158. A Trobriand Medusa?

Author(s): E. R. Leach
Source: Man, Vol. 54 (Jul., 1954), pp. 103-105
Published by: Royal Anthropological Institute of Great Britain and Ireland
Stable URL: http://www.jstor.org/stable/2793580
Accessed: 12-02-2018 10:43 UTC

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 [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

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

This content downloaded from 193.48.45.204 on Mon, 12 Feb 2018 10:43:26 UTC
All use subject to http://about.jstor.org/terms
 JULY, I954 AlIan No. i58

A TROBRIAND MEDUSA?*
by

E. R. LEACH, M.A., PH.D.

Faculty of Archaology and Anthropology, University of Cambridg

hanld information. Mr. H. A. Powell, an expert in the Kiri-

158 Objects of the type illustrated in
winan language, fig.
informs me thatithis
are to
expression has no obvious
be found in a number of museum ethno-
significance. He has never heard it used in any context in
graphical collections where they are variously listed as present-day Kiriwina.
dance shields, shields and war shields. The type is exclu- Linton and Wingert (I946), pp. I44f.: 'The shields are
sively Trobriand. Published variations include: Finsch ovoid in shape with marked convex longitudinal curvature.
(I888), Plate XII; Ratzel (I896-8), Vol. I, p. 236; Edge On the background of white a strictly symmetrical design
composed of curvilinear geometric elements is painted in red,
Partington and Heape (I890-8), Part i, Plate 345; Webster black and yellow. This has the characteristic fineness of all
(I900), figs. I3, I7 (the latter also shown in Chauvet Trobriand work, but although the patterns show considerable
(I930), Plate LIII, fig. I9I); British Museum Handbook variety within a basic design they are generally highly formal-
(I9I0), p. I33; Linton and Wingert (I946), p. I48. ized and ornate and lack the vitality and verve of other objects
from this area.'
Finsch (I888), p. I3, describes the type as already rare;
In what follows two distinct hypotheses are advanced
elsewhere (I89I, p. 35) he makes it clear that these shields
concerning the nature of this design, the second being
were at least sometimes used for warlike purposes, as he
dependent upon the validity of the first. The hypotheses
mentions a specimen in which were embedded a number
are: (i) that the seemingly abstract design on the shield is
of spear points. Malinowski (I920) explains these circum-
in fact a rationally ordered representation of a winged
stances:
'Very seldom, and only in the case of very brave and
anthropomorphic figure; (2) that the figure represents a
distinguished warriors, were the shields painted. Thus during flying witch (Trobriand mulukuausi) and that the reported
the last serious war between Omarakana and Kabuaku, in mythology concerning these beings is consistent with the
i899, only two or three men had their shields painted.. . . To observable character of the shield design and also the
have one's shield painted was a challenge, since it was a great
honour to split such a shield or to kill such a man. Therefore a
magical function of the design suggested in the above-
painted shield attracted many more spears than a plain one, quoted report by Malinowski.
and it was distinctly dangerous to use this form of bravado. I will discuss these two hypotheses separately.
One of such shields used in the above-mentioned war showed If we exclude the seemingly debased example of Edge
as many as fifty-six spear marks. The warriors were decorated
Partington's Album all published illustrations of Trobriand
with exactly the same feather headdress as is used in dancing....
'One very important factor of warfare [was] . . . war vai ova shields conform to the same general design, though
magic.... In each belligerent district there was a family of details vary. Thus all examples include the feature d in
experts in war magic, whose members handed down from fig. 2, comprising several parallel lines arranged sym-
generation to generation the sacred formulx. When all the metrically on each side of the shield, but the number of
men were assembled at the chief's bidding in the main village
the magician coram publico chanted over the shields so as to
lines varies from two on each side to four on each side.
impart to them the power of warding off all spears....' Again, all specimens include a motif a in fig. 2, but whereas
Given this social context it seems intrinsically probable that in some examples this is a clearly defined face, in others it is
the design painted on the shields was itself of symbolic featureless. I shall proceed therefore to analyse fig. 2 as if it
magical significance, and it becomes a legitimate question were representative of all particular examples of vai ova
to enquire why the design shown on fig. I should in fact shields. In the key to fig. 2, left, the various design features
be deemed by the Trobrianders to have these magical are listed a-k and interpreted as anatomical details, e.g.
properties. a-face, b ears, breasts. At first sight some of these inter-
The design has on several occasions evoked comment, pretations are likely to strike the reader as surprising and
sometimes favourable and sometimes unfavourable, but, arbitrary but the analysis will be found more convincing
so far as I can discover, it has always been taken for if reference is made to the more obviously anthropo-
granted that the pattern as a whole is an abstraction and morphic figure shown in fig. 2, right. It will then be seen
non-representational. that fig. 2, left, can be derived directly from fig. 2, right,
The following comments may be noted: by, as it were, 'folding the paper.' The indications are
Finsch (i888), p. I3: 'These rare Trobriand shields are self-explanatory, but perhaps it should be added that the
remarkable for the artistic painting (red and black on a white creature is supposed to have the wings and legs of a flying
ground) and for the altogether singular design. These shields
fox, a creature resembling a bat.
perhaps represent the most perfect works of painting made
anywhere by Papuans.' Whether this interpretation will be found convincing
Haddon (1894), p. 240: 'The shape of the Trobriand shield or not will presumably depend to some extent upon the
is very characteristic, sometimes the surface is quite plain. individual temperament of the reader. It is n6t, I think, a
When ornamented the design is simply painted on the smooth
matter capable of proof one way or another. It may be
and whitened surface of the shield, with black and red pig-
ments.' noted, however, that the design in its revised form (fig. 2,
Haddon, in this same reference, states that shields of this right) has a certain resemblance to certain other Melanesian
type were known as vai ova, but this cannot have been first- designs, for example Trobriand clubs of the type figured in
* With two text figures Firth (I936), p. 96, British Museum Handbook (I925),
I03

