BIG MEN AND

GREAT MEN
Personifications of power
in Melanesia
edited by

MAURICE GODELIER
and

MARILYN STRATHERN

The righl of the

Universas of Consbridge
io prini and sell
oil marine, of boaSs
N-07 granied hc
Henry VIII in 1534.
The Universo". has proferd
and published cominuously
suite 1584.

CAMBRIDGE UNIVERSITY PRESS

Cambridge
New York Port Chester Melbourne Sydney

EDITIONS DE LA MAISON DES SCIENCES DE L'HOMME

Paris

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, So Paulo, Delhi
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK

Contents

With Editions de la Maison des Sciences de l'Homme

54 Boulevard Raspail, 75270 Paris Cedex 06, France
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521390187

List of illustrations

Notes on contributors
Pre face
Acknowledgements
Map

(C) Maison des Sciences de l'Homme and Cambridge University Press 1991
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may Cake place without the written
permission of Cambridge University Press.

Introduction
First published 1991
This digitally printed version 2008
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data

Big men and great men: personifications of power in Melanesia/

edited by Maurice Godelier and Marilyn Strathern.
p. cm.
Includes bibliographical references.
1. Political anthropology Melanesia. 2. Political leadership Melanesia.
3. Power (Social sciences) Melanesia.
I. Godelier, Maurice. II. Strathern, Marilyn.
GN668.B54 1991
306.2'09995dc 20 90-1312

MARILYN STRATHERN

PART I

1 From great men to big men: peace, substitution and

competition in the Highlands of New Guinea

PIERRE LEMONNIER

2 Great man, big man, chief: a triangulation of the Massim

28

JOHN LIEP

3 Soaring hawks and grounded persons: the politics of

rank and gender in north Vanuatu

48

MARGARET JOLLY

ISBN 978-0-521-39018-7 hardback

ISBN 978-0-521-10229-2 paperback

PART II

4 Punishing the yams:

leadership and gender ambivalence
on Sabarl Island

81
83

DEBBORA BATTAGLIA

5 Great men and total systems: North Mekeo hereditary

authority and social reproduction
MARK MOSKO

97

CHAPTER 9

The fractal person

ROY WAGNER
Wagner re-opens the Highlands material via his own Austronesian perspective
from New Ireland. He poses a question about the different kinds of anthropological
understanding that have been brought to the depiction of great men and big men.
Big men have been seen as exemplars of sociological activity, as mobilising social
forces, for they appear to change the scale of men's actions from an individual to a
group dimension by virtue of the numbers they command. But great-men systems
force us to comprehend a pre-existing sociality, and a pre-existing totality, of which
any aggregate can be only a partial realisation. This totality is neither individual nor
group but a 'fractal person', an entity whose (external) relationships with others are
integral (internal) to it. However diminished or magnified, the fractal person, keeping its scale, reproduces only versions of itself. The great man thus represents the
`scale' of his culture rather than a scale-change to accommodate anthropological
attempts to ground it in principies beyond itself. If we have here an indigenous social
science, the question becomes how then to conceive big men from the point of view
of understandings of this kind that great-men systems are able to elicit from the
western social scientists.

We are indebted to the Italian Marxist Antonio Gramsci for the notion of
hegemonic ideas (1971), of concepts that have come to be taken so much
for granted that they seem to be the voice of reason itself. Such ideas are not
subconscious or out-of-awareness for the same reason that their validity is
not subject to question; they are the very form taken by our consciousness
of a problem or issue. Hegemonic ideas, then, are no more subject to proof
or disproof than are Kuhnian paradigms, for in both cases entering the discourse is tantamount to replacing the question of whether things work that
way with one of how they work that way. Hence anthropologists with an
investment of research interest in the hegemonic motif, say, of the necessarily social dynamics of human thought, might be expected to fault and
misunderstand a challenge to the motif in terms of its failure to provide a
convincing 'how', without perceiving the irrelevante of their objections.
The opposition of individual and society, a product of western jurisprudence and political ideology, is not merely coincidental to the hegemony of
`social' thinking, but identical with it. It is based on the necessarily ideal,
and practically unrealisable, notion of the 'social concept', and the necessarily substantive, physical and material, notion of the person as object.
159

