Property relations

Renewing the anthropological tradition

Edited by

C. M. Hann
Uniaersity of Kent at Canterbury

CANTNRIDGE
UNIVERSITY PRESS
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 2RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org
Information on this title: www.cambridge.orglgT 80521593892

@ Cambridge University Press 1998

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.

First oublisn"d
't \/Íífr-)
A catalogue recordfor this publication is availablefrom the British Library

Library of Congress Cataloguing in Publication data

Property relations: Renewing the anthropological tradition / edited

by C. M. Hann.
p. cm.
Includes bibliographical references and index.
rsBN 0 521 59389 I (hb)
l. Property. 2. Economic anthropology. 3. Right of property.
I. Hann, C. M., 1953- .
GN449.P76 1998
306.3'24c21 97-35811 CIP

ISBN-13 978-0-52 l-59389-2 hardback

ISBN-10 0-521-59389-l hardback

ISBN-l 3 978-0-521-59636-7 paperback

ISBN- l0 0-521-59636-X paperback

Transferred to digital printing 2007

4 Property and social relations in Melanesian
anthropology

James G. Carier

Chris Hann is correct when he says that property is not an issue that
excites a great deal of anthropological interest. Certainly úis is the case
in the anthropology of Melanesiâr ffiy concern in úis chapter. §7hile the
published work on the region deals with a range of topics, properry has
never loomed very large. The area of interest that seems to be closest is
land tenure. Frequently, however, úat takes úe notion of property in
land for granted, and concerns itself instead wiú kinship and the
practicalities of social relations (e.g. Crocombe L97l; I-awrence 1955,
1984: ch. 3).
The topic that most anthlopologists outside úe region associate with
Melanesia is probably íxchange.'lThis was, after all, úe focus of
Malinowski's .rílrgo nauts b*e" WeSiem Pacific (1922). Exchange is also
the most plausible claimant to the status of premier topic within
Melanesian anthropology itself. Perhaps spurred by Marshall Sahlins'
(1963) paper layrng out úe big-man system, a form of leadership based
on exchange, the 1970s and 1980s saw an impressive number of works
on úe topic. §ühile úis wave of interest had many sources, it was
riggered partly by a growing awareness úat úe lineage model that had
been imported from African ethnography could not explain the struc-
ture of the societies of the Highlands of Papua New Guinea, a region
that became open to anthropologists only in üe 1950s (Barnes 1962).
Increasingly, anúropologists came to conclude that exchange was
important (see A. and J. Carrier l99l: 8-22). §7hile ethnographers
initially linked exchange to group formation, as a substitute for úe
discredited concept of patrilineal descent, they soon began to see it as
being of great importance for understanding Melanesian societies more
generally. Indeed, it became so central to anúropological interpretations
of the region that one commentator in úe middle of úe 1980s could say
that 'exchange itself is úe central dynamic' of Melanesia, apparently
without feeling úe need to argue the point §rhitehead 1986: 80).
This flourishing of interest was concerned lyith-the way that exchange,l
was implicated in úe formation of social Sí-9"p.j, It úerefore had little I
 ttó -lunrcs (i. Ourncr

J direct bearing on úe notion of property. Howevcr, sincc rtrorrrrtl l ()t10,

' studies of exchange have shifted úeir orientation to a diÍItrcnt qucstion:
úe way that exchange is implicated in social identiry and relationships.
This is most prominent in úe writings of Marilyn Suaüern , whose The
Gender of the Giít (1988) is probably úe most widely cited work in the
Melanesian anúropology of this period.
§ühile the focus remained exchange, the change in orientation
brought úe topic closer to conceptions of property. Anthropologists
increasingly investigated the ways that objects*õíEiffi (in the extended
sense) are associated with different peoflé"in ditre;ént settings. After all,
when people exchange, they pass obiects back and forth and the owner-
slip of the obiects changes by úe act of exchange. In addreísín§ -fhe
ways that people identifu and relate úemselves to one another through
úe giving and taking of üings, the literature on exchange necessarily
includes, even if only obliquely, information on Melanesian notions of
property.
II My purpose in this chapter is to explore some of úe üings that úe
exchange literature can tell us about notions of property in Melanesia. I
: do not, however, intend üis chapter to be a survey of the pertinent
anúropological literature. It is too vast. Instead, I mean to draw out and
illustrate one of úe central conceptions of relationships benveen people
and things. Briefly put, that conception is that Melanesian societies are
characterized by aa. inclusiae notion of property, w{r-etein an object is
embedded in and reflects durable relationships between those people
implicated in its past. This attribute is taken to distinguish Melanesia
from the modern West, where property is taken to be-íxclusiaerlndçf
úe sole control of and associated only with úe person wtro happens to
own it at úe moment. ÊIowever, it is important to bear in mind úat úis
theme of inclusive property does not emerge spontaneously from a naive
contemplation of societies in Melanesia. Rather, it is motivated by úe
intellectual assumptions and interests that shape ethnographic per-
ception and production. Consequently, I will address those assumptions
and interests to show how úey are manifest in the ethnographies úat
they inform. Ultimately úis entails considering úe ways that people
think of property in úe §(Iest.

