ANNUAL

REVIEWS Further
Quick links to online content
Ann. Rev. Anthropol. 1977. 6:2SS-.:tJ1
Copyright © 1977 by Annual Reviews Inc. All rights reserved

SOCIAL EXCHANGE +9595

Harumi Befu
Department of Anthropology, Stanford University, Stanford, California 94305
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org
Access provided by McMaster University on 01/30/15. For personal use only.

INTRODUCTION

We have long been aware that exchange of goods, services, etc plays a vital part in
our own social life as well as in the life of other peoples we study, and in fact, as
Belshaw states ( 1 6, p. 7), that exchange penetrates through the social fabric and may
be thought of as a network holding society together. Such awareness, however, was
not translated into systematic thinking-with a few exceptions-until very recently.
As for these exceptions, Ekeh (37, pp. 2 1-24) cites Sir James Frazer's Folklore in
the Old Testament (49), published in 1919, and Chavannes's 1884 Studies in Sociol­
ogy as representing early thinking on social exchange. While, as Ekeh points out,
one may well find in these works features of exchange theory propounded by later
theorists, their direct impact upon modern thinkers of social exchange is not very
clear; it appears that they were the Mendels of exchange theory in that although
they were on the right track, they were forgotten by more recent theorists who had
to "rediscover" the concept.
Three anthropologists stand out as having had disproportionate influence in the
development of exchange theory. The first is Mauss, who has carried on the Durk­
heimian French intellectual tradition and applied to it the phenomena of gift-giving
in his essay on gift, first published in 1925 (89, 90). In seeing gift exchange as an
obligatory act, Mauss focuses on normative rules. Pervasiveness of gift-giving in
primitive societies leads him to propose the concept of "total prestation." These
ideas provide some of the conceptual ingredients for Levi-Strauss, the second major
contributor to the anthropology of exchange, to develop a theory of cousin marriage
(78, 80). Basic to the theory is the distinction between restricted exchange, which
is only capable of connecting pairs of social groups, and generalized exchange, which
integrates indefinite numbers of groups. The third major contributor is Sahlins (110,
111), whose conceptual distinction among generalized exchange, which (not to be
confused with Levi-Strauss's concept) parallels Mauss's "total prestation," balanced
exchange, which is epitomized by monetized market exchange, and negative ex­
change, which is characterized in extreme by "something for nothing" stealth, has
been widely taken up, applied, and tested with ethnographic data.
There are several empirically oriented studies which are to a major extent orga­
nized around the exchange concept. Probably the best known among them is Mali-
255
 256 BEFU

nowski's analysis of the Trobriand kula ring (87). In more recent years, Schwimmer
( 1 16 ) has conceptualized Orokaiva data in terms of a cycle of exchange which is
mediated by objects such as land, pigs, etc, and causes social relations to move back
and forth between hostility and reconciliation. Although very different in certain
crucial details, both Strathern ( 127) and Young ( 132) see big-men's competition for
status in Mt. Hagen and Kalauna, respectively, in Melanesia as played out in the
arena of ceremonial exchange, known as moka among the Hagener and abutu in
Kalauna. It is interesting that all these works and others (92, 94) which take
exchange as an organizing concept come from Melanesia. Outside Melanesia,
Kapferer (65) presents data from a factory from Africa, giving us a detailed and
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

first-rate analysis in the framework of social exchange primarily derived from Blau
Access provided by McMaster University on 01/30/15. For personal use only.

( 1 9). Barth's early analysis of Norwegian fishing crews and Swat Pathans (7, 8),
couched in "transactional" terms, is essentially a study in social exchange, as is
made evident in a volume edited by Kapferer (67), in which Barth's transactional
approach is further developed and revised by a number of authors. Finally, Belshaw
( 16) in the mid 1960s and H. Schneider ( 1 15) some 10 years later have devoted
considerable space to summarizing theoretical works and empirical studies on social
exchange.
Turning to major contributions from social psychology, Homans and Blau sys­
tematically build their theories from experimental psychology and economies, Ho­
mans relying on psychology more than Blau, and Blau making more serious use of
economics than Homans, as seen in his use of the concepts of marginal utility and
indifference curves and his frequent reference to Boulding (23). Boulding himself has
recently offered a theory of social exchange (25). Application of economic theory
is further carried forth by Heath (56), whose central concept for building an ex­
change model is rationality. Less influential, but still systematic in constructing
models of exchange, are Burns (28), Emerson (40, 41), Foa & Foa (44), Gergen (SO),
and Meeker (9 1).
In addition to these original contributions, a series of critiques of major works
in social exchange have appeared. Since they are all in social psychology, I shall limit
myself to merely mentioning them. Sociological Inquiry published a special issue
dedicated to critiquing Homans' theory (124). This was later revised and published
as a separate volume under the editorship of Turk & Simpson ( 130). Mulkay (93)
has published a critical analysis of Homans' and Blau's theories of exchange. Fi­
nally, Ekeh (37) provides us with a penetrating and contrastive examination of the
intellectual roots, philosophical foundations, and theoretical structure of Homans'
and Levi-Strauss's theories.
While psychological literature on exchange abounds, most of it, particularly
experimental findings, are couched in such hypothetical terms that they are scarcely
useful to anthropologists in helping analyze real-world data. These will be by and
large omitted from this review. On the other hand, anthropologists can benefit from
theoretical contributions in the social psychology of exchange and should be encour­
aged to be conversant with that segment of the literature which has potential
relevance to analysis of real-life situations. Some of these materials will be reviewed
here.
 SOCIAL EXCHANGE 257

DOMAIN OF INVESTIGATION

There is temptation for anyone engrossed in construction of a theory to make it as

comprehensive as possible. When this is carried too far, however, there is the danger
of making the theory tautological. That is, when a supposed counter example is
presented, concepts of the theory are often redefined and stretched to fit the facts.
Altruistic and coercive acts are such "test cases" lying at the fringes of the domain
of social exchange acts. This distinction among "altruistic," "exchange," and "coer­
cive" parallels Etzioni's "normative," "remunerative," and "coercive" (42).
We may begin our consideration with Boulding's conceptualization of economic,
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

political, and social systems (24). He suggests that human behavior, insofar as it
Access provided by McMaster University on 01/30/15. For personal use only.

involves necessity, chance, and freedom, should be seen as constituting population

systems, exchange systems, threat systems, learning systems, and finally love sys­
tems, the last being added only tentatively because it has not been as well under­
stood. What Boulding does here is to propose to exclude altruism (Boulding's
"love") and coercion (his "threat") from the realm of social exchange, in contradis­
tinction to many exchange writers who include either or both of these elements in
their models. For example, Mauss (89, p. 7 1) regards as untenable Malinowski's
contention (87, pp. 177-80) that some gifts are "pure gifts" not requiring reciproca­
tion. For in Mauss's scheme there cannot be gifts which are "pure," i.e. which do
not call for return, since he posits three obligations surrounding gifts, namely, to
give, to receive, and to return. Meeker (91, p. 480) also considers altruism as a type
of exchange. For her, altruism is an exchange rule which assigns maximum value
to the total payoff of the receiver, which implies that in its limiting case, one would
receive nothing at all for what he gives. This is tantamount to saying that social
exchange can take place when there is no exchange taking place.
Blau acknowledges ( 19, p. 16) that there are "virtual saints" who "selflessly work
for others without any thought of reward and even without expecting gratitude,"
but excludes such altruistic acts from the domain of exchange. Such saintly behavior.
may not be as rare as Blau thinks, if Titmuss's report of the motives of hundreds
of blood donors in England is true. According to him, for most donors "there is no
formal contract, no legal bond, no situation of power, domination, constraint, or
compulsion, no sense of shame or guilt, no gratitude imperative, no need for peni­
tence, no money, and no explicit guarantee of or wish for a reward or return gift"
( 1 29, p. 89).
Heath takes a stand that is essentially Blau's with respect to freely given gifts,
excluding them from his consideration of social exchange, for "by definition these
gifts are not given with an eye either to past or to future benefits but given out of
a sense of altruism or of a wish to help those in need (56, p. 60)." In other words,
from the point of view which Heath adopts-of rational choice and maximization
as a conscious strategy-altruism does not make any sense. Gregory (54, p. 83), too,
proposes a category of acts involving no obligation, asserting that there should be
a place in the anthropological lexicon for such concepts as "pure gifts."
Finally, Simpson ( 12 1, p. 14) makes a clear and concise statement regarding this
issue: "our theorists all indicate that altruistic behavior can be psychologically
 258 BEFU

profitable; but its rewards come from inside the altruist himself and not from the
beneficiary of the altruistic behavior, or else it is not altruism at all but approval
seeking. There is no room for altruism in a theory that restricts the source of rewards
to one's interaction partner."
At issue in this controversy is the status of intent on the one hand and of cultural
norm on the other. The first question is whether or not, when people offer gifts or
services, they indeed mean it to be a pure gift without any expectation of return.
If the answer is "yes," and if the exchange model is to be based on a motivational
system, the act should be excluded from the domain of exchange. We must not be
too hasty, however. For as Parkin notes (101, p. 172), the ideal of altruism can act
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

as an effective cover for utilitarian interests. Thus we should examine the case
Access provided by McMaster University on 01/30/15. For personal use only.