160

The fractal person

Big men and great men

Thus the ideal of `corporateness', an ostensible merging of individuals into

a single social `body', becomes, in its failure to achieve complete realisation,
a substantive group of individuals. And the notion of a totally integrated
`culture' of collective representation within the individual becomes, in its
failure of realisation, a mere `culture-concept', an ideal. The point is not
simply that a flawed and unrealistic opposition of thought and su bstance
reproduces itself as measurable social fact, that social groups and idealised
cultures are mass-produced as a map of socio-cultural variation and problematics. It is, more importantly, that a naively hegemonic dependence
upon individuality and plurality underlies and articulates the manner in
which idealised concept and substantive object are brought into play. This
dependence makes the fact-and-problem-producing failures of concept to
be fully realised, of substantive object to the conceptually tractable, seem
like stubborn fact, seem to be the very fabric of social reality.
Thus to make a statement such as 'no society works perfectly', or even
'the reason no society works perfectly is just that its members expect it to
do so' is to describe the expectations of anthropologists themselves rather
than those of their subjects. For what is described is the manner in which
social scientists work to make their subjects interesting, statistically variable and problematic. It is by no means clear that the subjects think of themselves in this way, or think of their social interactions as interesting because
they can be mapped into paradigms of social groupings and individual
varia bility.
The idea of a social mechanism or that of the individual as its natural
resistance did not grow indigenously in Melanesia; it was brought there
together with other mechanisms by self-conscious `individuals'. And so the
proposition that a society might work or not work is the same sort of surprise in indigenous terms as that an automobile engine should work or that
it might not work. But the failure of an automobile engine, or of the society
of western construction, does not entail a complete overhaul of our
assumptions about mechanics; it entails an overhaul of the engine, the
model, before the mechanics get to work. A hegemonic of individual/
society mechanics, with its underpinnings of the particular/general, shifts
automatically from questions of `why?' into questions of 'how ?'.
Hence a discovery that, at least for some Melanesians, the part/whole
distinction and its systematic entailment is inapplicable, does not automatically imply that those Melanesians belong to a race of mathematical
wizards. If such a discovery suggests that the individual problem- and
person-producing failure of social concept, and the system-producing
failure of individual autonomy, are wrongheaded constructions of the

161

wrong `engine', this may simply mean that Melanesian thinking is too
elegantly simple, rather than too complex, for western expectations. An
engine with no moving parts at least avoids the nemesis of friction. And friction may well be the effect that social scientists have mistaken for social
leverage.
Or so at least the received conception of the big man would suggest: an
emperor of social friction who uses society against itself to reinstate the
essential individual at the top of the heap. In his identification of the
phenomenon of the great man, Godelier posed a profound challenge to our
understanding of Melanesian societies. Introduced as a type or another
kind of leader, the great man provides a counter-example to the big man
that familiarity and overuse have inflated far beyond Sahlins's (1972)
sophisticated characterisation. But typology alone can only trivialise the
challenge, which takes its weight and authority largely from the context of
Baruya ethnography. For The Making of Great Men proposes a vivid
antithesis to the self-excusing notion of `loosely structured' societies that
has entertained ethnographic speculation for many years. The larger challenge is that of a more holistic manner of thought than that implied in structure, and the great man is its holistic counterpart.
Is the big man his equivalent in another kind of society, a more open,
competitive and loosely organised one? Or is this type-casting of the big
man itself the error of another way of approaching society, and therefore
not a typological contrast at all? Let us consider an ethnographic locus

classicus.
Neither individual nor group
The anthropologist has often been obliged, even pleased, to construct
social forces out of the evidence of a big man assembling, say, his resources
for a moka. As long as he can be seen to be making a kind of solidarity, helping the group to happen, the imputed sociology has an immediate and
obvious realisation. The question posed by the idea of the great man is that
of what to do when society and its solidarity are already in place. Then, of
course, the big man's efforts have to be reconsidered or re-entitled; he is not
enacting the answer to a sociological question, because that question has
already been answered. But if we should suggest that he is realising his own
individual aspirations, the projection of western political economy has
another easy answer. Sociology is then seen to emerge from the conjoint
effects of individual competition.
Anyone who has ever tried to determine the definitive locus of