Gift and property

The reorientation of úe Melanesian literature on exchange drew on
many sources. However, the easiest point of entry, particularly for those
interested in property, is Marcel Mauss's The Gift, an explicit point of
reference for much of that literature. At the most general level, Mauss
 l'roper ty ulrrl social rclations in Mclanesia 87

was conccrncd with úe ways in which people understand and organize

their relationships to úings, especially as these are manifest in their
relationships with each other.
ln The Gift Mauss described three different types of societies, which
he arrayed in terms of a continuum that is developmental and even
evolutionary: archaic sqcieties, societies of the gift, societies of the
commodity. His description of societies of the gift draws on material
from Melanesia and it attracted attention from anthropologists of the
region. Gift societies are dominated by groups üat are constituted by
kin. In them, kinship defines people and úeir relations to and obliga-
tions towards each other. Just as people are defined in terms of üeir
relationships with others, so are things, wiü the result úat people and
objects are, in important ways, inseparable. This link benveen people
and úings is apparent in Mauss's discussion of the'spirit of the gift', the
Maori hau, probably úe best-known and most debated part of The Gift
(Parry 1986: 462-6; Sahlins 1974: ch. 4). In this famous passage Mauss
(1990: ll) relates úe words of Thmati Ranaipiri, a Maori, as reported
by the ethnographer Elsdon Best. To paraphrase what Ranaipiri said: if
a person gives me an important item that I úen pass on to another
person, and if that oúer person then repays me with a different item, I
need to return that different item to úe first person, because it contains
the spirit or essence of what the first person gave me. If I fail to do so,
'serious harm might come to me, even deaú'.
This passage points to a fact úat is central ro anúropological
interpretations of úe way that Melanesians think of properry. Obiects
are not considered absuactly, as neuual things; neither are úey defined
solely by úe power úat their current possessor has over úem, in the
manner of exclusive conceptions of property. Instead, úe previous
possessor has a claim on the obiect and its fruits even after it leaves his
or her possession, a claim that, if denied, can even result in death. In
other words, here objects are invested wiú a durable relationship wiú
úose who possessed them in the past.
In úe case of important items, this relationship can be contained in
úe histories associated with the obiects in question. As a result of this
histõry, tli-e-ítêm is not iust a fungible thing, an anonymous member of a
class of like obiects. Instead, it is unique. One type of important item in
Melanesia is üe valuables that are transacted in úe kula exchange in
what is now Milne Bay Province in Papua New Guinea. As Mauss
(1990: 24) says of úese valuables, úey 'are nor unimportanr rhings,
mere pieces of money. Each one has its name, â personality, a history,
and even a tale attached to it.' Tbe kula valuable is, úen, like an
heirloom, an obiect that bears the identity of those people and evenrs
88 Jumas (i. Oumer

úat were part of its past. -lhis identity can bc so strong thut thc pcrson
who possesses the object possesses its past as well. 'ltris, says Annette
§feiner, occurs with a particularly valuable type of Maori heirloom, the
taonga. These carry not iust the identity of úe people and groups in
úeir past, but úeir power as well, so that to gain one is to gain that
power and identity. To acquire anoüer's taonga'is to acquire anoúer's
rank, name, and history' (§fleiner 1992:64).
Taonga and important kula valuables are, by their definition, extra-
ordinary obiects. Flowever, the principle that applies to them applies to
lesser objects as well, albeit to a lesser degree. The ordinary kula
valuable of modest age has an ordinary history, but a history none the
less, so úat it carries úe identity of úe ordinary kula traders who
possessed it. For more mundane items, this social identity is weaker still,
but as Mauss says, when such items are given from one person to
another, they 'are never completely detached from those carrying out
the exchange. The mutual ties and alliance they establish are compara-
tively indissoluble' (l 990: 33).
As my sketch of the basic Maussian model of the society of the gift
indicates, when he constructed a model of a type of exchange, Mauss
also constructed a model of the relationship between people and
objects, which is to say a construction of a kind of property. i ffrings are
not construed solely as neutral, alienable chattels under the control of
úe person who happens to have them at the moment, in üe manner of
exclusive conceptions of property. Rather, úey are inalienable, 'part of
úe self, somehow attached, assimilated to or set apart for the self'
(Beaglehole 1931: 114). Furúermore, and because of this inalienability,
they tend to remain attached to the people in their paslr and e_specially
thope from whom one receives úem. Here, then, the obiect is con§trüeã
üã.rirrg and being a part of one's relationships with orhers, in the
"r
maÍIner of inclusive conceptions of property.
It is important not*to idealize this Maussian construction of societies
of úe gift by seeing such societies solely in terms of the logic of their
beliefs and values. To say úat an object bears the identity of those in its
past, and most notably those from whom a person receives it, is to say
that the obiect is inalienably linked with that person. However, that link
does not exist in itself. Raúer, it is maintained only because people
remember it and act in ways thqt reaffirm it. In the kind of society
Maus§-ãéscribes, úere is no in§titution úat guarantees memory and
enforces inalienability in the way that úe modern state registers and
enforces types of ownership. Consequently, the link benveen a person
and an obiect can be denied, and those who are persistent and powerful
are Jikely to be able to make such a denial stick. Those who wish to
 trro;lerly ntt(l r(x utl rclutions in Mcluncsia