carefully both at the face value of what is proclaimed to be the intent of the
give-and-take and also what is at the subterranean though not necessarily subcon­
scious level of a self-serving scheme.
If one's interest is in constructing a systemic model, ignoring native's intent or
taking native's intent only as partial data, then there is room to include free gifts
and altruistic acts as relevant for the model. For example, native norms may
proscribe desire to receive returns for gifts to certain categories of individuals, e.g.
parents from their children for providing sustenance. Yet the norm of the same
culture may demand the recipient of the gift to return the favor, e.g. for children
to give some kind of return for the sustenance received. As a social system qua
system, reciprocation does take place in such a situation, and the case should
certainly be included in the model.
Turning now to the opposite end of the continuum from altruism, while Boulding
admonishes us to leave coercion or threat out of exchange systems, many exchange
theorists have included these and other "negative resources" as constituting ele­
ments of social exchange. Homans, for example, speaks of exchanging punishment
to one another (59, p. 57). Blau, too, refers to punishment for wrongdoing, imprison­
ment of criminals, or employment office workers penalizing uncooperative clients
( 19, pp. 227-30). Sahlins ( 1 1 1, pp. 148-49), in his well-cited work on the model of
primitive exchange, defines "negative reciprocity" as "the attempt to get something
for nothing with impunity, the several forms of appropriation, transactions opened
and conducted toward net utilitarian advantage," and in this rather broad and
amorphous classification of acts, Sahlins includes "various degrees of cunning, guile,
stealth, and violence to the finesse of a well-conducted horse-raid." The "flow" of
resources here may be, as he admits, "one way."
What these examples indicate is that "negative" exchange as a category includes
a variety of heterogeneous phenomena. It includes (a) physical coercion or threat
thereof, (b) "something for nothing" stealth, and (c) withdrawal of positive re­
wards. Among them, I believe it is best to leave out of our consideration Sahlins'
"one-way flow" of "something for nothing" and genuine physical coercion (and
threat thereot). The former should be left out because it lacks the basic element of
return, and the latter because it denies volition and hence room for operation of
strategy, to be discussed below.
Concepts and constructs of exchange models are largely created for exchange of
positive rewards--desired resources only if because, as Homans says (59, p. 57), "the
 SOCIAL EXCHANGE 259

exchange of rewards takes a larger share in social behavior than the exchange of
punishments." Accordingly, none of the five propositions on human exchange on
which Homans elaborates (59, Chap. 4) has to do directly with negative exchange.
Ethnographic cases of exchange, too, by and large concern gift-giving, payment of
bride price, potlatching, giving of armbands and necklaces in the kula ring, etc,
which all illustrate exchange of desired goods. Conceptualization of negative ex­
change thus lags far behind that of positive exchange and is a task left for the future.

CONTEXT, NORM, RULES, AND STRATEGY

At this juncture I would like to introduce the concepts of sociocultural context, the
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org
Access provided by McMaster University on 01/30/15. For personal use only.

norm 0/ reciprocity, cultural rules, and strategy.

1. Sociocultural context, or the abbreviated term "cultural context," refers to the
cultural and social environment within which a model of social exchange is con­
structed. It is that part of an exchange model which is assumed as given insofar as
the model is concerned; it is the environment within which the model is embedded.
No exchange model can operate in a cultural vacuum. Specification of the cultural
context is what brings life to an exchange model. When Salisbury ( 1 13, p. 43) refers
to "microeconomic analysis of transactions within a single moral community," as
against transactions across moral communities, and when Kapferer (66, p. 12)
exhorts us to pay attention "to particular cultural understandings which underlie
maximization or any other tactic or strategy," they are pointing to the importance
of cultural context in understanding exchange behavior.
To illustrate the concept with data from India, the jajmani system has often been
analyzed as an exchange system in which higher and lower castes exchange goods
and services (88, 97). In analyzing the give-and-take between jajmans and kamins,
inequality in caste ranking as god-given, the ritual superiority of the Braman,
specific caste ranking, economic superiority of Bramans stemming from ownership
of land and other forms of wealth (in certain cases), paternalistic relationship
obtaining between particular jajmans and their kamins, the cultural ascriptions of
hereditary occupations to various castes constitute some of the major elements of
the social and cultural context, without which discussion of exchanges between
jajman and kamin is meaningless (70).
In short, the sociocultural context sets the stage on which participants can act
out their exchange behavior according to "the script," i.e. dictates of the exchange
model. Without the stage, actors cannot act; without the cultural context, exchange
analysis is vacuous.
2. Gouldner, in his seminal work on the norm o/reciprocity, posited the universal
norm of reciprocity, which according to him, "makes two interrelated, minimal
demands: 1. people should help those who have helped them, and 2. people should
not injure those who have helped them" (53, p. 1 7 1). The norm as stated by
Gouldner is an etic concept and an abstraction in approximation of culture-specific
norms exhorting participants to reciprocate. These culture-specific norms are stated
in varying degrees of explicitness. In the Philippines, the Tagalog concept of utang
na loob (56, 58, 68) and the Tausug concept of buddi (69, pp. 65-66) both imply
moral compulsion to return a favor. Similarly, the Japanese on and giri are concepts
 260 BEFU

implying normative necessity of one who has received a favor to give an appropriate
return (12, 13, 74). American culture does not have an explicit normative concept
of reciprocity which is as salient as the Tagalog utang na loob or the Japanese on
and giri. That does not at all mean that Americans lack the norm. As exemplified
in numerous experimental and field studies (e.g. 19,59,126),reciprocity is a power­
ful norm which plays a critical role in structuring the American society.
3. Rules ofexchange refer to specific cultural rules governing what should or may
be given or returned in a given type of situation defined in terms of the specific
relationship between the participants, occasion for the exchange, etc. In contrast to
the norm of reciprocity, which is a generalized expectation to reciprocate irrespec­
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

tive of what or how much is given or returned, rules of exchange specify what the
Access provided by McMaster University on 01/30/15. For personal use only.

norm leaves undefined.

In Hagen society, for example, the main items in moka exchange are specified
as "pigs, six kinds of shell valuables, bird of paradise plumes and other feathers used
.
for decorations, sale, decorating oil, stone axe-blades and hafted stone axes, mar­
supial furs, and red ochre paint" (127, p. 101). In many East African societies bride
price is paid in cattle. These are cultural rules governing exchange behavior. Cul­
tural rules need not define the precise kind and exact amount of goods and services
to be given. Instead, generally they prescribe a range of acceptable kinds and
quantities of resources. As a return for a dinner party, an American couple has a
range of recipes to choose from. A Hagen chief giving a moka can vary the quanti­
ties of items to be given away. However, there is always a culturally prescribed limit
to the range. An American couple cannot normally invite to a hamburger lunch a
friend who invited them to a formal dinner and consider the exchange even. Hogbin
tells us that a headman in Ontong Java became an object of ridicule by villagers
because he was too ostentatious in giving a betrothal for his daughter (57, p. 37).
4. As rules of exchange define what the norm of reciprocity leaves undefined, so
does the strategy ofexchange specify the indeterminate part of cultural rules. Since
cultural rules specify a range of possibilities as to what may be given and how it may
be given, an individual must make decisions as to the exact content of the resource
to be given, its quantity, and the manner of prestation. These decisions are often,
if not always, made on the basis of the individual's desire to maximize the opportu­
nity for his benefit/profit.
An individual normally exercises his strategy for a maximum return within the
cultural context (which is to say, within the normative and institutional framework
of the culture) and within the bounds set by cultural rules (31). It is important to
distinguish between cultural rules and strategy. One refers to prescribed conven­
tions, which for participants of a culture are by and large given; the other has to
do with the operation of an individual's motivations by which cultural rules are
exploited.
Anthropologists interested in exchange are generally concerned with constructing
a model which manipulates cultural rules of exchange, such as rules of marriage,
whereas social psychologists are interested in discovering strategies individuals use
in exchange, either setting aside the question of cultural rules or by choosing a
problem where explicit cultural rules do not exist and people behave at the dictates
 SOCIAL EXCHANGE 261

of the norm of reciprocity and within the framework of the cultural context. But
such observation also applies to anthropologists interested in uncovering exchange
strategies, as Parkins notes:

The central tenet of transactional analysis, however, appears to be that the maximization
of self-interest is an axiomatic and universal feature of the human condition that underlies
the exchange nature of social relations. Transactional analysis does not attempt to identify
and analyze those social circumstances in which this supposed axiom of human behavior
is raised to a level of cultural consciousness and identification (WI, p. 173).

TYPOLOGY OF EXCHANGE RELATIONS

Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org
Access provided by McMaster University on 01/30/15. For personal use only.