162

Big men and great men

`individual' and `corporate group' in the planning and making of these

competitive exchanges, fairly soon realises that individual and group are
false alternatives, doubly so implicated because each implies the other. It is,
after all, difficult or impossible to define the successful (or unsuccessful)
maker of moka as either individual or group, because the big man aspires
to something that is both at once. One might say that the Hagen big man
aspires to the status of great man that the moka produces variant
examples, equally valid however successful or unsuccessful, of the great
man. It is a matter of the realisation of something that is already there, as
the pigs and shells are already there.
Would it make any difference, then, to argue that the status and the
society are never really there, that the image is always realised for the first
time, or even that it may never be realised at all? None whatsoever. Hagen
society is there or not there whether or not the moka is realised, the big man
remains a big man regardless of the form of his achievement. If this were
a matter of `making' society, then the failure of a moka would make a
difference.
I have borrowed an illustration from Hagen society (cf. chapter 11),
and purposely made our normal projection of motivation and agency
into its actors obligue and difficult for a very specific purpose. This is to
develop, in the course of this essay, Marilyn Strathern's concept of the
person who is neither singular nor plural. In introducing her idea, Strathern
(1990) borrowed from Haraway (1985) a most ingenious application
of the classic science-fiction term `cyborg' the integral being who is
part human and part machine. For my purposes, and for reasons that
shall become apparent presently, I shall re-entitle the concept as that of
the fractal person, following the mathematical notion of a dimensionality that cannot be expressed in whole numbers. I shall not be concerned
with the degree of fractality here, the terms of the ratio or fraction, but
simply define the concept of a fractal person in contrast to singularity and
plurality.
Although the idea of fractality may appear abstract, it is in fact no more
so than singularity or plurality, or statistical analysis. Its effects are
altogether familiar to the fieldworker as the problem, for instance, of the
big man's aspirations being at once individual and corporate. It is that
problem, apprehended as a solution. It lies at the root, too, of what is commonly misconstrued as the `extension' of kin-terms, exemplified in the
Siane usage (Salisbury 1964) whereby any daughter of a unit to which the
class of `father' had given a bride becomes a hovorafo (`father's sister's
daughter'), a potential spouse. As Salisbury correctly deduced, father is not
necessarily identified with a so-called primary kin term here, and is neither

The fractal person

163

singular nor plural. The term has a fractal implication, equally applicable
to both situations.
A fractal person is never a unit standing in relation to an aggregate, or an
aggregate standing in relation to a unit, but always an entity with relationship integrally implied. Perhaps the most concrete illustration of integral
relationship comes from the generalised notion of reproduction and
genealogy. People exist reproductively by being `carried' as part of another,
and 'carry' or engender others by making themselves genealogical or reproductive `factors' of these others. A genealogy is thus an enchainment of
people, as indeed persons would be seen to `bud' out of one another in a
speeded-up cinematic depiction of human life. Person as human being and
person as lineage or clan are equally arbitrary sectionings or identifications
of this enchainment, different projections of its fractality. But then enchainment through bodily reproduction is itself merely one of a number of
instantiations of integral relationship, which is also manifest, for instance,
in the commonality of shared language.
Is this not, then, a mere generic, a mathematical fiction like the 'modal
personality'? It would be indeed if I were concerned either to generalise or
particularise the relation between general and particular. But integral
relationship is not a matter of general and particular, nor of how one of
these might be made over into the other. The argument is not one of comparative reality or practicality, but rather one of how one's realities or practical issues are situated with respect to relationship. The only issue that
need detain us is that of how Melanesians themselves would seem to situate
them.
The issue requires evidence, and the best evidence I can think of pertains
to the way in which Melanesians indigenously speak of, order and conceptualise existence as identity. This entitlement of existence is quite simply
that of naming, for it is after all names, rather than individuals or groups,
that 'go on high' in the moka, that command awe, attention and responsibility in the Kula, that serve, as `big' or `small', for the identities of what we
are predisposed to call groups lineages, clans or whatever. Regardless of
their range of denomination, whether personal or collective, names are but
names, but it is a name that is at once the individual and collective aspiration of the big men. A Daribi friend once observed, 'When you see a man,
he is small; when you say his name, he is big'.
The example I shall use is that of Daribi naming. A Daribi name, nogi, is
always an instantiation, and also a simplification, of the relation designated by the participle, poai, of the verb poie, `to be congruent with'. Two
persons, or a person and a thing, that share a name are tedeli nogi poai, `one
name congruent'. Two beings that share the same kind of skin are tedeli tigi