rnaintain thcir links with obfects of any value need to have úe power to
tl«r so.
Equally, in such societies úe social identities and relationships that
obiects embody are not valued simply for their cultural preeminence or
intrinsic merit. Even those who are temperamentally uninterested in
such úings are obliged, to attend to them, if only because they are
important materially. In these sorts of societies, social rights and access
to people and to resources are governed by social relationships, espe-
cially kinship and gift relations. As Mauss says, in the gift 'all kinds of
institutions are given expression at one and the same time - religious,
iudicial, and moral, likewise economic' (1990: 3). C.A. Gregory (1982)
makes úe same point in a different way when he describes how, in gift
societies, control of material resources comes úrough úe conuol of
people.
The place of úis Maussian, inclusive conception of property in
anthropological understandings of Melanesia is well illusuated by
Marilyn Straúern, particularly in her discussion (1988: 148-65) of
Lisette Josephides' The hoduction of Inequality (1985).2 Josephides
analysed exchange and gender relations among the Kewa, a society in
the Highlands of Papua New Guinea. Put most simply, she argued that
women's labour was crucial for rearing the pigs that were a cenual part
of Kewa ceremonial exchange. However, she said, when these pigs were
given in exchange, it was úese women's husbands who decided how
the pigs should be given and who received the credit and prestige of the
giving. Unlike rü(/estern societies, then, where alienation occurs in úe
process of production, among the Kewa it occurs in the process of
exchange, for it is úere that women are dispossessed of úe products of
their labour.
Suaúern reiected Josephides' analysis, arguing in effect üat
Josephides was using an exclusive raúer than an inclusive conception of
property, a \üíestern raúer úan a Melanesian notion of the identity of
objects. Flence, for Straüern, Josephides' analysis was culturally
inappropriate. Strathern argued instead that Melanesians see things as
they see people, as 'the plural and composite site of the relationships
that produced úem' (1988: l3). The pig in quesrion, rhen, does nor
belong in any unitary and restrictive sense to the woman who reared it
or to the man who gave it away. Instead, it 'exists as a specific
combination of other identities' (1988: 159), which include the identi-
ties of úose involved in its rearing. Because of this complex, inclusive
identity, úe pig is inseparable from úe woman, so úat it cannot be
alienated from her when úe man gives it in exchange. This reconsidera-
tion of Kewa production and exchange is itself open to criticism.3
90 Junrcs (i. Ourncr

However, it does illustrate nicely the distinctivencss ol'thc cotrccption of

property that is common in Melanesian eúnography. Foll«rwing Mauss,
úe obiect is seen to bear the identities of úose involved in its past,
which includes those from whom one received it.
I have said úat anthropologists have commonly accepted Mauss's
notion of societies of the gift as a basic representation of Melanesian
social life, and I have shown how úis contains a model of property that
stresses its inclusive nature. §7hile the Maussian approach has become
dominant in the regional eúnography, úe model of property within it
has remained implicit. Anthropological interest has focusçd'on exchange
and its links wiú social identity and relationships. Froperty as an
analytical issue in its own right has not attracted much attiertion.

Ponarn property
I want now to illustrate the way that property is considered inclusively in
one Melanesian society. I stress that my purpose is only illusuation, to
provide empirical examples of the raúer abstract conceptual scheme
that I have laid out, and through úese examples to indicate some of the
limitations of the Maussian approach as it has been used in regional
ethnography. The society is Ponam, the people of a small island of the
same name that lies off the north-central coast of Manus Island, in
Manus Province, Papua New Guinea (for basic Ponam eúnography see
A. and J. Carrier l99l; J. and A. Carrier 1989).
At the time of fieldwork, roughly 1979-85, úere were about 500
Ponams, of whom about 300 were resident on úe island, while úe
remaining200 were migrant workers or úeir dependants, living in other
parts of Papua New Guinea.a As úis high rate of migration suggests,
Ponams were closely involved with the national economy and were
heavily reliant upon it for their survival. Islanders were fortunate
because migrants typically had desirable iobs úat paid well, so úat úey
were able to send substantial amounts of money home. Previously,
Ponams had survived by fishing and trading within the Manus regional
economy. Islanders had control over extensive areas of reef and sea, but
apart from some limited subsistence production they seem never to have
paid much attention to agriculture. Because the island is a sand cay, its
soil had been poor even before much of it was covered wiú crushed
coral concrete to make an air base during úe Second §7orld §lar.
Ponam social organization was complex enough to satisfr úe most
exacting ethnographer. The 500 villagers were divided into fourteen
named, land-owning patrilineal descent groups, called kamals (hamal =
male), themselves diüded into constituent lineages and sub-lineages.
 lrroperly nn(l xor lul rclutrorrr ur Mclrurcsi:r 9l
I;urther, each islundcr bclongcd to onc of nine matrilineal totemic
groups, or kowuns. -I-hese groups were important for regulating women's
t'ertiliry and for life-crisis rituals. Avoidance of group totemic plants and
animals was important for members' general health and well-being.
IrinallS each islander was a member of a number of cognatic stocks or
ken sis (ken = base; sl - one), descended from every ancestor who had
children. These were activated primarily during ceremonial exchangêr â
frequent occurrence, though they were also important in recruiting
assistance for any proiect larger than a household could undertake on its
own. Of these úree types of island social units, úe patrilineal_ kamals
were úe most important for the ownership and transmiss ion ofrandly
concern here.5 (*---"-',
Other than areas set aside for the village church, school, cemetery and
streets, Ponams used land in only t'wo important ways. One was as a site
to build dwellings, almost all of which were in a dense cluster úat
covered only a small part of the island. The bulk of the rest of úeir land
was used for subsistence agriculture. Land owners planted coconut
palms and gathered the resulting nuts or cultivated small patches of
pumpkins. Such land was also important as a source of firewood.
Ponams said that originally their island was settled by people from úe
Manus mainland. In most cases islanders no longer knew the names of
these settlers, where they came from in Manus or how úey were related
to living Ponams, nor did they really care. §ühat islanders did care about
were those ancestors considered to be the first owners of the different
parts of the island. These were the men who were remembered as being
responsible for the first clearing of the land and úe first building of
houses on it. fllflomen'ttid not feature much in their tale§.) Thê§e men
were the ancestors who had fundamental rights in that land, to plant
and build on it, to manage it and to exclude others from it, and who
passed úose rights on to úeir children. It is worú noting that the
notion of right, of jural authority, was extremely important to Ponams
and thef"ffiê<iúently used the English word or its Pidgin equivalent I
('rait')Àúeir debates about who could and should be where and doing [,
what.6
At various points in úe past, men who owned land built men's
houses for úemselves and úeir sons and úe patrilines descended from
them, pauilines úat were the ke kol (ke = base or origin, kol = place) of a
particular piece of land. However, even úough kamals were land-
owning groups, they did not manage úeir land centrally or redisuibute
it to members in each generation. Instead, land passed from a fathet to
his sons or, in úe absence of sons, to his nearest male agnates. His
daughters had úe incontestable right to use his land so long as they
 92 Juntcs (i. Ournicr
remained unmarried. Once úey married, cach hnd tfic ilceptcsrable
right to use her husband's land. The right to land could be clisposed of
permanently only if it were given in compensation, rypically to allies
who had suffered losses in pre-colonial fighting. Such únsfers seem ro
have been rare.
However, usufruct rights, the right to use land belonging to another,
,i- could given
be rffipôia*íiry, and frequently were. .Temporarily, meant
i' until such time as the giver or his heirs revoked
úe gift. However,
revocation was difficult, especially after a few decades had passed.
Consequently, such gifts tended to be permanent in effect, though
;. Ponams were careful to distinguish benveen the alienation of land in
compensation and the revocable gift of use rights. The gift of such rights
was heritable following the same rules as úe inheritance of the unquali-
fied right to land. Further, rights received as a revocable, gift could
úemselves be given revocably to someone else. (This situation re-
sembled what Macfarlane says in chapter 5 on user rights in medieval
Japanese and English land tenure: see p. I l9). The result, particularly in
úe populated area of the island, was úat people were commonly using
land that did not fully belong to rhem, and üe re-granring of use rights
meant úat úree or four granting groups could stand benveen those
identified as the original, fundamental owners of a piece of land and
üose who were curently using it.
consequently, many of those who occupied land did so, and saw
themselves as doing so, only because of nvo sorts of social relationships
úat were embedded in úe land itself. The first of these was the link
between úe original land owner and the person to whom he had
granted use rights. Such grants commonly occurred in two situations.
One, which effectively ceased after úe imposition of colonial control in
the early part of the twentieú century, was benveen a land owner and
refugees whom he had taken in. In this case the grant solidified a relation
of patronage with úe recipient. The second situation, though never very
common, did not disappear after colonization. Here a land owner whose
daughter had married a man with relatively little land, and especially
wiúout adequate land for a house site, would grant the right to use land
to the daughter, and úis right would pass to her childr.rr. Th. second
set of relationships embedded in the land linked living r)ónams to úese
original donors and recipients of ee granr of use rights. Most granrs
were several decades old, so úat the original parties were dead. The
present occupant could claim continued use of the land only through
descent from the ancesual recipient.
The consequence of this granting and re-granting was that most
peopl.e were not living on land that was wholly úeir own, land of
which
 l'rolrt't ly urtrl nur tul trlultottr ttt Mclrtttcsia 9J