1. Perceived dyadic exchange. Normally in social exchange, one assumes a rela­

tionship to take place between two existing beings (individuals, groups, etc). How­
ever, if one were to analyze exchange as taking place strictly on the basis of
"Person's" perception as to who "Other" is (to use Homans' jargon), Other need
not physically exist as long as Person believes Other exists and behaves accordingly.
Indeed, this is the very reason why in many psychological experiments on social
exchange, subjects do not interact with any individuals, but are made to believe that
they do through devious designs using confederates or simply written messages
(presumably coming from subjects' partners, but in fact written before the experi­
ment begins and simply handed over to subjects in outright deceit).
This consideration is relevant to ethnographic accounts of exchange in analyzing
interaction with the supernatural world. For example, the Buddhist conception of
merit, in which one accumulates merit through contribution to Buddhist institu­
tions, giving of alms, etc, as a way of improving one's karma, implies an exchange
relationship with the supernatural world ( 1 25). Again, Foster (46, 47) includes
dealings with saints in his discussion of "dyadic contract." Although Foster does
not formulate the concept in exchange framework, nevertheless the give-and-take
of resources between the mortal and a saint and between mortals-in the "col­
league" type contract, for that matter-can readily be reinterpreted in the light of
exchange theory. In dealing with supernaturals, obviously, a saint's giving of re­
sources such as blessings, curing of disease, bringing good fortune, etc are perceived
acts: they are merely perceived by the mortal as a saint's doing. However, insofar
as supernaturals are perceived to be real and certain phenomena are perceived as
emanating from them as a result of the mortal's propitiation, promise, etc, such
interactions are no different in substance from a mortal's interaction with another
mortal, and the same theory of exchange should be applicable to both. Lebra's study
(75) shows how reciprocity-based supernatural sanctions can effectively change
human behavior.
2. Motivational vs institutional. Earlier we have referred to those interested pri­
marily in elucidating strategy of exchange based on motivational process, compared
with those interested in cultural rules constituting institutional process. I have
elaborated on this distinction as "individualistic" vs "collectivistic" (37). Taking up
Homans and Levi-Strauss as representatives of these respective approaches, Ekeh
 262 BEFU

derives their approaches from the differing intellectual climates of Anglo-America

and France. To him, the traditions of the individualistic and collectivistic ap­
proaches are traceable back to Calvinism and Catholicism respectively, Calvinism
being characterized by atomistic world view, conception of the society as being
posterior to the individual, realism, sobriety, and rationalism, and Catholicism by
an organic world view, conception of the society as being prior to the individual,
symbolism, emotionalism, and mysticism (37, pp. 14-19). It is these differences
which are manifested in Homans' theory of social exchange on the one hand and
Levi-Strauss's on the other.
For Homans, and most social psychologists, social exchange is predicated upon
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

human sentiment, and his theory is a theory of motivation. The starting point as
Access provided by McMaster University on 01/30/15. For personal use only.

such is obviously the individual. He is concerned with strategies of exchange for

maximizing rewards and minimizing costs. Thus it is that Abrahamsson (2) has,
characterized Homans' utilitarian theory as "hedonistic," true to the Benthamian
tradition. For Homans, norms and institutions arise out of the individual behavior,
resulting from patterned relationships based on rewards and costs, and are main­
tained and supported only insofar as they are able to satisfy individuals. As such,
Ekeh argues (37, p. 181),Homans' notion of group is derivable from and reducible
to individual psychology.
Levi-Strauss's theory, on the other hand, ignores human sentiment and motiva­
tional processes. It is satisfied that motivational processes exist and dictate human
action in the form that complies with structural rules. The starting point for Levi­
Strauss is the group, and his concern is not the strategy of action but'arrangements
and interrelations of cultural rules of exchange and consequences of application of
these rules. In all these respects, Levi-Strauss is starting from a set of premises which
at the very least have nothing to do with Homans, according to Homans &
Schneider (61),or, as Ekeh would have it, are antithetical to Homans, not allowing
any possible meeting ground. Levi-Strauss's institutional approach has been fol­
lowed up in a series of writings by Rosman & Rubel (105-108) and by Schwimmer
(116).
Is the relation of group to individual to be conceptualized only as antithetical and
contradictory? Or is there a possibility of bridging the gap? Several writers, all
starting from the individualistic approaches, have proposed ways of dealing with
this question. Blau's solution, espoused by Kapferer (65,pp. 6-7), is to posit the
concept of "emergent" (19,p. 3): "Emergent properties are essentially relationships
between elements in a structure. The relationships are not contained in the elements,
though they could not exist without them, and they define the structure." The
relationship which develops out of social exchange is emergent, as is the group
defined by such relationships. But a crucial difference from Homans' approach is
in Blau's contention (19, p. 4) that "the emergent properties of social exchange
consequent to this interdependence cannot be accounted for by the psychological
processes that motivate the behavior of the partners." Yet, as Ekeh points out (37,
p. 185),this concept of emergent group should not be confused with Levi-Strauss's
concept of structure (79), which, derived from Durkheimian sociology, has an
existence sui generis. As Kapferer notes (66, p. 15), intellectual derivation of the
 SOCIAL EXCHANGE 263

"emergent" concept is rather to be found in George H. Mead's social psychology,

especially in his concept of "the definition of situation."
Paine also tries to cope with this "individual vs group" dichotomy by taking a
hint from Barth's notion of "incorporation" and developing the concept of "I mode
of exchange" in contradistinction to the " T (for transaction) mode of exchange."
According to Paine, in the I mode of exchange, "value optimum" is sought for the
sum of the partners, whose solidarity may be based on a contract or simply on
common prior commitment to a certain value (100, p. 69). This concept closely
parallels Meeker's definition of "group gain," namely "an exchange rule that assigns
the maximum value to the sum of P's [Person's] and O's [Other's] total payoffs (91,
p. 480)."
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org
Access provided by McMaster University on 01/30/15. For personal use only.

Somewhat different from the individual-group dichotomy, but still focusing on

the same issue of how to conceptualize group in exchange terms, is Heath's distinc­
tion (55, p. 30), earlier developed by Kuhn (71) and also by Olson (96), of private
and public good: ''A p rivate good is one whose enjoyment can be restricted to those
who have puid for it, whereas a public good is one which, if it is provided at all, must
be made available to all potential beneficiaries whether they have contributed or not"
(italics original). The conventional consumer good is a private good. The classic
example of a public good is national defense, which tax evaders and tax-exempted
citizens as well as tax payers equally benefit from. However, in order for the theory
of public good to work, according to Olson, there must be in addition "selective
incentives" which differentially favor those who contribute more and/or differen­
tially punish those who contribute less or do not contribute. In this system, those
who contribute to the group cause most stand to gain most from the group benefit,
and those who contribute the least benefit the least, giving rise to a hierarchy of
statuses (i.e. differential reward system) in the group, based on the degree of contri­
bution to the welfare of the group. This is essentially tantamount to exchange of
prestige and other intangible "status" rewards with material resources, as Leach
(73) has argued with respect to marital exchange and Cancian (30) with the
Mesoamerican cargo system. If we are to use this line of thinking, blood donation
in England ( 129) may be regarded as a public good. When Gregory defines "the
expectation of circumstantially balanced reciprocity (ECBR) as "a belief that those
having much (or having more) should share with those having none (or having less)"
(54, p. 74), he is conceptualizing an exchange relationship between the individual
and the community, since expectation of balance is not directed toward any par­
ticular individual, but to the pooling of the sum total of community resources.
In short, balance is systemic at the group (community) level, not between any two
individuals.
Another solution to this problem is offered by Ekeh, who first distinguishes
between "mutual reciprocity" and "univocal reciprocity" (37, pp. 205-6), the for­
mer being dyadic exchange between two individuals and the latter involving at least
three individuals in indirect exchange (to be discussed below). Ekeh's "univocal
reciprocity" is somewhat like Levi-Strauss's "generalized exchange" (80), but is
crucially different in one respect. Namely, in generalized exchange there is an
ordered transfer of resources from A to B, B to C, etc, and back to A. In Ekeh's
 264 BEFU

univocal reciprocity, there is no such assurance. One can only hope someone would
do you a favor of the kind you have done someone else. "If I see burglars in my
neighbor's house, I have the duty of doing something about it . . . because I expect
any neighbor of mine to do the same thing if he sees burglars in my own house. Thus
the concept of univocal reciprocity leads directly to the conception of generalized
duties and rights" (37, p. 206, italics original). According to Ekeh, it is only this
concept of univocal reciprocity which can lead exchange theory to such high-order
conceptions as citizenship. Thus Ekeh derives commitment to a group (nation state,
in this case) through the concept of generalized duties; but he does not tell us how
the concept of generalized duties is obtained through exchange, or whether it is part
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

of the sociocultural context and must be accepted as a given.

Access provided by McMaster University on 01/30/15. For personal use only.

3. Generalized vs balanced exchange. It was Sahlins (111) who introduced the

tripartite division of exchange phenomena: generalized, balanced, and negative
reciprocity, these being ends and a midpoint of a continuum. "Generalized reciproc­
ity" here is essentially the same as Mauss's "total prestation" (89, p. 3) and should
not be confused with Levi-Strauss's concept of generalized exchange, which, as will
be discussed below, is a different idea altogether. Concerning generalized reciproc­
ity, Sahlins says:

"the expectation of a direct material return is unseemly. At best it is implicit. The material
side of the transaction is repressed by the social . . . the counter is not stipulated by time,
quantity, or quality, the expectation of reciprocity is indefinite . . . Receiving goods lays
on a diffuse obligation to reciprocate when necessary to the donor and/or possibly for the
reCipient. The requital thus may be very soon, but then again it may be never" (111, p.
147).