164

Big men and great men

ware poai, `one epidermis congruent'. Anything designatable by a word

stands in a poai relation through any conceivable point of resemblance.
Furthermore any two persons or objects that each share any conceivable
point of resemblance with a third, are related as poai through that third.
Poai is universally commutative, and because a poai relation can simply be
bestowed, through the giving of a name for whatever reason, it is also universally applicable. Poai eats the world, and it also eats itself. For when an
infant goes unnamed for an intolerable period of time after birth, usually
out of fear for undesirable consequences of naming, it will acquire the
designation poziawai, `unnamed'. The infant acquires an immediate poai
relation with all things unnamed (non-congruent), but, of course, since
poziawai is a name, it acquires another with all things named.
The infant, in short, becomes an embodied hinge between the world of
names and that of unnamed things. And though poziawai is by no means
uncommon as a name at Karimui, this is no reason to accede to one patrol
officer's private musings that the Daribi are a prime example of negative
thought. For it turns out that the designation poai is virtually as popular as
a personal name. Unhinging as these examples may prove, they serve to
direct our attention to the social recognition of the name, the only real grip
afforded the Daribi on an otherwise frictionless surface.
Essentially, any recognition or bestowal of a name is always the fixing of
a point of reference within a potentially infinite range of relations, a designation that is inherently relational. As an instantiation of poai, it always
implies, through that relation, something that is both less (one of many
potential relations) and more (a class, a range of objects or beings) than the
person designated. A man, for instance, named for the cassowary, can
claim such words as tori, kebi and ebi as his names, since they are all equally
names for the cassowary. Also, since the cassowary is poetically and colloquially the ebi-haza, the `cassowary-animar through its non-avian proclivities, the man could well claim haza, 'animal', as a pagerubo nogi, a
(somewhat droll) basing-name or nickname. And if, as is usually the case,
the man was named for someone else, or someone else is named for him, the
name is always a section, like the conceptual person or body, taken from a
genealogical chain and implicating that chain.
Hence the particular points of convergence that other Melanesian
regimes of naming may share, or may not share, with Daribi naming are
somewhat beside the point. As long as words are polysemic (and naming,
of course, makes them so), and people relate by reproduction, any system
of identities developed by sectioning and referencing such a relational field
is intrinsically fractal apparent differentiation developed upon universal
congruence and interchangeability. And since denomination is our surest

The fractal person

165

map or model for the apprehension of identity, the case for the indigenous
conceptualisation of fractal units is manifest. It is 'individual' and `group'
that are arbitrary, imposed and artificial.
The concept of currency, money that demands accounting in terms of
singularity and plurality, is likewise a non-fractal imposition upon a regime
of exchange based on sectionings taken from human productivity and
reproductivity. Pigs, pearl shells, axes, bark cloaks are already relational
and implicated in the congruence that underlies the remaking of human
form, feeling and relationship. Shells and shell wealth (which Daribi think
of as `eggs' through which human beings reproduce) are engaged in the
reciprocity of subjectives involved in display and concealment, just as axes,
meat and other adjuncts of production and reproduction place human sustenance and replication in reciprocal exchange. When such relational
points are treated as representational, as commodity-aggregates on the
model of currency, or when the currency substituted for them is taken
literally, integral relationship is denied and distorted. Minus the congruence that keeps the scale of their essential unity through all permutations of
categorisation, names become merely representational categories of social
designation and classification. And minus the sense of their essential unity
with body and life-process (in their subjective as well as objective enhancement), items exchanged become the mere `wealth objects' of a like categorisation a `representation' of human values through utility, a 'classification' of utilities through human value.
Money, as the cutting edge of the world-system, entails the counting of a
resource-base. Where the resource is itself relational, the commodity, so to
speak, of relation, it will exert its own reflexive effect upon the terms of
assessment. Hence bridewealth and childwealth inflate prodigiously in the
attempt to make assessment into a form of relating, spending representational literalism in the service of what is fundamentally a rhetoric of
assertion.
Is the `economic' image of the big man merely the effect of this rhetoric
when magnified via the literalising commensuration of objects and their
assessment? Thus our very image of the big man inflates him through the
imputation of his own inflation, whereas his distinctive indigenous attribution is as a rhetorician (Reay 1959: 113-30). For ultimately the final
arbiter of money as well as law and court cases, ethnography as well as
indigenous status, is talk. And talk, a concept that is generally inclusive of
language for Melanesians, is by no means the same thing as description,
assessment, information or language itself. It is the medium of their fractality, that which expands or contracts the scale of recognition and articulation to fit all exigencies, making language equal to all occasions by