thcy wcrc thc rcsitltrul owncrs or,lec Aol. Rather, thcy were living on land
to which use rights had bcen granted at some time in úe past. Thus, in
thc early 1980s thcrc were fifty-eight households headed by married
rncn or single women. In only sixteen cases, less úan a úird, was úe
household head patrilineally descended from the person who had owned
the land three generatio4s previously. Of course, úere is no reason to
think that all of those ancestral owners were úemselves residual owners.
Of the nineteen households headed by widows, in only seven was úe
widow's deceased husband descended patrilineally from úe ancestral
owner (J. and A. Carrier 1989: 5l).
For Ponams, úen, land was neiúer a neutral commodity nor some-
üing that was linked only with its owner. Rather, when Ponams walked
across úeir village they were walking across a map of social relations
üat linked them to those who occupied the land and úat linked the
occupiers to úose who were the original owners. As might be expected,
these social relations were more visible at some times than at oúers. In
úe normal daily round, islanders did not pay much overt attention to
the social identities and relations embedded in the place where úey
were walking. Likewise, in úe normal daily round, islanders did not say
much about úe place on which úey had planted their palms or built
their houses. However, úis relative silence is not a mark of indifference.
Instead, it marks the way that their knowledge of property and its social
aspects was taken for granted.
This knowledge came to úe fore on ceremonial occasions, when
people's social identities and relations werê-t-éiiig celébfated in one way
or anoúer. One such occasion was in the activities of intra-village
ceremonial exchange, which was invariably organized and executed by
kin groups. §úItren, as a part of such an exchange, a group of villagers
carried a collection of gifts to úe intended recipients, they did not
casually walk úrough the village. Rather, they were careful to walk only
where úey had an undeniable right to walk. There were nvo such
places. One was the main streets of üe village, úree of which ran east to
west and three of which ran north to south. These, togeúer with a few
other pieces of land, were public, effectively free to anyone at all times.
The other place where úese villagers could go with impunity was on
úose lesser paús úat crossed the land of people to whom they stood in
the appropriate social relationship. The villagers carrying a brideprice
payment from the groom's relatives to the bride's would avoid úe path
úat went across úe land of those who were much more closely related
to the bride's family than to the groom's, as üey would avoid paths úat
went across the land of those with whom they were in dispute or to
whom üey were only distantly related. Equally, they might very well
 94 .fu»uts (i. Ourncr
intentionally take a path that wcnt across land rhut thcir nnccsr.()rs
granted in usufruct to anoúer, as a public statement and remindcr
of
their relation to the land and its currenr occupants.
Because kamals, patrilineal descent groups, were the main institution
associated wiú land, attention to property in land was pronounced
in
kamal activities. Most notable was the building of a new kamal men,s
house. The building and associated festivities were crucial for estab-
lishing úe legitimate, autonomous existence of the descent group, for
establishing that úey really were a kamal and not merely a disaffected,
dependent rump of another kamal. As a common part of the festivities
that celebrated üe completion of a new men,s house, a piece of cloth
was thrown on to úe thatch of the roof. This was publi" ,."ognition
úat üe men's house was built on land that had been loaned, rather úan
belonging to the kamal itself. The cloú was collected by a represenrative
of úe heirs of úe man who had loaned the land in úe filrst place, a
collection that reaffirmed the loan. This ceremony was a common
feature of men's-house building because historical changes in the
placement of these houses meant that almost none built after
the
Second World §7ar were on the kamals own land.
Although land shows most clearly the way that ponams located
objects in a web of social relations, the same theme is visible in the
collection and distribution of §ifts io1".r.*onial exchange. The average
Ponam was engaged in such an exchange about one day out of four.
§7hile these ceremonies were important expressions oi ponam selÊ
conception and cultural values, úey had a material importance as well.
Erik schwimmer (1973: 49) pointed úis out in his study of úe
Orokaiva, of mainland New Guinea:
§Testerners often criticise Melanesians for being too grasping and mean
in gift
exchange. Absurd though úis criticism may..ã*, it arisãs frorn real
cultural
"
difference: §flesterners depend on institutions other úan gift exchange for the
acquisition of desired scarce resources. Hence úe instiútion of ei'changing
Christmas presents need serve no other end but úe fostering of social
exchange
\i relations. For úe Melanesians gift exchange must serve economic as well
as
social ends.

