
(b) Hartland (1894-6; 1909-10) assembled a vast
collection of mythological tales from all over

the world which related to the magical

conception of ancestral heroes and herodeities.4 Some of these stories resemble very
closely indeed the account given to Roth by
the Tully River Blacks of how ordinary
human births occur. Hartland thought that
such stories were survivals from a state of
Almost everyone
primeval ignorance.
would now reject such an interpretation. But
if the existence of European tales about
ladies who became pregnant after eating
magical fish is not now held to imply that
Europeans are, or were, ignorant of the facts
of physiological paternity, why should such
stories have this implication in the case of the
Tully River Blacks?
3. A third reason for rejecting the supposition of
simple minded ignorance is the judgement of the
more recent ethnographers. Meggitt (1962) for
example remarks that the answers which a Walbiri
makes to questions about conception depend upon
who is asked and in what circumstances. 'In ritual
contexts, men speak of the action of the guruwari
(spirit entities) as the significant factor; in secular
contexts they nominate both the guruwari and
sexual intercourse. The women, having few ritual
attitudes, generally emphasise copulation' (Meggitt
1962: 273).
4. Outside Australia the only society for which
'ignorance of physiological paternity' is commonly
thought to have been well established is that of the
Trobriand Islands. Malinowski's original statements on this point were very dogmatic. He asserted that 'knowledge of impregnation, of the man's
share in creating the new life in the mother's womb,
is a fact of which the natives have not even the
slightest glimpse'. But later on he became much
more guarded. The 1932 version was: 'The
Trobriandersdo ilot suffer from a specific complaint
ignorantiapaternitatis. What we find among them
is a complicated attitude towards the facts of
maternity and paternity. Into this attitude there
enter certain elements of positive knowledge, certain
gaps in embryological information. These cognitive ingredients again are overlaid by beliefs of an
animistic nature, and influenced by the moral and
legal principles of the community. . .' (Malinowski
1932: 21). As much surely could be said of any
people in the world?
We should also note that in his original account
Malinowski stated that the Trobrianders, like the
Tully River Blacks, recognised the significance of
copulation in animals though not in men. Moreover, as a result of investigations carried out in
1951, Dr Powell has reached conclusions strikingly
similar to those of Professor Meggitt for the
Walbiri and these findings are supported by
independent observations recorded in Austin (1934).5
Despite the fame of Malinowski, the Trobriand
Islands do not provide a supporting case illustrating
the possibility of total ignorance of the facts of life.

40

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

Where does this get us to? Ignorance is a relative

matter and obviously we are all ignorant in some
degree particularly about sex. But I think that
anyone who reads through the Australian ethnographic evidence with a mind reasonably free from
prejudice must agree, not only that the balance of
evidence is now heavily on the side that the aborigines
were not 'ignorant about the facts of paternity' in any
simple sense, but that the balance of evidence has
always been that way. That being so the fact that a
long line of distinguished anthropologists, which
includes Frazer and Malinowski as well as Professor
Spiro, have taken the opposite view is a very intriguing
fact. Spiro's attitude is in fact typical. He is
positively eager to believe that the aborigines were
ignorant and he accepts their ignorance as a fact
without investigating the evidence at all; at the same
time he displays an extreme reluctance to believe that
the products of aboriginal thought can be structured
in a logical way. This, of course, is a very ancient
tradition. In anthropological writing, ignorance is
a term of abuse. To say that a native is ignorant
amounts to saying that he is childish, stupid, superstitious. Ignorance is the opposite of logical rationality; it is the quality which distinguishes the savage
from the anthropologist. When Professor Spiro
writes that 'Religion persists because it has causesit is caused by the expectation of satisfying desires'
(Spiro 1966: 117) he is simply rephrasing the old
Frazerian argumentthat hope springs eternal even in a
context of total illusion. It is the same argument as
that which affirms that if men perform magical rites
before the start of a fishing expedition it is because
they are deluded into believing that fish can be influenced by words and actions at a distance and they
go on believing this because they desire to catch fish.
In other words religion and magic persist in a context
ofi gnorance.
Now the interesting thing about this argument, as
exploited by Frazer and also by the neo-Tylorians, is
that it is applied only to primitive contexts. Frazer,
who had such unbounded contempt for the ignorant
savage magician, offered no objection to the nightly
recital of Latin grace in Trinity College Hall. One
cannot legitimately interpret the reading of grace as
evidence that the parties conicerned are either superstitious or devout; why then should we make different
assumptions when 'ignorant natives' engage in
'meaningless ritual'? If an Australian aboriginal
woman announces her pregnancy by bringing into
camp a frog of a particular kind or by vomiting after
taking food from her husband this is not evidence
that she believes these actions to be the 'cause' of her
pregnancy in a physical sense. They are signs not
causes. Grace in a college hall 'says' that the meal is
about to begin or that it has just ended; the actual
word content is totally irrelevant. Similarly the
ritual actions described by Roth serve to 'say something' about the social situation and the social condition of the parties involved, they do not express the
sum of aboriginal knowledge.