166

Big men and great men

making those occasions over into talk. Hence talk is like a poai relation
intrinsic to thought. Law and money, singular and plural, individual and
group, even ethnography, are supposed to be the places where it comes to
rest, but talk about law and money, even ethnography, never rests, and talk
itself, as Goldman's recent study of Huli rhetoric exemplifies (1983), never
dies. This is the fractality of the Melanesian person: the talk formed
through the person that is the person formed through the talk.
Neither singular nor plural

When the arbitrary sectionings cut from the whole cloth of universal congruence are taken literally as data, they become the social categories that
we identify as names, individuals, groups, wealth-objects and informationbearing sentences or statements. Taken at face value this way they lose any
sense of fractality and merge with the western hegemonic of social orders
constructed of substantive elements, cultural systems made of representational categories. This does not mean that the fractal possibilities of scale
retention are not there, for they are evidenced by the poai relation and its
many equivalents. But it does carry a strong guarantee that the indigenous
awareness and use of these possibilities will be discounted, overlooked or
misread as rustic attempts at social construction.
To put it into the structuralist terms that have become an argot of the
social anthropologist's craft, the possibility remains that social and cultural
phenomena might be collapsed along a number of axes to yield scaleretaining understandings of unsuspected elegance and force, the generalising forms of concept and person that are neither singular nor plural. This
would implicate Benoit Mandelbrot's fractal dimensionality, perhaps the
general case of holography, as a `fractional dimension' or dimensional
`remainder' that replicates its figuration as part of the fa bric of the field,
through all changes of scale. Fractality, then, relates to, converts to and
reproduces the whole, something as different from a sum as it is from an
individual part. A holographic or self-scaling forro thus differs from a
`social organisation' or a cultural ideology in that it is not imposed so as to
order and organise, explain or interpret, a set of disparate elements. It is an
instantiation of the elements themselves.
The phenomenality of meaning provides an apt parallel; there is no such
thing as 'pare of a meaning. Though we may well persuade ourselves,
through grammars, sign-systems, deconstructive ploys and the like, that
the means by which we elicit meaning can be eminently partible, the meanings so elicited do not and cannot have parts. It is not simply a matter of the
clich about wholes being greater than the sums of their parts, for if a mean-

The fractal person

167

ing has no parts, there is no sum to compare with the totality. One might as
well conclude that the whole is less than the sum, for it is only one. When a
whole is subdivided in this way it is split into holographs of itself; though
neither the splitting nor its opposite amount to an `ordering' function.
What we call an `order' belongs to the world of partibility and construction.
This calls to mind a more extended Melanesian example, that of
Mimica's remarkable study of the conceptual mathematic of the Iqwaye,
an Angan-speaking people who live near Menyamya. Mimica (1988)
describes an essentially recursive counting system, which includes only two
numbers, one and two, and is computed on the digits of the hands and feet.
A crucial facet of the mathematic is that digits are understood to be assimilated to the final number reached, a holistic sense of sum or totality for
which Mimica borrows the German term Anzahl (1988: 102). Thus, for
instance, the five digits of the hand become `one', in the sense of `one hand',
because they are assimilated to the final 'orle' in the series `one-two-onetwo-one'. 'Ten', the `one' at the end of the second hand, is also, of course,
`one', except that this is hand number two. The feet are likewise differentiated (`one foot', `two feet'), except that the unity at the conclusion of the
second foot becomes, oddly enough for an even number, one: `two hands,
two feet: one man'. Then we start again with the first finger of the first hand,
counting it as `twenty', or `one man' instead of `one finger'. When we have
counted twenty of these twenties, or 400, the Anzahl is once again `one', as
is 8,000 and so forth.
In fact, infinity is also `one', not so much through some privileged access
of the numeration, but simply because it is always counted on the body,
which always closes on one. But the reason for this also closes with cosmology, and with the kind of universal congruence or integral relationship
evidenced in the poai relation and in genealogy. According to Mimica
(1981), the Iqwaye cosmos was originally embodied as a single man,
Omalyce, folded in on himself, with his fingers interdigitated between his
toes and a penis/umbilicus connecting abdomen and mouth. Only when the
ligament was cut, and Omalyce unfolded, did plurality/reproduction, as
well as the fingers and toes on which to count plurality, come into existence.
It should not be a surprise, then, to learn that numeration and genealogy
have the same congruent basis for Iqwaye, that they characteristically name
their offspring (in order) for the digits of the hand.
Now suppose that a western demographer carne to make an accurate
census of the Iqwaye. No matter what number might be reached, and no
matter how painstakingly and accurately the census is carried out, it will
invariably be deficient by Iqwaye standards. For the Iqwaye totality, the