For Ponams, much of the wealth úat migranrs remitted to the island
was sent back as contributions for these ceremonial exchanges.
The items in an exchange were collected and distributed in a complex
pattern of accumulation and dispersal. I will illustrate úis pattern
wiú
only one example, the sort of distriburion made by the lÀader of the
group receiving the gift. I-ogically it is possible to make such a disuibu-
tion in many different ways. one could, for instance, place all of the
items. of üe gift in a heap, take each item in turn from the
heap,
 I)ro3rcrty urrrl rot'rul rcluliorts in Mclaltcsia 95

announce who is to gct it, and hand it over. Ponams, however, used a
m()rc complex system. For all but a type of distribution associated wiú
mortuary ceremonies, the distributor would lay out úe items in a series
of piles, arrayed in front of úe dwelling or men's house selected for the
site of úe distribution.T The distributor or a public speaker would üen
walk through úe display and announce úe name of úe recipient of
each pile. §ühile this may, at first glance, seem to differ little from merely
picking up one item after another and announcing the recipient, in fact
there was a significant difference. The piles of items were given to thsse
who were úe prime figures in úe disuibutor's kindred, and they were ,i
laid out in the form of a kinship diagram. Moreover, most of these prime
figures were long dead, and their share of the distribution went to úeir
living cognatic descendants (see A. and J. Carrier 1991: ch. 6). Thus,
úe method of disuibution stressed that people received items only i
because of their kin relationship to the person making the distribution, l,
who had items to distribute only because of his or her kin relationship to .'
the person making the gift in the first place.
I have described Ponam systems of land tenure and ceremonial
exchangé to illustrate the ways úat objects such as land and items
distributed in exchange are enmeshed in and embody a set of social
relationships. My intention has been only to flesh out with ethnographic
examples the basic Maussian model that I described. In my illusuations
I have, moreover, reproduced a key element of the way that the
Maussian model is used: I have stressed what might be úought of as the
jural and suuctural nature of Ponam property. The flurry of recent
resàarch on exchange in Melanesia úat utilizes úis Maussian approach
attends primarily to the basic principles or cultural understandings that
are said to underlie úe relationship bet'ween people and things. §ühile
úese principles and understandings certainly are important, the result is
rather bloodless. Simply because social identities and relationships are ,
important in shaping people's relationships with valued resources, one i{
would expect, and Ponams themselves certainly expected, that there l
would be disputes about üose social identities and relationships, that
there would be disputes about who is linked to whom and what, and in ,

which ways. The all-encompassing Maussian approach turns out to

leave little room for such dispute. ''i
Consider again Strathern's re-analysis of Josephides' description of
úe Kewa. If the pig bore the identity of the woman who reared it, then
perhaps it could not, strictly speaking, be alienated from her, or from
her husband or from their coniugal relationship. Furúer, as Strathern
notes in úis volume (p. 221), úe woman's claim is recognized (and
apparently extinguished) by the payment to the woman of a portion of
 9ó .|urr,.,s (i. Ournirr

Pig, a process analogous to the wage givcn t«r l'uctory wlrkcrs that
extinguishes úe workers' claim on what is produccd. I l.wcvcr, the
relatively frequent disputes bet',veen Kewa husbands and wives over the
disposition of the pig suggesrs that wives frequently do not see things
this way and are alienated from their pigs. Ttris indicates that attention
to basic cultural understandings about relationships among people and
things needs to be complemented by attention to patrerns óf piactical
social outcomes. It is in these outcomes that factors more mundane
than
inalienability or the spirit of the gift make their mark, and so help qualiff
what we know of úose understandings. Such outcomes have attracted
less attention úan the cultural understandings they seem to qualifr
and
complicate.s
Certainly Ponams had an articulate awareness of the distinction
between úe structure of their system and the processes of its practical
application and manipulation.e This distinction is clearest in àisputes
about land, which were very common. In fact any inheritance of land,
granting of usufruct rights in land or repossession of granted land
was
almost cerrain to lead to a dispute. Typically dispuies were at first
euphemistic. claims alleging the theft of coconuts, for example, were
covert claims about who had rights to úe land on which úe palm uee
stood. At úis early stage úe disputants simply quarrelled, using gossip
and oúer techniques to try to mobilize public supporr. Relatively few
disputes went beyond this informal stage to court proceedings, which
meant the Village Court, an entity authorized by national statuie to hear
minor disputes and apply customary law. Because Village Courts could
not hear cases about land, úe province of úe national l-and Mediators
and I-and Courts, úe disputes úat went this far still used claims of
theft
as a way of disputing ownership of land without seeming to do
so. Few
cases wenr beyond the Village Court.
Disputes were sufÊciently common for Ponams to recognize a set of
practical factors that complemented their formal rules of gift, grant
and
inheíita*nce in shaping lanà ownership. Their concern with rights, if you
will, was complemented by an articulate awareness or úoãgr.'Th.r.
factors can be codified. First, while sons have an inalienaUte iight to a
share of their father's land, úey can be disinherited by a strong troúer,
paternal uncle or agnatic first cousin. second, a land o*rr.i can
face
overwhelming social pressure to make or refrain from making a grant
of
use rights. Third, úe land of someoàe who dies without heiri *Iy go
,o
the strongest near relative rather úan úe nearest patrilineal réhtive.
Fourú, a strong group may be able to claim that it is the residual owner
of a piece of land not acrually its own. Fifth, rhe more distantly one is
linked to a person who holds land in usufruct, úe stronger one has to be
 I'roglcr ly urrrl ror rul rrlsttotrn ttr Mclattcsiu 97