I am here reminded of the fact that while the

ordinances of the University of Cambridge indicate
that young men obtain degrees because they sit
examinations, they do not specify that any prior
knowledge is required to achieve this desired end.
Yet Cambridgeundergraduatesare not ignorant of the
realities of the situation. The relationship between
ritual and copulation in aboriginal theories about
the causes of pregnancy appears to be of a strictly
analogous kind.
But in any case, as I have said already, what seems
to me interesting is not so much the ignorance of the
aborigines as the naivete of the anthropologists. It
seems evident that Western European scnolars are
strongly predisposed to believe that other people
should believe in versions of the myth of the Virgin
Birth. If we believe such things we are devout: if
others do so they are idiots. It is into this aspect of
the matter that I should now like to enquire.
Let me first state my personal bias. I find it highly
improbable on common-sense grounds that genuine
'ignorance' of the basic facts of physiological paternity
should anywhere be a cultural fact. It is true that
since human gestation lasts nine months and the first
evidence of pregnancy can only be experienced
several weeks after the act of intercourse which caused
it and since in any case intercourse is a necessary but
not sufficient cause of pregnancy, it is by no means
absurd to suppose that there might be human groups
which are quite ignorant of the role of the male; but
consider the probabilities.
Human beings, wherever we meet them, display
an almost obsessional interest in matters of sex and
kinship. Presumably this has always been the case.
Human beings have existed on earth for a very long
time during which they have displayed collective
problem solving intelligence of an astoundingly
high order. Ethnography admits that with a very few
exceptions all cultural communities now existing are
fully aware of the physiological connexion between
copulation and pregnancy. The exceptional groups,
which are alleged to be ignorant of this connexion,
appear to be fully the equal oftheir more knowledgeable
neighbours in such matters as technical ingenuity,
complexity of kinship organisation and so on.
Moreover, the allegedly 'ignorant' groups are not
living in lonely isolation in some fabulous ethnographical Shangri La, they are groups which have close
political and economic ties with other peoples who are
not 'ignorant of physiological paternity'.
My inference from all this is one of scepticismn. If
certain groups, such as the Trobrianders, have
persuaded their ethnographersthat they were ignorant
of the facts of life, then it is because that 'ignorance'
was for these people a kind of dogma.6 And if the
ethnographer in question believed what he was told
it was because such belief corresponded to his own
private fantasy of the natural ignorance of childish
savages.
We meet with alleged ignorance of physiological
paternity among peoples whom ethnographers rate

41

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

and would therefore be a much later development.

The evolutionist doctrine that systems of matrilineal
descent represent an 'earlier stage' in the evolution of
human society than systems of patrilineal descent
was tied in to this idea that matrilineal kinship is more
obvious than patrilineal kinship. It was not argued
that ignorance of physiological paternity must fow
prevail in all matrilineal systems, but onilythat it must
have prevailed in the 'original' human societies and
that these original societies would on that account
have been matrilineal. This whole set of ideas
predisposed ethnographers who were searching for
'very primitive' peoples to think they might discover
matriliny and ignorance of physiological paternity
in close conjunction. Moreover, even when the
matriliny wasn't obvious, this particular type of
ignorance could be taken as the ultimate mark of
primitiveness, and would thus confirm the atnthropologist in his hope that he had discovered a fossilised
living specimen ot primeval man-which is precisely
what the ethnographers of Australian aborigines
imagined they had discovered.
Evolutionism in its crude nineteenithcentury form
is no longer the fashion but versions of the evolutionist
argument are constantly being revived. This ties in
with the fact that the quest for the ultimate primitive
who is quite differentfrom civilised man appeals very
strongly to certain anthropologists. My own prejudices go all the other way. The data of ethnography
are interesting to me because they so often seem
directly relevant to my own allegedly civilised experiences. It is not only the differences between
Europeans and Trobrianders which interest me, it is
their similarities. This of course is a Malinowskian
precept, but let us try to put it into practice. Let us
go back to the Christians.
In the Christian case a careful distinction is made
between Jesus' legal status as a man and his essential
nattureas a god. As a man he is the legal son of
Joseph the husband of Mary and in this legal sense
he belongs to the lineage of David. In contrast, his
divine essence derives from the fact that the male
component of his conception was 'the holy spirit'
which entered Mary's body by an unnatural route.
The details of this were at one time the subject of much
learned speculation, the general consensus being that
Mary was impregnated through the ear.
The distinction between legal status and substance
appears also in the mnatrilinealTrobriand case in the
reverse sense. A Trobriand child is of the same legal
lineage as the holy spirit (the baloma) which magically
enitersthe mother's body by an unnatural route at the
monmentof conception but the child's human substance and appearance derives from the mother's
husband.
However, wh.enthe child is divine ratherthan human
there is another factor which comes in besides the
split between pater and genitor and that between
descent and filiation. In the theology of Christianity
it is not sufficient that Jesus as mnediatorshould be
ambiguously both human and divine, Mary must also

as veryprimitive. This 'ignorance' is deemed a mark of

the 'primitiveness'. In contrast the miraculous
birth of divine or semi-divine heroes is a characteristic
of the mythology of the 'higher' civilisations. Dioniysos, son of Zeus, is born of a mortal virgin, Semele,
who later became immortalised through the intervention of her divine son; Jesus, son of God, is born of a
mortal virgin, Mary, who . . .; such stories can be
duplicated over and over again. They do not indicate
ignorance.
The Frazer-Hartland generation of anthropologists
tended to adopt two mutually inconsistent attitudes
to such stories. On the one hand since virgin birth
is plainly a non-rational concept, the stories could not
have been invented by sensible civilised people-they
were survivals from an earlier primitive stage of
society. On the other hand it was implied that
theology of the 'higher religions' was not amenable to
anthropological investigation at all. Only the Catholic Fathers associated with the journal Anthropos
could consider the possibility tnat the religions of
primitive peoples might have theological merit in
their own right.
It is a striking fact that the five volumes which
Hartland devoted to the discussion of Virgin Birth
(Hartland 1894-96, 1909-10) contain scarcely a single
reference to Christianity and the corresponding
volumes of The golden bough (Frazer 1906), despite
their cynical tone, make no attempt to fit the details
of Christian theology into a cross-cultural schema
which also includes 'primitive' materials.7
Now in its Christian context the myth of the Virgin
Birth does not imply ignorance of the facts of physiological paternity. On the contrary, it serves to
reinforce the dogma that the Virgin's child is the son
of God. Furthermore, the Christian doctrine of the
physical-spiritual paternity of God the Father does
not preclude a belief in the sociological paternity of
St Joseph. Medieval Christians thought of St Joseph
as a cuckold . . . 'Joseph was an old man, an old man
was he' . . . but the authors of the Gospels of St
Matthew and St Luke combine their account of the
Virgin Birth with a pedigree which places Jesus in the
direct line of patrilineal descent from David through
Joseph. In other words the kind of interpretation
which I put on Roth's evidence and which Professor
Spiro finds so novel and unacceptable has been
orthodox among Christians for about 1600 years.
The myth, like the rite, does not distinguish knowledge
from ignorance. It establishes categories and affirms
relationships.
Ethnographers of the late nineteenth and early
twen-tiethcenturies were predisposed to discover cases
of ignorance of physiological paternity by their reading
in the theoretical works of MTcLennan, Morgan et al.
McLennan's fantasy of the begirnningof civilisation
was of a society in which men mated promiscuously
and the only kinship recognised was that through
females. Common motherhood, he argued, is a fact
very readily recognised whereas the recognition of
paternity would requirereflection and rational thought
42