Some of these descriptive features verge on pure altruism, as w�en Sahlins refers
to the suckling of children as an extreme case, or when he admits "a sustained
one-way flow of goods for a long period," suggesting that balancing of give-and-take
is of little conC(!rn. I submit that in the long run, there is probably more balancing
than meets the eye. For one thing, as I indicated above, even though it may be quite
unseemly for the giver to expect a return, the receiver may be fully expected to
return. Thus the imbalance from the perspective of the giver may well be corrected
to a large measure from the receiver's perspective. A counter case given by Pryor
& Graburn ( 1 04) will be discussed below.
In contrast to generalized reciprocity, balanced reciprocity is characterized by
precise balance: "the reciprocation is the customary equivalent of the thing received
and is without delay . . . . Balanced reciprocity may be more loosely applied t o
transactions which stipulate returns of commensurate worth or utility within a finite
and narrow period" ( 1 1 1, p. 148). This distinction is to some extent echoed in
Lebra's conception of exchange (76, p. 550). She uses the term "exchange" in a
generic sense, and under it distinguishes between "market exchange" and "reciproc­
ity," where the latter is characterized by "actor-loaded" or "socially charged"
exchange. If we regard Lebra's distinction as being ends of a continuum, her pair
of concepts parallel Sahlins' generalized-balanced pair. Sahlins' distinction between
"generalized" and "balanced" also parallels Parkin's similar distinction between
 SOCIAL EXCHANGE 265

"altruistic" and "negotiable" ideologies of transaction. At one point Parkin (101,

p. 172) equates his "negotiable" with Sahlins' "negative." But Parkin's conception
of "negotiable" as everything having its price and having to be paid for (p. 187) is
every bit like Sahlins' "balanced," and scarcely like the "something for nothing"
negative exchange which Sahlins speaks of.
Sahlins offers the generalized and balanced reciprocity as an end and a midpoint
of a continuum, implying, in his evolutionary, global-comparative thinking, differ­
ent cultures to fall in different places along the continuum. But within a given
space-time coordinate called a culture, different kinds of exchange patterns may be
observable at the same time. It is not clear in such a situation whether exchange
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

transactions are supposed to vary infinitely along the spectrum or whether they
Access provided by McMaster University on 01/30/15. For personal use only.

cluster and come in discrete types, as I think they do, representing separate spheres
of exchange (to be discussed below).
Sahlins identifies the three variables of kinship distance, sociability, and
generosity as determining the mode of exchange along the continuum of "general­
ized" through "balanced" to "negative" exchange. Lebra (76) criticizes Sahlins for
treating these three variables as correlating in one-to-one fashion. She points out (76,
p. 552) that "these variables are conceived as of unidimensional quantity progressing
from the plus extreme to the minus extreme, whereas special attention should have
been paid to the qualitatively different values involved in the variables, especially
in sociability." Thirdly, Lebra feels that Sahlins' model suffers from not considering
the interaction of two parties, and contingency of the actor's behavior upon the
alter's, as Sahlins concentrates his attention upon the actor's behavior alone.
Many workers have used Sahlins' model as a framework of analysis for their own
data (e.g. 12, 14, 34, 62, 64, 98) or to test it with their own field data. I shall review
a few of the empirical tests. Among them, Brady's and Dama's tests only partially
validate Sahlin's hypothesis. Subjecting the hypothesis to test with data from Ellice
Islands, Brady concludes (26, p. 314) that "moderate support has been given for
Sahlins' provisional hypothesis that the character of exchange events is likely to
change in proportion to social and sectoral distance." However, Brady suggests (26,
p. 314) that this hypothesis is likely to apply differentially "to individual as opposed
to group exchange events, and that the most positive relationships tend to occur
between generalized reciprocity and close-in social distance." His data tends to
invalidate (26, p. 314) "the possibility of a strict linear transformation from general­
ized to negative reciprocity as social distance increases." Here Brady echoes Lebra's
reminder that qualitative transformation along the spectrum should have been
considered. Brady also found (26, p. 315) that "the generalized sphere of exchange
for food is sometimes wider than the generalized sphere of exchange for other things,
particularly in hospitality outputs," suggesting that perishables and durables should
be considered to behave differently with respect to social distance.
Damas (33) finds in his Eskimo data that foodstuff is used for trade, i.e. balanced
exchange, a good deal more than Sahlins' model would suggest. Secondly, among
the Eskimo, kinship and community are not coterminous, and community consider­
ations override kinship considerations in distribution of resources, contrary to Sah­
lins' hypothesis. Thirdly, Sahlins' model would have us believe that hunters and
 266 BEFU

gatherers (such as Eskimos) should be primarily characterized by generalized reci­

procity; but Damas finds a good deal of balanced and negative reciprocity as well
as redistribution (which is supposed to be characteristic of a more complex, chief­
dom level society) among them. In spite of these criticisms and suggestions for
modifications, Sahlins' model at present stands as one of the most widely debated,
along with Levi-Strauss's.
The "generalized-balanced-negative" typology reminds us of another in the game
theory: they are positive-sum, zero-sum, and negative-sum games. In a positive-sum
game, as applied to exchange, both parties to an exchange relationship stand to gain
through the give and take, for both sides give up certain resources which they need
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

less in exchange for some other resources which they need more (and which the
Access provided by McMaster University on 01/30/15. For personal use only.

other party has and is willing to give). The perceived value of what one has obtained
thus is greater than the perceived value of what he has given up or what he had
before the exchange transaction. In this sense, exchange as a positive-sum game
enables both parties to profit from the relationship. In a zero-sum game, an example
of which is"Foster's "image of limited good" (48), ego's gain is alter's loss in precise
equivalence, a point of view which Parkin assumes with respect to exchange theory
and leads him to assert that men in exchange are necessarily in competition for
valued objects (101, p. 165). Lastly, the negative-sum game implies that loss or
injury is sustained by each party through exchange, and is of course intimately tied
in with the so-called "negative exchange." I have already noted the inadequate
understanding or integration of negative exchange in the model for social exchange.
The game-theoretical typology above is not strictly a parallel ofSahlins' typology.
Though it has seldom been used in exchange framework, it may prove to be a useful
frame of reference in organizing exchange data and beckons further research in this
area.
4. Direct vs indirect exchange. This distinction originates from Levi-Strauss's
"restricted-generalized" dichotomy (80). However, because Sahlins' usage of "gen­
eralized reciprocity" has been widely adopted, to avoid confusion we shall use the
terms "direct" and "indirect." "Direct exchange" refers to exchange of resources
between two parties whereby return of resources given by A to B is made directly
back to A from B. In indirect exchange, at least three units are involved in exchange,
such that A gives resources to B, who in turn gives resources to C, etc, until the
last unit gives to A, thus completing the cycles. Lebra's "circular transference"
corresponds to this type (76, p. 559).
As is well known, for Levi-Strauss this distinction was applied to forms of cross­
cousin marriage, such that bilateral cross-cousin marriage was an example of direct
(or restricted) exchange, and unilateral (either patrilateral or matrilateral) cross­
cousin marriage was a case of indirect (or generalized) exchange.
Ekeh considers Levi-Strauss's concepts of direct (restricted) and indirect (general­
ized) exchange not refined enough and has offered further elaborations (37, p. 209).
He suggests two forms of direct (restricted) exchange: exclusive, in which "the
dyadic relationships are totally isolated"; and inclusive, in which "the dyadic rela­
tionships are implicated in a network with other dyadic exchange relationships." In
my view, "exclusive, restricted exchange" is a vacuous category because except for
 SOCIAL EXCHANGE 267

dyads in imaginary worlds, such as Adam and Eve or Robinson Crusoe and Friday,
no dyad exists totally isolated.
Ekeh also proposes a subdivision of "indirect (generalized) exchange" into "chain
generalized exchange," which is the same as what we have defined as indirect
exchange above, and "net generalized exchange," in which each unit of a group acts
in relation to the rest of the group. There are two subtypes here. "Group-focused
generalized exchange," which may be represented as: (A � BCD; B � ACD;
C � ABD; D � ABC), meaning each unit contributes to the welfare of the group,
but does this separately. In the other subtype, "individual-focused generalized ex­
change," represented as (ABC � D; ABD � C; ACD � B; BCD � A), the group
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

helps each of its member units separately. Individual members of a group taking
Access provided by McMaster University on 01/30/15. For personal use only.

turns to stand in sentry may be an example of the group-focused generalized ex­

change. Villagers going from household to household helping in harvesting, as
Embree reports of kattari (38, p. 135) in Japan, or a hunting band sharing the kill
of a member [Denton (35), pp. 48-50] are instances of individual-focused general­
ized exchange.
While an indirect exchange such as cross-cousin marriage may appear to admit
of one-way flow of women only, it should be remembered that in most such mar­
riages, some form of bride compensation is given to the wife-giving group. This fact
yields two consequences. One is that there are two cycles moving in opposite
directions simultaneously, women moving in one direction and bride compensation
in the other. The Trobriand kula ring, as reported by Malinowski (87), where arm
bands circulate in one direction and necklaces in the opposite, is another example
of double-cycled indirect exchange. A second consequence of this system is that each
dyad in the system is engaged in direct exchange and that the whole cycle consists
of a series of pairs connected like links in a chain.
5. Horizontal vs vertical exchange. Equal or egalitarian exchange is one in which
values of resources exchanged are roughly the same, so that there is no enduring
or clear-cut debt that one side owes the other. In egalitarian exchange, as between
friends, there may be temporary "debt" but this is repaid in due course, restoring
the balance. The balance restored or maintained need not be exact; in fact, in some
respects it may be better not to strike the precise balance in that a small debt would
be a symbolic reminder to the continuing relationship between the actor and alter
(53). Such imbalance, however, is small enough and can be repaid easily with the
resources one controls. Such is not the case with vertical or hierarchical exchange.
Here resources controlled by the two parties are so vastly different in value that it
would not be possible for the subordinate to pay back what the superior has given.
What the superior controls may be employment opportunities, connection with
government offices, money, and other material resources, of which the subordinate
is in need, but to which he has no direct access.
If we combine the generalizedlbalanced dichotomy with the horizontal/vertical
dichotomy, we have the following two-by-two table, with examples filling the appro­
priate cells (Figure 1). Close friends are horizontal in relationship and generalized
in exchange, as defined by Sahlins, in comparison with a relationship which obtains
between salesman and customer, where there is no super- or subordination any m ore
 268 BEFU

Horizontal Verti cal

Melanesian big-man
General ized Friend-friend Polynesian Ch ieftainship
Patron-client

Balanced Salesman-customer Employer-employee

Figure 1 Types of exchange relations.

Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org
Access provided by McMaster University on 01/30/15. For personal use only.

than in a friendship, but the exchange relationship is strictly balanced. These rela­
tions again differ from vertical ones. The vertical-generalized category includes a
kinship-based rank system of the kind Sahlins discusses (109; 111, pp. 158-64),
examples of which are the Melanesian big-man and the Polynesian chieftainship. In
the latter, "where kinship-rank reciprocity is laid down by office and political
groupings, and becomes sui generis by yirtue of customary duty" (111, p. 164),
exchange takes on the character of chiefly redistribution, in the tripartite reciproci­
ty-redistribution-market classification of the substantivist economic anthropology
(102). Note that in this scheme of Sahlins', redistribution is no less an exchange
relationship than reciprocity through gift exchange among kinsmen, even though
the nomenclature juxtaposes "redistribution" with "reciprocity" as if they are anti­
theses. "Redistribution" in this scheme is a species of exchange imbedded in a
hierarchical structure implying a resource differential. Incidentally, Sahlins' hypoth­
esis concerning exchange and hierarchy receives confirmation from Brady (26, p.
314). In the Ellice Islands;'he finds that with respect to the relationship between rank
and reciprocity, "Greater existing rank, or aspirations for rank and prestige, lead
some individuals to give away items through the morality of noblesse oblige when
the initiator of a transaction intends to negotiate a 'trade' or obtain customary
equivalence in a commercial exchange. The donor thereby transforms a potentially
balanced exchange to a more generalized form."
Another generalized-vertical type of relationship, distinguished from big-man and
chieftainship and occurring in the absence of kinship ties, is the patron�clientship
and brokerage (131). Brokerage differs from patronage analytically in that a broker
is able to obtain from another source (through personal contact, etc) whatever
resources are desired by his client, whereas a patron possesses himself resources
desired by his clients. Empirically, of course, one is likely to find the two roles
combined in one person in varying degrees. The patron-client and brokerage rela­
tionships have been widely reported in ethnographic literature from Latin America
(45, 83, 128), Asia (11, 18, 27, 103, 117-119), Europe (21, 29, 120), and North
America (32, 99, 100).
Reinterpreting the traditional patron-client data in the exchange framework,
however, is a new endeavor, and attempts are few and far between (81, pp. 5-9).
Southeast Asian cases indicate changing relationship resulting from economic and
political changes. Institution of electoral democracy in recent years there, for exam­
ple, has necessitated erstwhile patrons, if they are to maintain a power base in the
 SOCIAL EXCHANGE 269

new political structure, to become patrons for a great many people at a grassroot
level in an effort to secure a large electoral base. Befu (14) argues that bureaucratic
structure is compatible with patron-client relationship, which may be a reason for
the widespread persistence of patron-client relationships in developing countries.
When a successful patron with much personal fund of resources and a loyal follow­
ing becomes a bureaucrat, an office holder, he has two alternatives. On one hand,
he can largely ignore the organizational goal of the bureaucracy and take advantage
of the power and privilege ascribed to the office for personal gain at the expense of
undermining the bureaucracy's goal. It is possible, on the other hand, for a patron­
office holder to use his personal resources for the furtherance of bureaucratic goals.
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org
Access provided by McMaster University on 01/30/15. For personal use only.

BALANCING

The question of balance is inextricably tied in with the concept of social exchange.
Balancing ne cessarily implies com pari son of at least two elements. This comparison
may be relative or absolute. In the former, A is said to be "greater" or "less" than
B; in the latter, comparison is based on the difference between two numerical values
assigned to A and B. Homans has opted for the former method (59, p. 43; 60, p.
221), Deutch's criticism (36, pp. 160-61) to the contrary notwithstanding. It is
understandable that Heath, who adopts economic theory as his model for social
exchange, would assume quantitative measurability, as when he asserts (56, p. 26)
that "some numerical probability can be assigned to the likelihood that a particular
outcome will follow a particular course of action" (italics added). Offhand it may
seem plausible to assume, as Blau does (19, p. 171), that "the worth (utility) of the
subordination he [a consultant] receives in terms of the worth of the cost in time
he incurs, decreases at an accelerated rate with increasing consultations" and to plot
this function in indifference curves (19, p. 173). This assertion assumes, however,
that contents and quality of advice are uniform and that the nature of subordination
does not vary from one case to another-an assumption which is not likely to receive
empirical validation.
Whether one adopts quantitative or relative valuation, there seems to be at least
two different approaches in assessing balance. One is to compare the cost one incurs
with the reward he receives in a given exchange relationship (which may involve
a number of transactions). "Cost" here may be defined as the resource one gives to
the other party, or it may be defined, as Homans does (59, pp. 57-61), as alternative
courses of action which one gives up in favor of the exchange one in fact has
transacted. If one perceives the cost to be equal to the reward, then one may speak
of a balance in this exchange relationship. A second way of conceiving balance is
to compare one's own profit (reward minus cost) with the other party's profit. If the
profit values are about the same, then a balance may be said to have been struck.
It should be noted that valuation in this context refers to subjectively perceived
valuation and comparison, and that the same resource may be differentially valuated
by the giver and the receiver. In fact, in many cases, Person gives up what he has
because for him the resource for him is dispensable (i.e. in some sense unneeded),
and gives it to Other because he believes Other will find it useful. Thus the same
resource is more highly valued by the one who receives it than by the one who gives
 270 BEFU

it. It is for this reason that social exchange is often regarded as a positive-sum game
in which both sides end up having more value than before the transaction.
In anthropological literature, this notion of perceived valuation is not often used.
Instead, objects to be exchanged are assumed to have some agreed upon, objective
value which is the same for all concerned. There is no question that society defines
the approximate "value" of exchange resources, whether it be a smile, a kiss, an
overnight stay at a friend.'s house. However, it is worth repeating that this societal
definition is only approximate, with a lattitude within which variation on the basis
of subjective perception is legitimate.
Central to the process of balancing is the concept of distributive justice: justice
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

in the distribution of rewards and costs (3, 59). In Homans' words (59, p. 75), "a
Access provided by McMaster University on 01/30/15. For personal use only.

man in an exchange relation with another will expect that the rewards of each man
be proportional to his costs-the greater the rewards, the greater the costs-and that
the net rewards, or profits, of each man be proportional to his investments." When
distributive justice is realized, one may say that a balance is struck in the relation­
ship. When not, according to Homans (59, p. 75), a man reacts with anger. Thus,
to Homans, the question of whether a person is engaged in fair exchange with
another is in the final analysis based on emotional reaction to his perception of cost
accounting, and has nothing fundamentally to do with moral justice.
Anthropologists are not likely to be satisfied with Homans' psychological reduc­
tionism. Notion of normative standards-publicly agreed upon values of resources,
approximate though they may be-weighs heavily upon ethnographic description.
How much a kin group would pay another for bride compensation, how much
assistance a revenge group would receive from friends, how willing a patron would
be to help one of his clients, etc, are not basically an individual psychological matter,
but a question about which there is considerable societal consensus. It is this
consensus which I called "cultural rules." What we have not seen much and are
badly in need of is an attempt at an ethnographic description which explicitly defines
cultural rules and at the same time examines individual strategies in an exchange
model.
Pryor & Graburn ( 104) raise a serious question regarding the nature and concep­
tion of balance. In their careful statistical analysis of Eskimo data on giving and
receiving of resources, they discovered that by and large balance is not struck
between a pair of giver and receiver, throwing grave doubts as to the utility of the
concept of balance and to the nature of exchange in egalitarian societies. What they
did find is that at an ideological level, Eskimos do exhort reciprocity and balancing
of give-and-take. In fact, Pryor & Graburn maintain that the ideological conviction
of reciprocity seems to serve as an effective cover or "blind" for empirical imbalance,
enabling the imbalance to continue. This suggests the necessity of separating cul­
tural ideology from actual practice, a rather elementary methodological point in
anthropology which exchange theorists have neglected to observe.