to claim rightltrl irrlrcritancc of residual ownership of úe land or to

rcpossess it.
These practical rules speak of strength . raúer than formal iustice.
Ponams did not úink úat the formal rules were only for show, and that
these practical rules were the real ones. Rather, they saw boú as real.
The formal rules were what proper people followed and what úey could
be criticized harshly for violating. However, islanders recognized that
people are not always proper, that úey are guided by selÊinterest as well
as rulpsç an{that power as well as iustice affects actions and outcorqes.
The,'power that was important in these disputes was of úree kinds. The
first iiiiid of power, and the most legitimate, was knowledge - úe ability
to produce good speakers who were reputablüêtdêr§ with a detailed
knowledge of patriline genealogy and land history. Those who could not
produce úis sort of knowledge stood little chance of defeating those
who could, if only because úey would be reluctant to take cases to úe
level of formal dispute.,lfhe second source of power was úe strengú of
numbe-§-,q5r_d persistence. Ponams did not regard any dispuê settlement
or court ruling as really final, and losing parties could re-open disputes
whenevei they chose. Thus, a numerous, wealúy and persistent patri-
line could keep up a steady stream of disputes and ultimately wear down
weaker oppositio/m, third kind of power was intimidation. This
ranged from the.semi-legitimate and overt (such as <i§uâôísm, which
could be very effective when used by a large group against a small one)
to the highly illegitimate and covert (especially sorcery).

Constructing Melanesia
I have devoted attention to land disputes on Ponam because they show
the way úat anthropological understandings of property principles, and
of cultural understandings more generally, need to be complemented by
anthropological attention to property practices, and to social practices
more generally. These raise questions about the ubiquity of inclusive
notions of property, about iust what it means to say that villagers like
Ponams can be understood adequately using úe Maussian model.
Certainly islanders understood orderly and proper land succession, as
they understood orderly and proper distribution of gifts, in terms of
principles that identified the obiects with the previous owner or with the
disuibutor and wiú the kin relationships benveen üat person and úe
recipient. My point is, however, üat üey understood practical succes-
sion and distribution as reflecting other things as well - as reflecting
forces that were shaped by economic, social or political interests úat did 1'
not march easily wiú úe principles of úe Maussian model. To say that
 gti .funrcs (). Ourntr

Ponams recognized these other Íbrces is neccssrlrily t() luy tltrt thcy
understood property in more úan one way.
These other understandings had a place in older Melanesian erhno-
graphy, particularly (and perhaps ironically) úe work on exchange and
group formation that preceded úe shift to a more Maussian orientation
(e.g. Finney 1973; sahlins 1963. A. straúern l97l). However, these
understandings seem to have slipped from view in the l9g0s, in the
exhilaration of pursuing úe more systemic, structural, cultural logic of
the Maussian approach. This ãpproach has-'"b-een siven a powerful
statement recently in Of Relations and the Dead (Barraud et al. 1994). In
úeir comparative study of exchange in four societies, the authors
reiterate their concern to take a Maussian view of society, which means
focusing resolutely on the structural logic of the whole of a sociery,s
system of exchanges in order to 'bring to light úe values at work' (1994:
108). §(zhile úey are right to argue against the utility and desirability of
extreme individualist approaches (1994: 104-5), their exrreme holism
leads úem into úe sort of difficulties úat I have described in this
chapter. In particular, they fail to consider ways in which úe idealized
structure of exchange and meaning that they elicit from úe field can or
cannot be said to exist independently of the contingencies, strategies
and circumstances that induce individuals to act in one way or another,
iust the sort of factors that Ponam Islanders recognized in their discus-
sions of how the inheritance of land works in practice rather than in
úeory.
..nt Certainly the more inclusive, Maussian conceptions of property did
not §qem to apply much to Melanesians when úey were dealing with úe
urban, capitalist economy. The Ponam women living in port Moresby
úho bought from indigenous vendors at the local food markers did not
seem to see the obiects they transacted as durably linked to the seller,
any more úan islanders who cashed a cheque thought the bank teller
had a moral claim on úe money that changed hands. one could, of
course, argue that these are alien introductions úat do not reflect the
inherent logic of Ponam life, or Melanesian life more generally. But such
an argument ignores the fact that Ponams, like others in Melanesia,
routinely and conventionally transacted in ways that do not look terribly
Maussian, as rhey thought about the obiects that úey acquired in ways
that do nor look terribly Maussian (spe Gell lgg}).ro
An example from Manus free trade will illustrate not just úe existence
of such transactions, but also how differing conceptions of properry
were made manifest in concrete circumstances. Manus üllagers had
long engaged in free trade, by which I mean transactions that took place
outside of úe regulating structures of village markets and trade parrner-
 l'ro;lt'r'ly nlrrl lor tul rclnltorrn rrr Meluncsiu 99