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

function as a mediator and must therefore have

anomalous characteristicswhen considered as a human
being. And what could there be that is more anomalous thani a human being
who is sinless and a
mother who is a virgin?
This is a point of some subtlety. If we take the
whole range of materials relating to what we consider
to be supernaturalbirths we find, at one extreme, cases
of the Trobriandtype, where virgin births yield normal
children from normal mothers. Next, we get myths
of magical pregnancy in which, say, an old woman
long past the age of childbearing is finally granted a
child-as with the biblical stories of the birth of Isaac
or of John the Baptist. Here the implication is that
while the child is predestined to be a hero, the mother
will remain a normal human being. Finally, at the
other extreme we meet with virgin mothers of the
Christian type where both child and mother are
thoroughly abnormal.8
Theologians delight in paradox of this kind and the
trouble with early anthropological discussion of the
topic was simply that the anthropologists could not
bring themselves to believe that their primitive informants might be armed with the ingenious sophistication of a Jesuit priest. But can we go any further
that that? Can we offer any general explanation as to
why people should maintain a dogma which seems to
reject the facts of physiological paternity, or is each
case peculiar to itself ?
Tylor, Frazer and the latter day neo-Tylorians
assume that statements of dogma start out as mistaken
attempts to explain cause and effect in the world of
nature. Dogma then persists because these mistaken
ideas satisfy psychological desires. As a vulgar
positivist I repudiate such speculation about causes
which are inaccessible to observation or verification.
It may seem surprising that men persist in expressing
formal beliefs which are palpably untrue but you won't
get anywhere by applying canons of rationality to
principles of faith. All that the analyst can usually
do is to observe the circumstances in which the untrue
dogma is now affirmed and study the context of this
affirmation in other ways. As with the recital of
grace in College Hall we learn what the recital
'means' by studying the situation not by studying
the words.
Is there then anything which a dogma of Virgin
Birth 'says' about the society in which it is affirmed?
Well, let us consider first the simple logic of the
matter. The Christian myth is compatible with a
social system that is essentially patriarchal, in which
it is taken for granted that the rulers are so vastly
superior to the ruled that class difference almost
ossifies into caste, a society in wlhich the lords never
marry into the lower classes, but in which they will
graciously deign to take slave concubines and elevate
their sons to the ranks of the elite. Such societies have
in fact repeatedly emerged in Christendom, notably
in Byzantium and eighteenth century Brazil, both
countries where the cult of the Virgin was exceptionally
well developed. It would need a lot of careful research

to discover whether this correlationi is other thani

accidental but it does seem to be a striking feature of
Catholic colonialism (which distinguishes it sharply
from the Protestant variety) that the rulers, with their
bias towards Mariolatry have tended to pull their
half-caste sons into the ranks of the elite. In contrast,
the Protestant colonists who generally speaking tend
to reject the myth of the virgin birth have always
pushed their bastards into the ranks below, insisting
that the status of ruler-god is exclusive to the pureblooded. God and Jesus fit well enough into the
English Public School ethos; the Virgin-Mother has
no place at all.
I fully realise that many people will find this sort of
treatment of Christian ideology quite shocking and
even more objectionable than Weber's celebrated
tie-up between Protestantism and the rise of Capitalisnm,but I am quite serious. The British ninieteenith
century evolutionist anthropologists were mostly
Presbyterian Scots soaked in a study of the classics
and sharing, as far as one can judge, most of the
paternalistic imperialist values characteristic of the
English ruling class of the period. Their theories
reveal a fantasy world of masterly men who copulated
indiscriminately with their slave wives who then bore
children wlho recognised their mothers but not their
fathers (McLennan 1865: ch. 8). This fantasy had
some indirect resemblance to features of American
chattel slavery, but it bears no resemblance whatever
to the recorded behaviour of any known primitive
society or of any known species of animal. It was
justified simply by a priori reasoning. In the beginning men would have been unable to recognise a
causal connexion between coitus and parturition so
that although men would dominate and satisfy their
lusts by violence the only form of recognised kinship
would be consanguinity tnrough female links. The
result was a theory appropriate to Protestant niot
Catholic imperialists.
McLennan's arguments were accepted by his friend
Robertson-Smith who passed them on to Frazer who
passed them on to a host of admiring ethnographic
correspondents. The whole argument was recapitulated in quite explicit form by Hartland and swallowed
hook, line and sinker by Malinowski. That was back
in 1913 and, as we have already seen, Malinowski
later modified his position very considerably. But he
retained his high regard for aristocracy and even at
the end of his life he still thought of 'culture contact'
as a kind of patronage extended by paternalistic
colonial powers towards their more primitive subjects.
The ignorance of the Trobrianders was a necessary
element in their continuing primitiveness.
By 1932 Malinowski's theory of Trobriandignorance
had been partly abandoned but there was still the
matter of Trobriand mythology. There was for
example, 'The marvellous land of Kaytalugi ("satiated
copulation") peopled exclusively by sexually rabid
women. They are so brutally profligate that their
excesses kill every man thrown by chance upon their
shores, and even their own male children never attain