RESOURCE TYPES

Since just about anything under the sun from smile and expression of respect to
giving of advice and material rewards can all serve as resources for exchange, how
 SOCIAL EXCHANGE 27 1

are we to conceptualize or systematize all different kinds of resources? The most

basic division in this respect is between positive and negative resources, the former
being those which are sought after and the latter those avoided. We have already
elaborated on the problematic nature of negative exchange and shall pass this topic
here. One way of dealing with the question of typology is to reduce all resources
to one category of phenomena. Nord (95), for one, argues that social eXchange, in
the final analysis, is a matter of seeking and conferring approval. Approval to him
is the psychological currency of social exchange. On this assumption, be brings
together a large body of research findings, mostly in social psychology, and reana­
lyzes them in an exchange framework. A problem with this approach is that people
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

seek to obtain many other kinds of things in social exchange besides approval. A
Access provided by McMaster University on 01/30/15. For personal use only.

recent study by Hurlbert (63), for example, illustrates the importance of social status
expressed in prestige, in faculty appointments and doctoral training in an­
thropology. Doctoral degrees are the valuables to be exchanged among departments.
If two departments can give each other their products, then they each confer prestige
upon the other. If a recipient department cannot reciprocate and give a PhD to the
department from which it receives one, then prestige flows one way.
For cultural anthropologists, what is important is the realization that prestige,
status approval, respect, love, affection, sympathy, etc, are English words, each of
which has a somewhat different meaning. To try to group all these under a single
heading of positive sentiment of some sort would do injustice to the semantic and
symbolic realities of these expressions.
Another mode of conceptualizing resource types is Blau's pair, "intrinsic" vs
"extrinsic" ( 19, pp. 35-38, 95-96), or Kapferer's pair, "sociational" vs "instrumen­
tal" (65, p. 164), the two pairs being essentially synonymous. The former, in Kap­
ferer's words, refers to those transactions "which are ends in themselves and not
detachable from their source," and the latter to those "which are essentially detacha­
ble from their source and are means to ends rather than ends in themselves." a
difficulty of these schemes is that concrete behavior itself is being classified as being
one or the other, when in reality concrete behavior is usually a combination of the
two elements in varying degrees. Befu's (14) pair, "expressivity" vs "instrumental­
ity," though in substance the same as Blau's and Kapferer's, avoids this difficulty
by defining expressivity and instrumentality as aspects of an act and by assuming
a given exchange act to manifest both qualities in varying proportion.
Using this distinction, Befu and his associates investigated the relationship be­
tween these two variables in Japan. Their as yet unpublished findings show that an
exchange act (or resource) which is regarded as high in instrumentality, such as
serving as guarantor or extending a personal loan, is also high in expressive value,
and vice versa. Also, a person who is a source of much expressivity is one who
provides a great deal of instrumental resources, and vice versa.
This set of findings contrasts with Foa & Foa's findings (44) from the United
States indicating possible differences in cultural disposition toward social exchange.
Foa & Foa classify resources into six categories: love, status, information, money,
goods, and services. These six are arranged in a "circumplex (44, p. 82, see Figure
2), such that the two opposites (e.g. love and money or status and goods) are least
likely to be exchanged for each other and that the two adjacent categories, such as
 272 BEFU

Mo re

•
Love

• •
Status Services
E
.!!?
�
'3
·M
(ij • •
n..
I nformation Goods

•
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

Money
Access provided by McMaster University on 01/30/15. For personal use only.

Less
Less Mo re
Concreteness

Figure 2 The cognitive structure of resource classes.

love and status or money and information, are most likely to be exchanged. Foa &
Foa's findings are not directly comparable to those of Befu and his associates, thus
requiring caution in interpretation. It appears, however, that Japanese tend to
exchange instrumental values and expressive values simultaneously and that as their
relationship becomes more and more intrinsically satisfying to one another, they
become more and more useful to each other. Foa & Foa's findings suggest that in
contrast to the Japanese pattern of exchange, in the United States instrumental
relationship, for example involving exchange of money, is kept separate from expres­
sive relationship where affect is exchanged.

SPHERES OF EXCHANGE

Bohannan (20) ha� divided Tiv economy into three spheres. The first, in order of
moral ranking from the lowest to the highest, consists primarily of foodstuffs and
is associated with the market. The second consists of prestige items, such as slaves,
cattle, metal bars, ritual offices, etc. The third is largely restricted to rights in
women, who are exchanged in marriage. Essential in Bohannan's definition of
sphere are (a) moral ranking of spheres and (b) nonoverlapping of categories of
resources among spheres. While exchange spheres manifesting these features are
known elsewhere in the world, e.g. Tikopia (43) and New Guinea (1 12), they are
far from universal. Barth (9, p. 157), for one, failed to find them among the Fur.
Against Bohannan's definition, Kapferer (65, p. 190) uses the term "transactional
sphere" to refer to a set of individuals involved in transactional activity who place
similar value on the benefits and services transacted.
Problems arise with Bohannan's definition in cultures, such as with the Pedi
( 1 14), where money is used both in market exchange and in bridal transactions.
Kapferer's definition also has difficulty in that one pair of individuals can engage
 SOCIAL EXCHANGE 273

in exchange in two different spheres, as when they are employer and employee and
are friends at the same time, thus engaging both in market and social exchange. My
proposal is to define "sphere of exchange" in terms of a system of valuation of
resources, such that a given resource would have approximately the same value in
a given sphere, but receives a different valuation in another. Thus a cigarette would
have a certain definable value in the market, but acquires a different one when given
to a friend as a gift. Even money, when it enters a social sphere of exchange, is no
longer interpreted on its face value but is evaluated in part on the basis of social
circumstances (64). Distinction between the market and the social sphere of ex­
change is an important one, as Bennett has cogently pointed out ( 1 7, p. 305). Blau
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

( 19, p. 94) and Heath, following Blau (56, p. 56), have claimed that the two spheres
Access provided by McMaster University on 01/30/15. For personal use only.

are distinguished by the element of trust, being present in the social sphere and
absent in the market. This is a gross distortion of the reality. Market transactions
cannot be carried out without mutual trust among participants, be they bankers and
account holders or buyers and sellers.

SYMBOLIC EXCHANGE

I have alluded to the importance of semantic and symbolic aspects of exchange

above. Conceptualizing exchange in terms of transmission of messages is a tradition
going back to Levi-Strauss and Mauss on the one hand and to symbolic interaction­
ists of the Chicago school on the other ( 1 , 122, 123). Reflecting this interest, half
of the papers contributed to the recent Kapferer volume deal with transaction of
meanings and symbols (32, 5 1, 88, 100, 101, 1 14). Among them, Sansom's concept
of "signal transaction" is of particular interest ( 1 14). Among the Pedi, as mentioned
above, bridewealth is nowadays paid in cash. But in order to keep bridewealth
payment out of the sphere of monetized market exchange, they use a metaphor of
domestic animals like cattle and sheep in reference to units of money to be paid in
bridewealth. In Jasper, according to Bennett ( 1 7), the reverse is true. There farmers
exchange labor, borrow each other's equipment etc in a system of exchange entirely
apart from the national market economy, but use the metaphor of the latter sphere,
such as "buy," "sell," "loan," etc to refer to their intracommunity exchange transac­
tions. Finally, Goody's analysis of greetings among the Gonja (52a) shows that they
use greetings as symbolic gifts and gifts as symbolic greetings.

POWER

Though several definitions of power have been offered by exchange theorists, proba­
bly the most useful one is that proposed by Emerson (39-41 ) and essentially fol­
lowed by Anderson & Befu (4), Blau (19), and Foa & Foa (44). Emerson conceives
exchange relationship as being predicated upon dependence of two parties upon each
other's resources. To the extent that Person is unwilling to voluntarily surrender
resource desired by Other and able to use this resource as a "carrot," to force,
coerce, or induce compliance upon the other party, one is said to have power over
the other party. This fact then allows both 'parties to have power over each other.
 274 BEFU

Thus power is not conceived of as a unilateral relationship where if Person has

power over Other, then Other does not have power over Person. Instead, power is
a bilateral relationship of each potentially having power over the other. In contract
negotation between employer and labor union, for example, the former's resources
include ability to improve labor conditions such as wage and fringe benefits and the
latter's resources, power to withhold labor through strike. Thus both sides can be
seen to have potential power over the other. When there is a clear differential
between the degree of dependence of one party over the other, then one may say
that there is power advantage of one over the other.
An important consequence of this definition of power is that the amount of power
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

becomes a function of the availability of alternative sources from which the compli­
Access provided by McMaster University on 01/30/15. For personal use only.