ships. F'rcc truclcrs wcrc sirnply villagcrs with a large quantity of some-
thing that úcy wantcd to sell. They would transport üeir wares to a
place where they thought there would be people who wanted to buy.
During the early 1980s, the typical free traders who came to Ponam
were villagers from mainland Manus carrying betel nut or bundles of
dried sago, the staple starch.
In December 1980, one young man came to Ponam with 170 bundles
of sago that he had purchased from his village church group. Sago was in
demand on úe islands of úe north coast of Manus, and he loaded them
on to a hired motor canoe and went to Ponam. He said úat he planned-
to sell as much as possible there before moving on to oúer islands
further east. This young man came to Ponam simply because he knew
that people there were likely to want to buy sago, and not because he
wanted to trade wiú kin or uade partners. He landed on a public beach,
whereas a uade partner would have landed close to the house of his
Ponam counterpart. He displayed his sago bundles on úe open ground
iust above the beach in front of his canoe, each bundle wiú a price
clearly written on it. He sold to anyone who wanted to buy, and did not
appear tô give special discounts to anyone. During the afternoon and
night he spent on Ponam he sold all of the sago. To do so, however, he
lowered his prices twice, partly because Ponams complained that the
sago was too expensive, and partlg he said, because he felt sorry for
Ponams, who, as fishing people on an island wiú poor soil, had no food.
§7hat does úis event tell us about Ponam conceptions of property?
First, úe mere existence of free trade indicates that úese villagers
engaged in transactions that did not conform to úe Maussian model,
but instead resembled the sorts of impersonal buying and selling, the
transacting of fungible equivalents, úat occurred in the urban, capitalist
economy. §íhile many islanders knew this young man, they did not
transact with him as a relative in a relationship of mutual obligation and
identity. Second, this event shows how üe understanding of transactions
and objects can be a source of disagreement and conflict, with one party
claiming and úe other denying úat úe obiect transacted is inclusive
property, embedded in and expressive of a social relationship. r I The free
uader said that his initial prices were fair, but that he lowered them out
of pity for hungry Ponams. He was asserting, then, that his lower prices
meant that he was rading in light of a social bond between himself and
Ponams, in which he gave and they benefited. Ponams, however, dis-
agreed. They said úat his initial prices were too high, úat úey were not
suited to a neutral exchange of equivalents. Ponams were correct to a
degree, as his initial prices were high compared to úe common price of
sago in the market between Ponams and their mainland market
 l(X) .fu»rcs (i. Ouritr
partners, though they wcrc not high in c.rnpuris.n to the pricc
prevailing in úe cash market at Lorengau, the provincial capital,
four
hours away by motor canoe. ponams saw no need to be ou.rly generous
with this man, no kinsman of theirs, and bought little of his r"go. It was
only after he lowered his prices twice úat islanders thought rhat úey
were getting a fair price in a transaction that bound neiúer parry to
úe other.
I do not mean by this example to argue that islanders routinely
engaged in totally impersonal trade of anon5rmous objects. In fact, in
úe
pre-colonial period free trade required kin relationships, for only
they
could provide the safety that free traders needed if thãy were to travel
and trade away from home. However, free trade existed úen, as it
existed at the time of fieldwork, and it was distinguished from mosr
oüer forms of transaction precisely because it amounted to the
exchange of equivalents benveen people wiú no durable links to each
other or to the obiects that they uansacted. Here, unlike ceremonial
exchange and trade partnerships (though not unlike market trade), úe
thing that the purchaser bought had no significant social identig.$1,
leasUS relation to the seller, it was the buyer,s exclusive property, rather
úan inclusive.
The argument I have put forward in this section brings me back to a
point I made near the beginning of úis chapter, úat the Maussian
úeme of inclusive property in Melanesia does not emerge from the
naive anúropological contemplation of societies in úe region. Rather,
it
reflects úe assumptions of úe anthropologists doing the ãontemplating.
one cluster of assumptions concerns me particularly úose úãt stem
-
from an apparent desire to treat Melanesian societies as exhibiting a
common essence, one'that is defined in its opposition to a conception of
modern §Testern societies that is equally essentialist. My
what I see as anthropological occidentalism, a concept"or..*.eflects
that draws on
Edward Said's (1978) arguments about Orientalism. I have discussed
this in detail elsewhere (1995b), so here I will sketch the argument
only briefly.
A key element in Said's Orientalisn is that Oriental Studies scholars
tended to construe the Middle East as a uniform and timeless region that
was boú radically separated from and radically different from úe strest.
The Orient became something like.a mirror image of the Occident as
those scholars understood it. \üühile Said took pains ro argue that úe
resulting construction of the Middle East is incorrect, he did not attend
to úe accuracy of scholars' constructions of the §7est, the Occidentalisms
úat exist in dialectical relationship to úeir Orientalisms. In my view
something similar has happened for Melanesia, and I have looked
 l'roperty urrrl ror tul rr'luttrurx ttt Mcluncsiu l0l
particularly ut thc irrrugc «rl'thc rJíest úat ethnographers of Melanesia
invoke. These twinncd anthropological Orientalisms and Occidentalisms
are particularly apparent in úe Maussian rendering of exchange in
Melanesia, and so are pertinent to my arguments in this chapter.
It is important to remember that the Maussian society of úe gift is
not a neutral descriptign based on ethnographic evidence. Rather,
Mauss developed and presented it as part of a scheme that was intended
to lay out key differences betrveen the modern §7est and societies in
oúer times and places. Mauss was hardly the only social scientist to
undêrtake this project, which seems to have been almost ubiquitous in
úe early periods of sociology and anúropology. Likewise, úrosq
studying Melanesia are not the only modern anúropologists to make
use of such a tactic. It is also notable among, for example, Souú
Asianists, including úose who have influenced Marilyn Strathern (e.g.
Marriott and Inden L977 and especially Dumont 1970, L977; on this
aspect of Dumont's work see, e.g., Fuller 1989; Macfarlane 1992193;
Spencer 1995). But this oppositional stance produces úe same kinds of
distortions úat so exercised Said.
Thusr'ethnographers in Melanesia have been reluctant to investigate
the sorts of transactions that I have described in this chapter üat do not
fit úe Maussian model. Consequently, they find it hard to consider the
implications of those transactions for their rendering of Melanesian
understandings of the relationship among people and úings. For
example, the authors of Of Relations and the Dead say úat 'exchanges are
all activities in úe course of which something is seen to circulate'
(Barraud et al. 1994:5, emphasis omitted), yet úe two descriptions of
Melanesian societies mention no activities that seem other than strictly
traditional and alien. Likewise, when Marilyn Suathern wrote an article
about marriage exchanges for Annual Reaiew of Anthropologt, she said
her task was 'understanding úe role of exchanges in items other úan
persons when these items are part of or move in coniunction wiú
transactions (such as marriage) conceptualised as exchanges of persons'
(Suathern 1984a: 42).However, she made no mention of money, a key
item of exchange in much of the region, and so could not consider úe {
ways úat Melanesians might understand money, the ways it is trans-
i