43

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

theological propositions can only serve as data not

as explanation. So we are pushed back to vulgar
positivism. What sort of positivist analysis is
appropriate to the sort of ethnographical data which I
have been discussing?
The method which I advocate is the one which
Levi-Strauss calls 'structuralist'. Structuralism (in
this sense) entails fitting the pieces together to form a
pattern. The pieces in each pattern must come from
a single context and we cannot accept the technique of
Frazer's comparative method where snippets of
evidence were drawn from here, there and everywhere. On the other hand we do not need to go the
functionalist whole hog and describe all the evidence
from every context. Our aim is still comparative,
we want to distinguish the variety of forms in which
a single ethnographical pattern can manifest itself and
then examine the nature of these variations.
The following examples show the importance of
seeing such patterns as a whole rather than as a
summation of odd elements. Frazer devoted two
whole volumes of Thegolden bough to discussing the
theme of Virgin Motherhood in a general muddle of
corn-spirits, fertility goddesses, dying gods and whathave-you. He felt certain from the start that the
virgin mother goddess was a survival from a matrilineal stage of social evolution. When he met with
contrary evidence he either suppressed it or else
invented quite imaginary forms of social organisation
to account for the discrepancy. At one point he
twisted some remarks made by the Sicilian Greek
Diodorus concerning the status of Egyptian women so
as to demonstrate that the ancient Egyptians had a
system of matrilineal descent in which men married
their sisters and succession passed from a man to his
own son in the dual status of sister's son and daughter's husband (Frazer 1914: 5 44n; 6: 213 sq.; cf.
Diodorus 1: 27: 1-2). I regret to have to report
that our President's justly celebrated essay on the
structure of unilineal descent groups overlooked
this interesting system (Fortes 1953).
Frazer however suppresses the same Greek
author's remark that 'The Egyptians hold the
father alone to be the author of generation and the
mother only to provide a nest and nourishment for
the foetus', which is clearly an extreme patrilineal
doctrine (Diodorus 1: 80: 3-4). So also, when Frazer
comes to spread himself in a mildly salacious account
of south Indian temple prostitution (Frazer 1914: 5:
63) he suppresses a large part of the evidence because
it conflicts with his basic matrilineal proposition, and
also with his theory that the temple prostitute is rated
as a goddess incarnate (Frazer 1914: 5: 71). By
rescuing Frazer's suppressed evidence from oblivion
I can perhaps display the merits of a structuralist
procedure.
The locale is south Kanara. As in neighbouring
Kerala most of the ordinary castes are matrilineal.
The Brahmans are patrilineal and consider this fact
to be an index of their social superiority. The original
account comes from Francis Buchanan writing in

maturity before they are sexually done to death.

Yet these women are very prolific, producing many
children, male and female', Malinowski (1932: 156).
Notice how closely this Trobriand version of the Land
of the Amazons resembles the anthropologists' own
fantasies. McLennan's imagination ran to rabid
men raping promiscuously their servile females; the
Trobrianders'fancy devised a world of rabid women
raping promiscuously their servile males. McLennan's dream emerged from the context of a patriarchal
ruling class which expressed horror at the thought of
any marriagebetween an upper class male and a lower
class female; the Trobriandversion belongs to a society
with virilocal matriliny, likewise class stratified, in
which women are given as tribute to political leaders.
Professor Spiro will be wholly unimpressed. He
will still ask: But how can you prove that these associations of facts are relevant? Well, quite frankly, I
don't claim to prove anything at all. In my vulgar
positivist fashion I just want to put the pieces of the
jig-saw together. When the pieces fit, I am interested.
I think this sort of thing tells us something about
Trobrianders and also about factors which have
influenced the development of anthropological theory.
But I agree that we learn no more about the Trobriander's factual ignorance than we do about McLennan's.
But I still haven't fully explained what I am trying
to say. Since the whole business of belief in virgin
mothers and ignorance of paternity has been the
subject of a vast literature, what can be the point of
my just recapitulating odd snippets?
Well partly I am interested in the problem of
method. We are dealing here with statements which
we know to be untrue; how should we interpret
ethnographical statements about palpable untruth?
There are various kinds of answer which can be
offered to such a question. If we are Tylorians, we
accept statements of belief at their face value. We
can then follow the footsteps of Frazer and Hartland
and assemble huge archives of apparently similar
untrue beliefs from all over the world. We then have
to ask: Why do all these people believe in something
which is untrue? And if you imagine that there must
be a straightforwardsingle answer to such a question
I think your only way out is to say that these false
beliefs rest on childish ignorance. This was the
answer offered by the evolutionists and by the Malinowski of 1913 and, in modified form, by Professor
Spiro in 1966.
If we are not Tylorians we can say what Powell's
Trobrianidinformants said. There are different kinds
of truth. Which is also what good Catholics say:
'We know that virgins do not conceive; but we also
know that the Holy Mother of God was and ever shall
be an immaculate Virgin.' This is the sort of answer
offered by religious people and it is an answer which I
can respect. But I do not think it is the sort of
answer which should be offered by professional
anthropologists in the course of their professional
duties. We are social analysts not theolorgians.
From an anthropological point of view non-rational
44