This content downloaded from 193.48.45.204 on Mon, 12 Feb 2018 10:43:26 UTC
All use subject to http://about.jstor.org/terms
 No. i8 Man JUTLY, I954

Plate VIc (facing p. I26), and Chauvet (I930), fig. I92, Malinowski (I9I6), and Malinowski (I922), Pp. 237ff.
and the Solomon Islands shields shown in Leenhardt In both cases he stresses that the Trobriand belief closely
resembles that reported by Seligman (I9Io), Chapter
(I947), p. 44, and in Traditional Art of the British Colonies
(I949), Plate XX. XLVII, for Bartle Bay. As a result of his Mailu researches
If this interpretation of the Trobriand design be accepted,
Malinowski (I9I5), p. 648, was at first critical of Seligman's
we may proceed to enquire what the figure represents and analysis, but for the Trobriand data he seems to have
why it should be an appropriate decoration for a shield. accepted the correctness of Seligman's views. Fortune,
who reported oni very similar beliefs current in Dobu,
My hypothesis is that the figure is a 'flying witch' and that
also stresses the close parallels between the Dobuan data
observed by himself, the Trobriand data reported by
Malinowski and the Bartle Bay data reported by Seligman
(Fortune (I932), P. 297).
A Face

B Ears,. Breasts it

D C Wim
Feet, H a
E Amus Noa,I A -- ----------------
F Pu i hai-
G CIitor ris

H Buttocks P Womb

K Kne joints,

- ------A-- -----

} H ~~~~~~- - - - -- - - - - -

0 -j /~~~~~-E-- - -- -

H - - - - - - - - - - - - -

- - ~~~~~F

H -- - - - -

FIG. 2. AN INTERPRETATION OF DESIGNS ON THE

TROBRIAND SHIELD
FIG. I. SHIELD FROM THE TROBRIAND ISLANDS
Malinowski's report (I922, P. 238) is as follows:
Photograph: University Museum of Archceology and Ethnology,
'The orthodox belief is that a woman who is a yoyova can
Cambridge
send forth a double which is invisible at will, but may appear
in the form of a flying fox, or of a night bird or a firefly.
it is a shield decoration because of the poisonous emana- There is also a belief that the yoyova develops within her a
tions that are believed to be emitted by the vulva and anus something, shaped like an egg, or like a young unripe coconut.
of such witches. If this suggestion is correct, the witch This something is called as a matter of fact kapuwana, which is the
word for a small coconut. This idea remains in the native's
design on Trobriand shields had the same logic behind it
mind in a vague, indefinite, undifferentiated form.... The
as the story that Perseus carried on his shield the petrifyingly kapuwana is anyhow believed to be something which in the
beautiful head of the witch-dragon Medusa. nightly flights leaves the body of the yoyova and assumes the
Our main source of informationl concerning Trobriand various forms in which the mulukuausi appears....'
ideas about flying witches (mulukuausi) is, as might beThe parallel Dobuan belief is that 'if one sleeps touching
expected, Malinowski. The principal references are the legs of a witch a gigantic testicle within her body will
I04

This content downloaded from 193.48.45.204 on Mon, 12 Feb 2018 10:43:26 UTC
All use subject to http://about.jstor.org/terms
 JULY, I954 Man No. i58