acted and the relationships in which it is acquired.

This Orientalist construction of Melanesia takes úe form it does
partly because the region is understood in terms of its distinction from
an Occidentalist understanding of §Testern societies, one that reduces
úem to impersonal commodity transactions and a wholly exclusive view
of property. This is not the place to review úe voluminous work that
indicates that transactions in the §íest are not so uniform as this
 102 .lu»n,s O. Ounwr

essentialist rendering indicatcs (scc (larricr lt)r)5tr: 88 «)4),

thgugh it is
pertinent to note úat little of úis work has been produced
by anthro-
pologists. Suffice it to say that when anúropologists
concerned with
exchange in Melanesia created models of the region, their
models were
shaped by understandings of úe west that ;igno.. exchanges
and
productive activiries concerned wiú non-commodities, (Davis
1973b:
166). Jt1us, the basic Maussian distinction benryeen gift and
commodity
systems can be seen, without too much exaggeration, as
a surrogate for
an essentialist simplification of the distinction benveen the
modern §rest
and Melanesia. Although it makes scholarly work more elegant,
it
distorts the rendering of each area.

Conclusions
§(lhile anúropologists may seriously exaggerate the ubiquiry
of certain
sorts of transaction and certain social identities of objects in
Melanesia,
úey have produced a substantial body of descriptions and analyses that
bear clearly, if indirectly, on concepts of prop.rty. As I llustraied
with
my discussion of land and exchange in ponam society, úe Maussian
model provides a good summary of the way that *rrry Melanesians
think about objects and social relations in many situations. Likewise,
the
Occidentalist rendering of úe §7est that is part of the Maussian
model
points accurately enough to the ways that many §Testerners
transact and
think about obiects in many situations. And .rr.r, if úe model does
lead
to an oversimplified view, it has had the virtue of encouraging scholars
to investigate a set of issues and processes úat are important for
úe
understanding of property. As úese comments indicate, I
am not
arguing that the distinction between ffis and commodities
should be
abandoned, as Ariun Appadurai (19g6) suggesrs. Rather, I
think it
should be applied with greater sensitiüty.
In particular, I think that the main shortcoming of the Maussian
Melanesianists lies in their assumption that .gift, andlcommodity,
apply
to entire societies and even regions despite cÍris cr.goryt prl"i
Crggo,
1982) úat those studying the region need to .tt.rrd to uótn gifts
and
commodities. §7hen the idea that Melanesia is a gift sysrem is
iaken for
granted in the influential anthropological literature, it
tends to become
selÊsustaining. rt comes to define yhat is authenticalry
Melanesian,
what is worthy of anthropological aitention, and so indirectly defines
what is inauthentic and not worth consideration. The result is some-
thing that André Béteille (1990: 4go) identified in Indian studies
'shift from üe fieldview ro the bookview of society, culminating in -
a
úe
assignment of a privileged position to traditional structure
over conrem-
 l'rollct l y ntrtl rot tul t clut tonfi ttt Melttttcsttt 1 0,

l)()rilry rcality'. l)ut bluntly, thc l'onutn reality of free traders, trade
st()rcs, wagc labour, bank acc()unts and all the rest gets swamped by the
traditional structure of the Maussian society of the gift.
More sensible, I úink, is a disavowal of the essentialism úat has
accompanied úe Maussian model in Melanesian anthropology. Cer-
tainly most people in Melanesia are more likely to transact and úink
about obiects in gift terms more often úan are most people in úe §(Iest.
However, this observable difference in degree should gg-.ç..become úe
scholarly asserrion of a differõÊê-id t<ind and ijúahry. Mêtànesians have
commodities and exclusive property iust as surely as \üTesterners have
gifts and inclusive property, a point made recently in slightly different
terms by ]ohn Davis (1996). §fle ought, úen, to be cautious about
investigating the logic of Melanesian gift systems, just as we ought to be
cautious about investigating the logic of \üíestern commodity systems.
tVe ought to be cautious until we spend more time investigating a prior
question, one that is raised by Maurice Bloch and Jonathan Parry
(1989) in úeir discussion of different transactional orders. That ques-
tion is: when and under what circumstances do people in a given society
transacr ôbjects in a web of social relations and think about obiects in
inclusive terms, and when do they ransact obiects impersonally and
think about objects in exclusive terms? Vlhile such an approach doubt-
less has its disadvantages, it has the prime advantage of making scholars
more open to the diverse forms of property that exist in the societies
they study.