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

Ancestor) who is the recognised genitor of the child

while it is the human husband (who alone has sexual
access to the woman) who is the pater-a paradoxical
reversal of the ordinary pater-genitor distinction. In
the Trobriand case the impregnating spirit children
come from Tuma, the other world. They originate
as deceased ancestors of the mother's sub-clan. In
sharp contrast to all the other cases discussed the
mother and her child are here intrinsically relatedthey are of the same sub-clan throughout eternity.
There is no deity who is 'other than' the mother, and
tne woman's husband has no status vis-a-vis the son
either as recognised genitor or recognised pater.
But the husband provides the physical substance of
the child, and it is the husband who alone has sexual
access to the woman.
If we studied other examples of Virgin Birth we
should meet with the same variables in new combinations. We should also meet with new variables.
For example, in the Attis/Adonis mythology discussed
by Frazer there is a major reversal. The god is
female and the recognised genitor a human being, but
many of the other elements are similar to the Indian
and Christian cases. I cannot here pursue these
ramifications nor discuss in detail where they may
lead to. I would merely remark that themes of
descent filiation and sexual and/or marital alliance
between gods and humans necessarily have basic
relevance for the symbolisation of time and for our
topographical apprehension of the other world.
Let me try to pull the various threads of my argument together. The crux is this. From many
sources we learn of legends, traditions, ritual practices
which seem to imply a belief that women may sometimes be made pregnant by means other than insemination by a human male. The simplest way of
'explaining' such a belief is to say that it is due to the
ignorance of the believer. Many anthropologists
have argued this way and some are still inclined to do
so. They seem to gain reassurance from supposing
that the people they study have the simple-minded
ignorance of small children. That Frazer should have
thought that way is understandable, that my contemporaries should do so is extraordinary.
An alternative way of explaining a belief which is
factually untrue is to say that it is a species of religious
dogma; the truth which it expresses does not relate to
the ordinary matter-of-fact world of everyday things
but to metaphysics. It is plain, for example, that
Christians who say that they 'believe' in the doctrine
of the Virgin Birth or in the closely related doctrine of
the Immaculate Conception are not ordinarilyarguing
from a position of ignorance; on the contrary these are
doctrines which are compatible with positions of
extreme philosophical sophistication. This type of
explanation is to be preferred to the other. Frazer's
childish savage should be eliminated from anthropological discussion once and for all; in his place we
should put a slightly muddle-headed theologian no
less ingenious than the Bishop of Woolwich or even
the cleverest of my anthropological colleagues.

1801 (Pinkerton 1808-14: 8: 749). He tells us that

respectable married Brahman women when widowed
or neglected by their husbands might become wives of
the deity. The marriage rite consisted sinmplyof
eating some of the rice which had been offered to the
deity. If the woman then chose to live in the temple
she had the status of a sort of priestess and was fed and
clothed at the temple's expense. She was free to have
sexual relations with any Brahman but ordinarily
became the permanent concubine of a person of high
standing such as a government official. Her children
belonged to a special respectable section (Stanika) of
a special caste called Moylar. Women of nonBrahman caste could take up a similar profession but
had to live outside the temple; their children also were
Moylar but of an inferior grade. Within the general
caste of Moylar there were a number of distinct
endogamous sub-groups. Succession among all Moylar was from father to son but each resulting patrilineage followed the caste custom of its matrilineal
founding ancestress. The ordinary Brahmans, while
professing to despise the Moylar, in fact behaved
towards them as if they had high ritual status.
In this case the mother is a widow, the legal wife of
a god. She is made pregnant by the action of a human
being. The son is the founder of a patrilineal
descent line originating in his mother but having
status qualities derived from the mother's divine
husband. Although the mother and the physiological
father may both be Brahmans, the son is not legally
related to either of them since he is of separate caste.
This last detail creates complications and the theory
that the Moylar son should rate higher than the
Brahman father is disputed.
Taken together these facts add up to a pattern of
some complexity. But notice how the Christian data
form a very similar pattern with all the elements
reversed. In the Christian case the mother is a
Virgin not a widow. She is the legal wife of a human
being, not of a god. The Son is the last not the first
member of his patrilineal line. Although the mother
and the legal father are both human beings the Son is
substantially related to neither of them since he is a
god. This last detail creates complications and the
theory that the mother is a lower order of being than
the Son is disputed.
But we face the same kind of question as before.
What is the point of arranging the facts in this way?
How do I know that such patterns are significant?
I don't. I find them interesting. If we go back to
the Australian and Trobriand cases the data do not
add up to the same identical pattern but some of the
same variables recur.
In Australia the impregnating spirit child, like the
Holy Ghost, originates in God (the Rainbow Serpent).
In Australia the marriage ritual entails the woman
accepting a gift of food from her human husband; in
south India the woman eats food from the shrine of
her divine husband. In Australia the ideology of
filiation resembles that of Christianity. It is God (in
the form of the Rainbow Serpent and/or the Totemic
45