ant can obtain needed resources. This is illustrated in Barth's analysis of the Swat
Pathans (7), among whom the chief must be judicial in his exercise of power because
his followers can leave him and go to another chief and receive protection. Similarly,
in central Africa (52) when people are not confined to a village and instead can leave
one and goto another if the village headman is too demanding, the headman's power
is that much weaker. By the same token, for the power holder, availability of
alternative sources of labor, service, etc increases his power. When there is a large
pool of unemployed workers, it is the "buyer's market" in that the employer can
demand and expect more from his workers because the employer can fire the
recalcitrant and hire more obedient workers who are willing to work more for less
reward. Patron-client relationship is a special type of power relation which we have
'already discussed above, and will not be taken up in this section.
Two sources of power must be clearly distinguished: personal and institutional.
When the resources given out in exchange for compliance are personal possessions,
as they are for a patron in most cases, we speak of personal power. Institutional
power is found in an "office" which involves the right to exercise power. Thus
anyone who assumes that office is entitled to exercise power as prescribed for the
office, even though qualifications for assuming the office may not have anything to
do with one's ability to exercise the power of the office. The distinction here is best
illustrated with the Melanesian big-man and the Polynesian chieftain. The former
acquires his position as a big-man by dint of his own effort. Being a son of a big-man
often helps, as Strathern points out ( 1 27, pp. 210-1 1). "But a good start does not
guarantee success, for a son must be a good manager and speaker on his own account
if he is to become a big-man ( 1 27, p. 211)." On the other hand, when power is
institutionalized in an office such as a chieftainship, one can hold the office and
exercise the power of the office without having appropriate qualifications to do so,
e.g. merely by being born a son of a chief. Appreciation of this distinction is crucial
for a clear analysis of power in the context of social exchange. Another distinction
to be noted has to do with kinds of powers. The big-man's exercise of power is clearly
in the realm of social exchange in that his power is obtained, maintained, and
exercised in exchange for various benefits he provides his followers and rival big­
men. A chief or king's power, on the other hand, may not be entirely of this sort;
most likely it is based on a combination of coercive, normative and remunerative
sources, the first of which are, as we saw earlier, outside the domain of social
exchange.
 SOCIAL EXCHANGE 275

FUNCTION

Although as Mulkay (93) has demonstrated in his critique, social exchange theory
has developed out of dissatisfaction and rejection of the conventional structural­
functional approach, many empirical studies of exchange are still functionalist in
approach . Two major functions of exchange identified here are (0) saving and
investment and (b) social integration. These will be discussed in turn below.
To the extent that one has some control or knowledge as to what will be returned,
giving serves the purposes of savings and investment. Although in social exchange,
the giver is normally enjoined from specifying what is to be given back to him in
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

return, in most cases cultural rules specify in varying degrees what are appropriate
Access provided by McMaster University on 01/30/15. For personal use only.

categories of resources to be returned. Thus by knowing what one may get in re­
turn and who is likely to provide maximum return, one can maximize the invest­
ment in the giving of one's resources. This saving function is well illustrated in
reports by Beals (10), Bennett ( 1 7), Lombardi (82), Lomnitz (84-86), and Stack
( 1 26).
Lombardi's, Lomnitz's, and Stack's studies deal with the urban poor, and demon­
strate the critical function which informal social exchange of material resources
plays in keeping families above subsistence level, something which formal institu­
tions such as employment and social security programs alone cannot accomplish.
Bennett's study shows that North American farmers in Northern Plains are able to
rely on assistance in labor and equipment from fellow farmers because they have
provided similar assistance to them in the past. Similarly, in Oaxaca, according to
Beals, gifts are given consciously as investment, to be returned at strategic moments.
Although somewhat different, Gregory's reanalysis of Foster's concept of the image
of limited good in terms of "expectation of circumstantially balanced reciprocity"
is an attempt to see an exchange system as underwriting the economic security of
the community.
If savings and investment are the major functions of social exchange from the
"motivational" perspective, integration is the most obvious function in the "institu­
tional" approach, as Levi-Strauss has so abundantly shown in his theory of cross­
cousin marriage. When Paine (100) argues that what he calls the "incorporation"
mode, i.e. the institutional approach-but not the "negotiative" mode, i.e. the
motivational approach---of analysis would throw light on group integration, he is
echoing Levi-Strauss's position. The theme of alliance between exchanging commu­
nities is seen in Malinowski's account of the kula ring (87) as well as in Strathern's
analysis of moko in Mt. Hagen, New Guinea (127). The theme of community
integration is evident also in Vatuk's study of gift-giving in an urban neighborhood
in India and also in Rehfisch's study of competitive gift exchange between members
of different villages among the Mambila (t04a). Though somewhat different, John­
son's study of gift-giving among Hawaiian Japanese also falls in this category. Her
findings that status maintenance, generational continuity, conflict resolution, and
status equalization are accomplished by gift-giving indicate integrative functions of
social exchange, as does Young's interpretation that the institution of obutu (com­
petitive food exchange) serves the function of social control among the Kaluna in
Massim ( 132).
 276 BEFU

CONCLUSION

Looking at the state of the field, one gets the distinct impression that theory has run
ahead of empirical testing. A good deal of ink has been spilled over the conceptual
status and theoretical nature of social exchange without sufficient empirical check­
ing of the usefulness of concepts and models. We must make a concerted effort to
gather data and test propositions generated from our models.
A major contribution of social exchange theory is that it has offered a frame of
analysis at the collectivity level and also at the individual level and that it has at
least attempted to bridge the two levels, in spite of claims by Ekeh and others that
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

this is impossible. The field thus enables those working primarily in terms of cultural
Access provided by McMaster University on 01/30/15. For personal use only.

rules and those interested in strategy to converge and deal with common problems.
In this attempt we have been able to combine the disciplines of psychology and
anthropology without taking the conventional Freudian or neo-Freudian stand.
A strength of exchange theory is its linkage with economic and political an­
thropology. Relations to economics and economic anthropology are quite impor­
tant. How important it is would depend on whose approach one takes. For those
of Heath's persuasion, economics provides the foundation upon which social ex­
change theory is to be built. For those not so bent on the superiority of economics,
social and market exchange are both species of exchange, perhaps of coequal level
though belonging to different spheres. As such, at some abstract level, social and
market exchange are encompassed by a common theoretical umbrella. A rapproach­
ment between the two fields is inevitable as well as desirable.
Relationship to political science is also obvious from the point of view of the
definition of power. If politics has to do with "who gets what, when, and how," then
it is a question of acquiring resources and of strategies of acquiring them. Social
exchange provides a framework in which strategies are to be analyzed. Its usefulness
has been demonstrated by a number of empirical studies by political scientists (72,
77, 103, 1 1 7-1 19).
One of the areas in which little work has been done so far is the linkage of social
exchange theory with network theory (e.g. 5, 6, 22). A fruitful start has been seen
in Kapferer's study of African factory workers (65), but it remains the only major
attempt in this direction. Social exchange theory focuses on dyads as dyads. Even
when a third party is introduced, the basic framework is that developed in dyadic
exchange. Social network theory; on the other hand, begins with the assumption that
an individual is interacting simultaneously with a large number of people and
develops its frame of analysis on this basis. It goes without saying that what network
theory takes as given (multiple relations) is much more realistic than a series of
dyads which exchange theory takes as the starting point. Moreover, network theory
allows indirect relationships-friends of friends of friends-encompassed in its
theoretical constructs, something which exchange theory cannot do very well except
obliquely, for example, by linking a series of dyads in a chain. On the other hand,
social exchange theory is better suited to analyzing processes, strategies, and func­
tions, while network theory is better at developing structural framework. A combi.
nation of these two approaches then should yield a rather powerful theoretical basis
to analyze social phenomena.
 SOCIAL EXCHANGE 277

We hope that social exchange approach will receive more attention in the future
and be further developed and revised in the areas and directions indicated above.
There is no claim made here that social exchange theory will be a "unified theory"
which would account for most of social and cultural phenomena. The claim made
here is much more modest, yet even that modest degree of accomplishment awaits
future research.

ACKNOWLEDGMENTS

I wish to thank George Dalton and Theodore Downing for discussing various issues
in social exchange and offering helpful suggestions. I would also like to acknowledge
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org
Access provided by McMaster University on 01/30/15. For personal use only.

with gr atitude the generous grants and fellowships pr ovi ded by the National Science
Foundation (GS2370), the Japan Foundation, the John Simon Guggenheim Foun­
dation, and Stanford University Center for Research in International Studies, whose
sponsorship allowed me to carry out research which laid the basis for this paper.