168

Big men and great men

Anzahl instantiated by Omalyce, includes also all the Iqwaye who have
lived in the past, and all those to be born as well (Mimica 1988: 74). However high the number, it will always be less than the number embodied by
Omalyce, which is, of course, one. Each Iqwaye person, then, is a totality,
Omalyce instantiated, but any number of them is less than that. For
Iqwaye, in other words, counting/reproduction keeps its human scale,
which is by no means comparable to the a bstraction of western number.
The holography of reproduction grounds another extended example,
that of the Gimi of the Eastern Highlands described in chapter 10. Initially
Gillison delineates this holography through a kind of metonymic conflation of the contained foetus with the penis contained in copulation. Like
the penis, the foetus has an opening at the top, the unclosed fontanel,
whereas the mouth is covered by a membrane ( Gillison 1987: 177); the
foetus `grows' in the womb as the penis swells and erects in the sexual act,
and it `eats' the proffered semen through the fontanel (Gillison 1987: 178).
But the substance it eats flows from the head of the father, himself a
matured `foetus' and thus a penis, down through his urethra, so that the
`head' of the foetus eats the metonymic `head' of the father. Gimi note that
the entire male body becomes flaccid, penis-like, after intercourse.
A man is, then, a penis with a penis; but so is a woman, according to
Gillison, save that her penis is within her body, even before impregnation.
For Gimi understand that a female foetus is impregnated by its father as it
is formed, that 'the means by which the Gimi female is conceived and made
to grow inside the womb are the same as her "impregnation" . . . [s]he is
congenitally pregnant with her father's dead child' (Gillison 1987: 186).
This incestuous miscarriage is her internal penis, to be displaced by the
monthly visits of the moon's giant penis, causing a bloody discharge of the
miscarried substance, and then by that of the husband or lover, instantiating itself metonymically as another foetus.
The set of substitutions constituting a woman's internal penis, from the
holographic foetus within a foetus to menarche to that of impregnation and
pregnancy, is also the coming into being of legitimate procreation and kin
relationship out of its incestuous opposite. Its social legitimation in marriage has a familiar ring, for along with the bride and her implicit
internal penis, her father secretly bestows an 'externa!' penis. This takes the
form of a hollow bamboo tube filled with cooked meat, with an outlined
but uncut `mouth hole' that is decorated with a pattern also tattoed around
the bride's mouth before marriage (chapter 10). The groom must remove
the cooked meat and give it to his wife to eat, then excise the mouth-hole
and play the tube as a flute. A `penis' that is a female `foetus' already pregnant with substance from the bride's father, the tube has been `fed' through

The fractal person

169

its `fontanel', the hole in its end, whereas its embryonic mouth is still
covered by a membrane. And it is identified with the bride when cut, vagina
for mouth-hole, and was made by the bride's father as a replica of his own
flute, its `mother', which he plays in his own men's house.
The appropriate recompense for this externalised pregnant foetus is a
return payment made upon the birth of a child, for the child's head. This is,
returning to the beginning of the example, the metonymic `head' of the
father again, though like the bamboo tube it carried a number of equipotent
analogic strains, all divergent facets of a single motif. In Gillison's words:
Gimi kinship is created, in other words, by an arduous process of differentiating one
life-giving thing. This `thing' is either alive and moving upward as seminal fluid or
killed and flowing downward as menstrual blood, but it is always derived from and
synonymous with the Father's penis. (1987: 198)