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

The problem with which this muddle-headed

theologian is concerned lies at the core of speculative
philosophy. What is the difference between the
physical and the metaphysical? One way of viewing
the matter is to equate the not-now with the other
world; in that case past and future coalesce as attributes of the other in contrast to the present which is the
factual experience of real life. The relationship
between the 'here-now' and the 'other' can then be
seen as one of descent. My ancestors belong to the
'other' category and so do my descendants. Only I
am in the here and the now. Of course, there is much
more than that to the matter of alternating generations
and to the tendency to put grandparents and grandchildren into one kinship category. All I am saying
is that social distance in time and space and generation
is very frequently and very readily dovetailed in with
a distinction between the living and the dead, and this
is relevant for an understanding of the materials we
have been discussing.
The relationship between the here-now and the
other can also be represented in other ways, for
example as one of class status and power-the gods
are perfect and powerful, men are imperfect and
impotent; or as one of normality and abnormalityhence the supernatural births and immortality of
divine beings.
But the disjunction of the two worlds is not enough,
there must also be continuity and mediation. Cross
cutting the idea that impotent men are the descendants
of potent gods we have the incestuous dogma that gods
and men may establish sexual connexion. Dogmas
of virgin birth and of the irrelevance of human male
sexuality appear as by-products of such a theology;
I do not claim that they are thus caused, but this is

where they fit in, and this Is the case in primitive as

well as in sophisticated societies.
This lecture has necessarily been very incomplete
but the main points which I would like to put across
are these:
1. Anthropological theories often tell us more
about the anthropologists than about their subject
matter.
2. Let us remain sceptical and positivist. Try to
see connexions between the facts as we know them.
Don't inject magical causal explanations from
outside.
3. It is time that we finally abandoned the
traditional distinction between the stupidity of
savages and the theology of civilised men. Stories
about ignorance of paternity among primitive
peoples are of the same kind as stories about the
virgin birth of deities in the so-called higher
religions. And in neither case are the story-tellers
stupid. If we are to understand such stories we
need to consider them all together as variations on
a single structural theme.
4. If anthropologists are to justify their claim to
be students of comparative religion, they heed to be
less polite. So far they have shown an extraordinary squeamishness about the analysis of
Christianity and Judaism', religions in which they
themselves or their close friends are deeply involved.
Roth's Bulletin No. 5 on the North Queensland
Aborigines was an ethnographic document of
considerable interest; so is Chapter 1 of the Gospel
according to St Matthew. Serious anthropologists
should treat the two works on a par; both are
records of theological doctrine.

NOTES
(1937). Schmidt criticises Ashley-Montagu's conclusions but does not cite any additional material.
Those who accepted Hartland's position relied
principally on evidence provided by Spencer &
Gillen (1899: 122-5) and Roth (1903: 16-23) and
resolutely rejected evidence which had a contrary
implication, e.g. Purcell (1893: 286-9), Strehlow
(1907: 2, 52, note 7), Spencer & Gillen (1899: 265).
Nearly all the more recent detailed ethnography leads
to the conclusion that the formally expressed ignorance
of physiological paternity is a kind of religious fiction
-see in particular Warner (1937: 24), Thomson
(1933: 506), Roheim (1932: 96-7), Sharp in AshleyMontagu (1937: 162-3), Stanner (1933: 27-28), Meggitt (1962: 273). The case of Kaberry is discussed in
greater detail below. Those who have maintained a
version of the Hartland position down to recent times
-and these include Malinowski, Ashley-Montagu,
Stanner (1933) and Kaberry (1936, 1939)-have shown
themselves willing to accept even the flimsiest evidence
for the fact of ignorance, while evidence against,
even when it is most meticulously recorded, is repeatedly rejected on the ground that it must be due to

1. The literature on this topic is very large. Relevant

source material down to 1936 is quoted in AshleyMontagu (1937: 349-155). The details of the controversy as it developed from 1894 onwards are laid out
in chapters 1 and 10 of the same work. The two basic
standpoints remained unchanged throughout: Hartland and his followers held that on logical grounds
there must have been a time when primitive man did
not recognise a connexion between copulation and
pregnancy; Australian aborigines are in this respect
examples of primeval man surviving into modern
times. The other group which included Andrew
Lang, Westermarck, Pater Schmidt and Goldenweiser denied that the actual evidence suggested that
the Australian aborigines were ignorant in the sense
suggested. The general consensus of anthropological
opinioin in the period before 1914 agreed with Hartland. The adherents of Hartland's opinion included
Frazer and Malinowski (1913). The last-named
however was already concerned to show that an ideology of ignorance could have sociological significance.
Schmidt (1952) seems to be the only general review of
the subject which is later than Ashley-Montagu
46

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

child, who will take the country and the totems of the
woman's husband. There were also instances where
although the husband himself had found the djerin
he did not afterwards dream of the child. But his
wife would then assert that she had done so.' 'Questioned on the function of sexual intercourse natives
admitted that it prepared the way for the entry of the
spirit child. They asserted that a young girl couild
not bear children.' [Compare the similar materials
in Malinowski (1913: 210-212).]
I am very puzzled as to how anyone could interpret
such data as indicating 'ignorance of physical paternity'. Kaberry's informants are saying that conception is not predictable in advance but is recognised by
certain physiological signs after the event. They say
that sex relations are a necessary preliminary to this
condition, and they say, as do most Europeans, that
the foetal embryo has a soul. Their theological
arguments are concerned with the origin of this soul.
Since they are less attracted by the vitalistic fallacy
than are orthodox Christians, they fail to make any
precise distinction between the soul and the foetus.
Instead of saying that the semen is a constituent of the
foetus and that the soul is a metaphysical separate
entity, they speak of the foetus-soul floating in semen
'like a water-lily'. Finally they assert that the woman's legally recognised sexual partner has an exclusive status as father of the child. This set of beliefs
and attitudes seems to me to differ from that of the
British only on rather subtle issues of a metaphysical
kind, it certainly does not constitute physiological
ignorance in any simple sense. That a distinguished
anthropologist should once have thought otherwise
displays the oddity of anthropologists rather than the
oddity of the aborigines.
2. Spiro (1966: 112) maintains that the fact that a set
of ritual data are structured in a manner which is
directly parallel to a set of social relations is of no
relevance unless we have direct evidence that the
actors are consciously or unconsciously aware of the
significanceof this symbolism. Instead of looking for
patterns in the way that people behave Spiro would
adopt the naliveprocedure of asking the actor why he
behaves as he does-'and unlike some anthropologists, I believe him'. If Spiro tries this out in the case of
English marriage procedures he will get some most
astonishing results. These rituals are, as it happens,
structuredin an extremely clear and well defined way,
but not one bride in a thousand has even an inkling
of the total pattern.
3. Frazer (1910: 1: 336) 'To a Central Australian
father fatherhood means that a child is the offspring
of a woman with whom he has the right to cohabit,
whether he has actually had intercourse with her or
not. To the European mind the tie between a father
and his child is physical, to the Central Australian it
is social.' Cf. Malinowski (1913: Ch. 6); RadcliffeBrown (1931: 42); Ashley-Montagu (1937: Ch. 13).
4. Ishida (1964) reports on a wide range of oriental
materials which were not available to Hartland.
5. Powell's report was as follows: '(I was told) the