pass over, mount

designs are mostly representational. Given this hypothesis,
elephantiasis. The
it is a reasonable proposition to attempt to interpret as func-
seen, a ball
tionally meaningful designs, such as ofthat of fig. 2, which fir
kaiana fire of
at first sight seem to be total abstractions. In this case the wi
witches as they go through the night' (Fortune (I932), evidence that can be adduced is wholly circumstantial,
pp. 297 and 296). since the objects are no longer made and memory of them
The corresponding Bartle Bay belief is in a witchcraft no longer survives in their place of origin; in other cases,
substance called labuni which the witches send forth at however, investigation of 'abstract' designs for their
night (Seligman (I9Io), p. 640): contemporary functional significance might prove very
'it was said that the labuni existed in and was derived from, rewarding. This is certainly an aspect of material culture
an organ called ipona situated in the flank and literally meaning
to which social anthropologists might usefully pay greater
egg or eggs. The labuni was said actually to leave the body and
afterwards to re-enter it per rectum. Although labuni resemble
attention.
shadows they wear a petticoat which is shorter than that worn The second point of interest is that, so far as I know, no
by the women in this part of the country.' one has previously recognized the 'folding-up' of the
The Trobrianders, like the Dobuans, believe that the human figure (as here postulated) as an esthetic device. It
flying witches can be seen as fire flying through the night, is, however, only the logical opposite of the extremely
but the precise source of the fire is not specified and the important xsthetic device of 'unfolding,' which figures so
clothing of the witch is uncertain. Malinowski says prominently in the art of the American North-West
(I922, pp. 24If): Coast (see Boas (I927), pp. 224f) and which has had a
'According to some versions, the mulukuausi, that is the powerful influence on many twentieth-century European
witch in her flying state, moves about naked, leaving her skirt
artists, including Picasso.
round the body which remains asleep in the hut. Other
versions depict her as tying her skirt tightly round her when The Trobriand technique of 'folding-up' has interesting
flying and beating her buttocks with a magical pandanus psychological implications since it permits a graphical
streamer.' representation of the association of ideas. Thus in this case
If we accept the general hypothesis that our fig. 2 there is a graphical association: ears=breasts; anus==
represents a Trobriand flying witch, and that, as a shield navel; hands-feet; headcheart; vagina womb-afolded
design, the pattern is intended as a source of dangerousflower, etc. The whole seems to provide a sort of Melane-
emanation, it will be seen that the design and the mythology sian Rorschach test which, at the very least, should be of
fit very nicely. The exaggerated emphasis given to the anal interest to Kleinian psycho-analysts interested in the
and vaginal orifices becomes meaningful, and also the 'introjection' of 'good' and 'bad' objects through the
fiery red colour of the pubes, and the claw-like character ofmedium of the witch-mother's breast!
the arms and legs. The two curious egg-like objects (fig.
References
2j) are clearly the witchcraft testicles or 'coconuts'
F. Boas, Primitive Art (Oslo, I927).
(kapuwana).
My interpretation of the upper half of the design isBritish Museum, Handbook to the Ethnographical Collections, 2nd ed.
(London, I925).
perhaps much more doubtful, especially the curious S. Chauvet, Art de Nouvelle Guine'e (Paris, I930).
design identification between the 'ears' and the 'breasts'0. Finsch, Samoafahrten: Ethnologisches Atlas (Leipzig, i888).
of the creature. It is, however, logically correct that the 0. Finsch, Ethnologische Erfahrungen und Belegstucke aus der Sudsee,
ears should be strongly emphasized in any representation Part 2, supplement to Vol. VI of Annalen des K-K-Natur-
of a mulukuausi. According to Malinowski (1922, p. historischen Hofmuseums (Wien, I89I).
24I):
R. Firth, Art and Lite in New Guinea (London, I936).
'By a special sense... 'a mulukuausi' can hear, as the natives R. F. Fortune, The Sorcerers of Dobu (London, I932).
say, that a man has died at such and such a place, or that a A. C. Haddon, The Decorative Art qf British New Guinea (Dublin,
canoe is in danger. Even a young apprenticed yoyova will I 894).
have her sense of hearing so sharpened that she will tell her M. Leenhardt, Arts de l'Oce'anie (Paris, I947).
mother: "Mother, I hear, they cry."' R. Linton and P. S. Wingert, Arts of the South Seas (New York,
I946).
Conclusion B. Malinowski, The Natives of Mailu (Victoria, I9I5).
B. Malinowski, 'Baloma; the Spirits of the Dead in the Trobriand
The interpretation which I have given to an apparently
Islands,'J. R. Anthrop. Inst., Vol. XLVI (I9I6), pp. 353-430.
abstract Trobriand design, though highly hypothetical, B. Malinowski, 'War and Weapons among the Trobriand Islanders,'
seems to me to raise a number of points of theoretical MAN, I920, 5.
interest. B. Malinowski, Argonauts of the Western Pacific (London, I922).
In contradistinction to writers such as Boas and Haddon, J. Edge Partington and C. Heape, An Album of the Weapons, Tools,
Ornaments, Articles of Dress, etc., of the Natives of the Pacific
I hold that the designs of primitive peoples are seldom Islands (privately printed, I890-8).
abstract in any genuine sense. Designs, both as wholes and F. Ratzel, The History of Mankind, 3 vols. (London, I896-8).
as parts, usually have a definite functional significance for Royal Anthropological Institute, Traditional Art of the British
the artist who makes them. Frequently of course the Colonies, catalogue of an exhibition (London, I949).
C. G. Seligman, The Melanesians of British New Guinea (Cambridge,
design element has moved very far indeed from anything
I9I0).
that might be described as photographic realism, but W. D. Webster, Illustrated Catalogues of Ethnographical Specimens
nevertheless the 'realistic' element remains. Primitive (Bicester), Vol. II; No. 20 (I900).

I05

This content downloaded from 193.48.45.204 on Mon, 12 Feb 2018 10:43:26 UTC
All use subject to http://about.jstor.org/terms