It is important to keep in mind that the arduous process of differentiation

is as much a part of the holography like the penis that makes itself a foetus
to replace another foetus within an enlarged `foetus' as the motif itself.
This can be seen in a third example, taken from my work among the Usen
Barok of Central New Ireland (1986). Barok constitute each of their
exogamous matrimoieties in terms of the relation between them: a moiety
contains the nurturance of fatherhood proffered by the other, and begets,
penetrates and nurtures the containment of the other. It is this relation,
rather than the moieties themselves as social bodies, permutated through
the transformation of the feasting cycle, that gives legitimacy to all transfers
of status or property.
Barok orong, traditional feasting leaders, say that two things are replicated over and over in everything they do, kolume and gala. Kolume is
containment, as the womb contains a foetus or the earth a corpse, and is
concretised ritually in the stone-walled enclosure of the taun or men's
house. Gala is the elicitation of inception and nurturance, as the penis
penetrates to fertilise or the knife to distribute, and is realised ritually as a
rooted tree. But this imagery itself, an iconography that Barok call iri lotos,
`finished power', is the kolume, containment, of the whole, as feasting, the
elicitory process by which its meanings are realised, as its gala. The Barok
term for feasting is `cutting pig'.
It is the relation between kolume and gala, then, that both constitutes the
moieties and relates them. Understood in the broadest sense, kolume as a
containing iconography, gala as the elicitory protocol of feasting, however,
it is clear that each of these modes is in turn constituted by the relation
between them. For the iconography contains images of both kolume and
gala, each of which is, through the action of the other, further resolvable

170

Big men and great men

into kolume/gala. Thus the ground within the taun enclosure is cut by a
tree-trunk (the threshold log of the men's house) into feasting and burial
spaces, whereas the upright tree-trunk is cut by the ground into a subterranean (burial) and an a bove-ground, fruit-bearing (nurturant) half. And
the protocol of feasting begins with a kolume of feasters surrounding the
food, and proceeds to the gala of cutting the pigs and consumption a basic
format to be enacted in either a kolume (closed) or gala (open) variant.
The relation kolume/gala `keeps its scale', as the mathematics of fractals
would have it, regardless of the level of magnification. Kolume and gala are
fractal motifs that, very much like genders, stand between whole and part
so that each can equally encompass the total relation. The clinching demonstration comes in the transformational final mortuary feast, the Una Ya
(`base of the tree' ) Kaba. The tree-image of gala is inverted, the pigs for the
feast arrayed atop the burial section, the roots; atop the pigs, in the position
of the tap-root (the tree's `apical ancestress' ), the winawu, or neophyte
orong stands. Kolume and gala are shown to be equally effective if their
roles are reversed, and thus identical; a single image is made of the apical
ancestress's encompassment of the people from the past and the winawu's
encompassment of them in his future potential. In a sense, the winawu is a
great man, an encompassing rather than a statistical leader, who outflanks
memory from a future position.
The three examples of holography are drawn from different language
families and represent different geographical locations in Papua New
Guinea. There is considerable evidence that the phenomenon is
widespread. A notable instance is Mosko's study of the Bush Mekeo
(1985 ); chapter 5 shows that for them, as among the Barok, a single relation
replicates itself throughout a ritual format. But if holography has a significance in this discussion, it is not as an ethnographic phenomenon but rather
as a mode of understanding.
Neither part nor sum
In no case is the holography a matter of direct presentation; it is not perceived in the material so much as it is re-perceived as the sense of indigenous
intention to show phenomena in their self-constitution. Thus the Iqwaye
`make people' in counting, and likewise for them making people is a
counting-out, or instantiation or re-numbering of Omalyce. The Gimi
female embryo is already pregnant with a holograph of her father's penis,
with the transitivity of replication that, via its transformations, becomes
continuity. The gala of Barok ritual feasting elicits and nurtures the containment of its own relation to kolume, and hence of the moiety relation-

The fractal person

171

ship, which becomes the simultaneity of memory and reproduction.

Nothing is built up and nothing dissected in these examples; they are
neither construction nor deconstruction, but simply a further replication of
fractality in the ethnographer's understanding. One might say that the
indigenous holography is re-interpreting the anthropologist's ideas, and in
the process re-interpreting interpretation itself.
Reperception implies that the holography will not be apparent in the
kind of organic thought that distinguishes kin terminology as 'social' (or
`cognitive'), factionalism as 'political', horticulture as `technologicaP, or
that postulates an integration of groups, functions or categories into a
larger social fa bric. The crucial element is the fractality that prevents the
differentiation of part from whole, that keeps the imageries of understanding from collapsing into the individuals, groups and categories that constructionism bundles into wholes greater than the sums of their parts. Thus
it matters very much that we follow the indigenous modalities here, the
analogic cross-sections through which the whole grows itself. Without the
instantiation of the Anzahl, and a special sense of the body, Iqwayan counting is but a mathematical mistake; ignoring the transitivity of its impregnation and the transformations worked upon it, Gimi reproduction is just
a neat set of native categories, and missing the exacting protocols of feasting, Barok kastam is merely a Durkheimian solidarity-feast, a happening
that could take any number of other forms.
The holographic totalisation of the conceptual world evidenced in these
three examples amounts to a recognition of personal fractality through the
realisation of its relational implications. As such, it is not a `construction'
or even an `interpretation' on the plane of explanation, for it is not
mobilised as a forced uniting of disparate elements, a realisation of meaning via the unaccountable methodological magic of scale-changing.
A big man, in the standard and inflated anthropological clich, becomes
the organiser of sociological 'force' in his agglomeration of others' debts as
status, whether this status is seen as that of an integrator or simply a powerbroker. He undergoes a personal magnification when he changes from an
individual to a sociological scale. What is often termed the sociology of
small-scale societies produces its object as well as its solutions through the
means of scale-change, the successive grouping of individuals and individuation of groups. Each facet of the assumed social structure or organisation
involves such a shift from individual or household to lineage or village,
from lineage to phratry or society, region to areal integrate. And once this
principle is established as basic, as an analytical strategy, with the big man
as indigenous integrator and scale-shifter, a rationale for change of scale as
legitimate theoretical strategy is fixed in place. Special terminologies are