recent contaminating influence from mission stations

or other Europeans. Thus Stanner (1933) reports of
the Daly River Tribes that: 'It is clear that two
theories of sex exist side by side: (a) a mystical theory
of the type commonly found in Australian cultures and
(b) a barely understood, confused version of the
orthodox theory learnt from the whites.' Stanner has
no evidence that (b) was in fact learnt from the whites.
Thomson (1933) came down decisively on the other
side: 'I actually approached the present study with a
firmly rooted belief that the natives were entirely
ignorant of the fact of physiological fatherhood. It
was only after I was repeatedly made aware of the
facts that I became convinced of the reality of the
natives' knowledge.' Yet Kaberry (1936) could still
persuade herself that ignorance of physiological
paternity is a feature of ethnography reported from
all over Australia without exceptions.
Meggitt's statements are discussed in my main text.
Kaberry's two reports are especially interesting
because they bear rather directly on Roth's original
statement and because she herself quite uncritically
accepts the view that the native beliefs and practices
arose from 'ignorance of physical paternity'. Her
fieldwork was carried out in the Kimberley Division
of Western Australia about 800 miles due west of
Roth's 'Tully River Blacks'. All her discussions
with informants were conducted in pidgin English.
Close attention to the details of her account do not
support her theory that informants were ignorant of
the physiology of sex in any simple sense. What they
said, in quite explicit terms is that they considered
sexual intercourse a necessary but not sufficient cause
of pregnancy. A woman does not becomnepregnant
automatically as soon as she has sex relations; she
becomes pregnant because she conceives and it is
with the theology surrounding conceptionthat Roth's
account and also Kaberry's is largely concerned.
Kaberry's version (1939: 42-3) clearly indicates the
logic of Roth's story. The Kimberley natives said
that spirit children were created long ago by the
Rainbow Serpent and 'temporarily incarnated in
animals, birds, fish, reptiles. Some say the spirit
children are like little children the size of a walnut;
others that they resemble small red frogs. Conception occurs when one of these enters a woman. Its
presence in the food given her by her husband makes
her vomit, and later he dreams of it or else of some
animal which he associates with it. It enters his wife
by the foot and she becomes pregnant. The food
which made her ill becomes the djerin, conception
totem of her child.' In other words a woman recognises that she is pregnant when she experiences
'morning sickness'-which is true also of European
women.
'The husband of the woman is the social father of the
child and as a rule its spiritualgenitor, for it sometimes
happens that the woman finds the djerinherself or that
it is given to her by another man. The latter however
will not dream of the spirit child, nor have access to
the woman sexually nor exercise any rights over the
47

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

semen acts as a coagulant of the menstrual blood

producing a "clot" which a spirit child (baloma)
enters by way of the head or otherwise, and which
proceeds to grow after the "quickening" by the entry
of the baloma. My suggestions that this account
either contradicted that of Malinowski or else was
garbled "mission talk" were strenously denied by my
informants, who maintained that both accounts were
"true",but that they were "different". Malinowski's,
they said, was "men's talk" valid in formal situations,
e.g. in matters of land ownership and the like; the
account given to me was "women's and children's
talk", that is, it was what fathers or their sisters told
children as they became old enough to take more than
a childish sexual interest in the opposite sex." (Powell
1956: 277-8). Quoted with author's permission.)
6. The evidence for 'dogma' is very clear. In 1909 the
Bishop of North Queensland complained to Frazer
that ignorance of the relation between intercourse and
pregnancy 'forms a fact which has to be reckoned
with in the introduction of a higher standard of
morality among the aborigines, for they do not naturally accept the true explanation of conception and
childbirth even after their admission to the mission
stations' (Frazer 1909). In later references to the
bishop's statement Frazer cites only the above remarks
but the aboriginal reluctance to accept the European
viewpoint seems to have been a male peculiarity.
The bishop also told Frazer: 'We often have girls who
are sent to the mission enceinte and we never dwell on
any wrongfulness of their condition. We have no
trouble afterwards, neither have we found, at any rate
for many years, that the girls persist in the belief ...
that copulation is not the cause of pregnancy'. Even
more striking is Fortune's account (Fortune 1932:

239) of how he tried to stage a debate between

Dobuans who maintained the role of the father and
Trobrianders who denied it. 'But the head of every
Dobuan in the room immediately was turned away
from me towards the wall. They affected not to hear
the conversation; but afterwards when they had me
alone they were furious with me.' The argument was
plainly about doctrine not about knowledge. Theologians who debate the doctrine of transubstantiation
cannot usefully be accused of ignorance of the
elementary facts of chemistry.
7. Two pages from the end of his three volume work
Hartland remarks: 'I cannot hide from myself the
important bearing that some of the subjects dealt with
in these pages may have upon Christian controversy';
but his heretical daring gets no further than that.
Frazer is equally cautious. Frazer (1906: 349-50)
draws attention to the writings of a variety of authors
who have seen similarities between the Cult of Isis
and the Cult of the Virgin Mary, but he scarcely
comments on this point. Book 2 Chapter 1 of this
work, which is indexed 'Stories of Virgin Birth' never
mentions Christianity at all.
8. All the numerous 'mother goddesses' of the
Ancient World seem to have been classified as 'virgins'.
There may be complex problems of translation here.
In the Trobriands, a widow, after the conclusion of her
mourning, becomes a 'marriageable maiden' and is
classified as such. It is at least possible that the title
of 'virgin' accorded to such goddesses as Aphrodite
signified their sexual availability rather than their
physiological condition. Incidentally the original
Hebrew of Isaiah vii: 14 which is quoted in Matthew
i: 23 refers to a 'maiden' rather than to a 'virgin'.

REFERENCES
HARTLAND.E. S., 1909-10. Primnitivepaternity. 2 vols.

into being among the

M. F., 1937 Conming
ASHLEY-MONTAGU,

London: Folk Lore Society.

ISHIDA,
E., 1964. Mother-Sondeities. Hist. Rel. 4, 30-68.
KABERRY.
P. M., 1936. Spiritchildrenand spirit centresof the
North Kimberley Division, West Australia. Oceania,
6, 392-400.
KABERRY, P. M., 1939. Aboriginalwoman. London: Routledge.
LEACH, E. R., 1961. Golden bough or golden twig? Daedalus
Spring 1961: 371-87.
MALINOWSKI,B., 1913. The family among the Auistralian
aborigines. London: London University Press.
MALINOWSKI,
B., 1932. The sexual life of savages in north
westernMelanesia. (3rd edition with special foreword.)
London: Routledge.
MCLENNAN, J. F., 1865. Primitive marriage. London:
Quaritch.
MEGGITT, M. J., 1962. Desert people. Sydney: Angus &
Robertson.
PINKERTON,J., 1808-14. A generalcollectionof the best and most
interesting voyages & travels in all parts of the world.
17 vols. London: Longman,Hurst, Rees & Orme.
POWELL, H. A., 1956. An analysisofpresentday social structure
in the Trobriands. Thesis, London.
PURCELL, B. H., 1893. Rites and customs of Australian
aborigines. Z. Ethnol.25, 286-89.

Australianaborigines. London: Routledge.

AUSTIN,L., 1934, Procreation among the Trobriand Islanders.

Oceania5, 102-18.

R. F., 1932, Sorcerersof Dobu. London: Routledge.

FORTUNE,
DIODORUS,see Oldfather,C. H.. Diodorusof Sicily. English
translation(10 vols.) Loeb Classical Library. (1932, vol.
1: 84-87, 274-75.) London:Heinemann.
M., 1953. The structureof unilineal descent groups.
FORTES,
Am. Anthrop.55, 17-41.
FRAZER.J. G., 1905. The beginningsof religionand totemism
among the Australian aborigines. Fortnightly Rev. n.s. 78,

162-72, 452-66.

Adonis,Attis, Osiris. 2 vols. in 1 (later

Vol. 5 and 6 of Frazer 1914.) London: Macmillan.
FRAZER,J. G., 1909. Beliefs and customs of the Australian
aborigines. Man 9, 145-7.
FRAZER,J. G., 1910. Totemismand exogamy, 4 vols. London:
Macmillan.
FRAZER,J. G., 1914. Thegoldenbough, 3rd edn. 12 vols.
London: Macmillan.
GEERT,C., 1966. Religion as a culturalsystem. In Anthropological approachesto the study of religion (ed.) M. Banton
FRAZER,J. G., 1906.

(Monogr. Ass. social Anthrop. 3.) London: Tavistock

Publications.
HARTLAND,E. S., 1894-6. The legend of Perseus. 3 vols.
London: David Nutt.

48

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions

A. R., 1931. The social organisation of

RADCLIFFE-BROWN,

Australian tribes (Oceania Monogr. 1). Melbourne:

Macmillan.
ROHEIM,
G., 1932. Psycho-analysisof primitiveculturaltypes.
Intern.J. Psycho-Anal.13, 1-224.
RoTH, W. E.,

1903.

Superstition,

magic

and medicine.

(N. Queensl.Ethnogr. Bull. 5) Brisbane:Vaughan.

SCHMIDT,W., 1952. Der Konzeptionsglaube australischer
Stamme. Intern.Arch. Ethnogr.46, 36-81.
B. & GILLEN, F. J., 1899. Thenativetribesof Central
SPENCER,
Australia,London: Macmillan.
Spmo, M. E., 1966. Religion: problems of definition and
explanation. In Anthropologicalapproachesto the study of
religion(ed.) M. Banton (Monogr. Ass. social. Anthrop.3).
London: Tavistock Publications.

E. H., 1933. The Daly River tribes: the theory

of sex. Oceania4, 26-8.

STANNER, W.

1907-21. Die Aranda und Loritja Stadnne in

Zentral-Australian. Frankfurt: Baer.

STREHLOW, C.,

See EpsteinI, TheBabylonianTalmud:

Seder Nashi VIII Kiddushin (1936 p. 149, translated
H. Freedman).London: Soncino Press.

TALMUD, Kiddush30. b.

D. F., 1933. The hero cult, initiationand totemism

THOMSON,
on Cape York: the knowledge of physical paternity.
J. R. anthrop.Inst. 63, 505-10.
W. L., 1937. A black civilisation. New York:
Harper.

WARNER,

49

This content downloaded from 143.106.201.154 on Fri, 27 Feb 2015 12:34:04 UTC
All use subject to JSTOR Terms and Conditions