172

Big men and great men

pressed into service to focus attention on the form of reduction or scalechange intended behavioural, psychological, symbolic, economic or ecological. The result is that as many forms of heuristic `order' are attached to
the subject as scale-changing heuristics can be imagined: that once system
and order are assumed to be what society is doing, the anthropologist is
given carte blanche to propose alternative heuristics.
Indigenous forms of thought and action thereby cease to be their own
subjects in the process of becoming many subjects, a virtual kaleidoscope
of scale-shifts. At the core of this strategy is the hegemonic dogma of the
disparate and distinctive quality of the individual in relation to any form of
generalisation or grouping, any system, that might be applied. It underwrites and guarantees systematising as the basic task for anthropologist as
well as subject.
But the evidence presented here indicates that for some Melanesian
peoples at least the forms of social and cultural conceptualisation keep their
scale through all ritual and pragmatic permutations. For in such a fractal
or scale-retaining conceptualisation the concept itself merges with the
space of its conceiving, and there is nothing to be gained by remapping the
data onto artificial and introduced scalings. If most social and cultural
problems depend upon the western hegemonic for their very imagining,
this suggests that the exigencies of living and thinking in many Melanesian
cultures are rather different than social scientists have understood them to
be.
The task of the great man, then, would not be one of upscaling individuals to aggregate groupings but of keeping a scale that is person and
aggregate at once, solidifying a totality into happening. Social form is not
emergent but immanent. If this calls to mind Louis Dumont's powerful
evocation of holism in the Hindu caste system, with its fractality of
Brahmanic unity, it also resonates with Marriott's concept of the `dividual'
person the person, like the society, that is whole and part at once.
In the end we come down to a question of pieces that are cut differently
from the fabric of experience than we might expect them to be. Fractality
deals with wholes no matter how fine the cutting, and it is for this reason
that I have insisted on the themes of scale-change and magnification. For
the issue of great men and big men is ultimately one of magnification. The
big man as a product of ethnographic inflation is the result of statistical and
sociological magnification, an apparent gatherer and disperser of persons.
But the fractal conception of a great man begins with the premise that the
person is a totality, of which any aggregation is but a partial realisation.
The totality is, in other words, conceptual rather than statistical. The great

The fractal person

173

man, non gender-specific, is great as a particular instantiation or configuration of a conceptual totality; one can have kinds of great men as one can
have variants of a myth.
Godelier's study of the Baruya has given usa number of eloquent exemplifications of this point. But I should like to close with a final example from
the Usen Barok of New Ireland. The Barok orong, beginning as a neophyte
winawu, is a leader of feasting, articulator of the cycle through which the
holographic totality of iri fofos is made manifest for all to witness. Indeed,
the Kaba feast, in which the manifestation is realised, can only be held
`because the orong wishes it, and for no other reason'. Put more simply, the
orong 'kifis pigs' for the cutting-of-pig that defines feasting. The umri, the
traditional Barok war leader, 'kifis men' for another kind of feasting, that
of the ararum taun, a 'closed' or kolume variant of the public feasting cycle.
Ararum feasts, held in a space defined by the convergence of feasting and
burial functions, are restricted to salup, men formally defined as already
deceased by having undergone their mortuary feasts while still alive. They
are already ancestors, great men like orong and umri, and thereby variants
of a single myth or holography.