Literature Cited

1. Abbott, C. W., Brown, C. R., Corsbie, statement. France-Asie/Asia 188:

P. V. 1973. Exchange as symbolic in­ 161-77
teraction: for what? Am. Sociol Rev. 13. Befu, H. 1968. Gift-giving in moderning
38(4):504-6 Japan. Monumenta Nipponica 23:
2. Abrahamsson, B. 1970. Homans on ex­ 445-56
change: hedonism revisited. Am. J. 14. Befu, H. 1974. Power in exchange:
Social. 76:273-85 strategy of control and patterns of com­
3. Adams, J. S. 1965. Inequity in social pliance in Japan. Asian Profile 2(5-
exchange. Adv. Exp. Soc. Psychol 2: 6):601-22
267-99 15. Befu, H. 1977. Structural and motiva­
4. Anderson, S., Befu, H. 1970. Power in tional approaches to social exchange. In
the context of social exchange. Proc. Social Exchange: Advances in Theory
Int. Congr. AnthropoL EthnoL Sci.. 8th. and Research. ed. K. F. Gergen, M. J.
Tokyo 2:247-49 Greenberg, R. H. Willis. New York:
5. Aronson, D. R., ed. 1970. Social net­ Wiley. In press
works. Can. Rev. Sociol Anthropol 7: 16. Belshaw. C. S. 1965. Traditional Ex­
No. 4 (entire) change and Modern Markets. Engle­
6. Barnes, 1. A. 1972. Social networks. Ad­ wood Cliffs, NJ: Prentice-Hall
dison- Wesley Module Anthropol. 26
17. Bennett, J. W. 1968. Reciprocal eco­
7. Barth, F. 1959. Political leadership nomic exchanges among North Ameri­
can agricultural operators. Southwest J.
among Swat Pathans. Monogr. Soc. An­
Anthropol. 24:276-309
thropol 19
18. Bennett, J. W., Ishino, I. 1963. Pater­
8. Barth, F. 1966. Models of social orga­
nalism in the Japanese Economy: An­
nization. R. Anthropol. Inst. Occas. Pap.
thropological Studies of Oyabun-kobun
23
Patterns. Minneapolis: Univ. Minne­
9. Barth, F. 1967. Economic spheres in sota Press
Darfur. In Themes in Economic An­ 19. Blau, P. 1964. Exchange and Power in
thropology, ed. R. Firth, pp. 149-74. Social Life. New York: Wiley
Assoc. Soc. Anthropol. Monogr. 6. 20. Bohannan, P. 1955. Some principles of
London: Tavistock exchange and investment among the
lO. Beals, R. L. 1970. Gifting, reciprocity, Tiv. Am. Anthropol. 57:60-70
savings, and credit in peasant Oaxaca. 21. Boissevain, J. 1966. Patronage in Sicily.
Southwest. J. Anthropol 26:3:231-4 1 Man 1: 1 8-33
11. Befu,H . 1964. Ritual kinship in Japan: 22. Boissevain, J., Mitchell, J. C eds. Net­
.•

its variability and resiliency. Sociologus work Analysis: Studies in Human In­
14:150-69 teraction. The Hague, Paris: Mouton
12. Befu, H. 1967. Gift-giving and social 23. Boulding, K. 1955. Economic Analysis.
reciprocity in Japan, an exploratory New York: Harper
 278 BEFU

24. Boulding, K. E. 1962. The relations of 41. Emerson, R. M. 1972. Exchange the­
economic, political and social systems. ory, part II: exchange relations and ex­
Soc. Econ. Stud. 1 1 (4):351-62 change networks. See Ref. 40, pp. 58-87
25. 1973. The Economy of
Boulding, K. E. 42. Etzioni, A. 1969. A basis for compara­
Love and Fear: A Preface to Grant's Eco­ tive analysis of complex organizations.
nomics. Belmont, Calif: Wadsworth In A Sociological Reader on Complex
26. Brady, I. 1972. Kinship reciprocity in Organizations, ed. A. Etzioni, pp. 59-
the Ellice Islands: an evaluation of Sah­ 76. New York: Holt, Rinehart & Win­
lins' model of the sociology of primitive ston
exchange. J. Polynesian Soc. 8 1(3):290- 43. Firth, R. 1939. Primitive Polynesian
316 Economy. London: Routledge
27. Breman, J. 1974. Patronage and Exploi­ 44. Foa, U., Foa, F. 1974. Societal Struc­
tation: Changing Agrarian Relations in tures of the Mind. Springfield, Ill:
South Gujarat, India. Berkeley: Univ. Thomas
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

Calif. Press 45. Foster, G. M. 1953. Cofradia and com­

Access provided by McMaster University on 01/30/15. For personal use only.

28. Burns, T. R. 1973. A structural theory padrazgo in Spain and Spanish Amer­
of social exchange. Acta Social. 19: 1 88- ica. Southwest. J. Anthropol. 9: 1-28
208 46. Foster, G. M. 1 96 1 . The dyadic con­
29. Campbell, J. K. 1964. Honour, Family, tract: a model for the social structure of
and Patronage, A Study of Institutions a Mexican peasant village. Am. An­
and Moral Values in a Greek Mountain thropol. 63: 1 173-92
Community. New York, Oxford: Ox­ 47. Foster, G. M. 1963. The dyadic con­
ford Univ. Press tract in Tzintzuntzan, II: patron-client
30. Economics and Pres­
Cancian, F. 1965. relationship. Am. Anthropol. 65:
tige in a Maya Community: The Reli­ 1 280-94
gious Cagro System in Zinacantan. 48. Foster, G. M. 1965. Peasant society and
Stanford Univ. Press the image of limited good. Am. An­
31. Cancian, F. 1966. Maximization as thropol. 67:293-2 1 5
norm, strategy, and theory: a comment 49. Frazer, J. G. 1 9 19. Folklore in the Old
on programmatic statements in eco­ Testament, Vol. 2. London: MacMillan
nomic anthropology. Am. Anthropol. 50. Gergen, K. J. 1969. The Ps;,chology of
68(2):465-70 Behavior Exchange. Readmg, Mass:
32. Cohen, A. P., Comarolf, J. L. 1976. The Addison-Wesley
management of meaning: on the phe­ 5 1 . Gilsenan, M. 1976. Dying, honor, and
nomenology of political transactions. contradiction. See Ref. 67, pp. 19 1-219
See Ref. 67, pp. 87-107 52. Gluckman, M., Mitchell, J. C., Barnes,
33. Damas, D. 1972. Central Eskimo sys­ J. A. 1 949. The village headman in Brit­
tems of food sharing. Ethnology 1 1(3): ish Central Africa. Africa 19:89-106
220-40 52a. Goody, E. 1972. 'Greeting,' 'begging,'
34. Davis, W. G. 1973. Social Relations in and the presentation of respect. In The
a Philippine Market: Self-Interest and Interpretation of Ritual, ed. J. S. Fon­
Subjectivity. Berkeley, Los Angeles: taine, pp. 39-7 1 . London: Tavistock
Univ. Calif. Press 53. Gouldner, A. W. 1960. The norm of
35. Dentan, R. K. 1968. The Semai: A Non­ reciprocity: a preliminary statement.
violent People of Malaya. New York: Am. Social. Rev. 25: 1 61-78
Holt, Rinehart & Winston 54. Gregory, J. R. 1975. Image of limited
36. Deutch, M. 1964. Homans in the Skin­ good, or expectation or reciprocity?
ner box. See Ref. 1 30, fP. 8 1 -90 CUrf. A nthropoL 1 6: 1 :73-92
37. Ekeh, P. P. 1974. Socia Exchange The­ 55. Heath, A. F. 1976. Decision making
ory, the Two Traditions. Cambridge: and transactional theory. See Ref. 67,
Harvard Univ. Press pp. 25-40
38. Embree, J. F. 1 939. Suye Mura, A Japa­ 56. Heath, A. F. 1 976. Rational Choice and
nese Village. Univ. Chicago Press Social Exchange: A Critique of Ex­
39. Emerson, R. M. 1962. Power-depend­ change Theory. Cambridge Univ. Press
ence relations. Am. Sociol. Rev. 27: 57. Hogbin, H. I. 197 1 . Polynesian cere­
3 1-4 1 monial gift exchange. In. Polynesia:
40 . Emerson, R. M. 1972. Exchange the­ Readings on a Culture Area, ed. A.
ory, part I: a psychological basis for so­ Howard, pp. 27-45. San Francisco:
cial exchange. In Sociological Theory in Chandler
Progress, ed. J. Berger, M. ZeJditch, 58. Hollnsteiner, M. R. 1967. Social struc­
2:38-57. Boston: Houghton Mifffin ture and power in a Philippine munici-
 SOCIAL EXCHANGE 279

pality. In Peasant Society, A Reader, ed. reappraisal of the concept of "on." Psy­
J. M. Potter, M. N. Diaz, G. M. Foster, chologia 1 2 : 1 29-38
pp. 200-12. Boston: Little, Brown 75. Lebra, T. S. 1972. Reciprocity-based
59. Homans, G. C. 196 1 . Social Behavior: moral sanctions and messianic salva­
Its Elementary Forms. New York: Har­ tion. Am. AnthropoL 74:39 1-407
court, Brace & World 76. Lebra, T. S. 1975. An alternative ap­
60. Homans, G. C. 1964. Commentary. proach to reciprocity. Am. Anthropol.
Sociol. Inq. 34:22 1-31 77:3:550-65
61. Homans, G. C., Schneider, D. M. 1955. 77. Lemarchand, R., Lemarchand, K. L.
Marriage, Authority and Final Causes: 1972. Political clientism and develop­
A Study of Unilateral Cross-cousin Mar­ ment. Compo Polito 4: 149-78
riage. Glencoe: Free Press 78. Levi-Strauss, C. 1949. Les Structures
62. Hostetler, J., Huntington, G. E. 1967. EIementaires de 1a Parente. Paris:
The Huterites in North America. New Presses Univ. de France
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

York: Holt, Rinehart & Winston 79. Uvi-Strauss, C. 1953. Social structure.
Access provided by McMaster University on 01/30/15. For personal use only.

63. Hurlbert, B. M. 1976. Status and ex­ In Anthropology Today, ed. A. Kroeber,
change in the profession of an­ pp. 524-53. Univ. Chicago Press
thropology. Am AnthropoL 78:272-84 80. Levi-Strauss, C. 1969. The Elementary
64. Johnson, C. L. 1974. Gift-giving and Structures of Kinship. Boston: Beacon
.

reciprocity among the Japanese Ameri­ 81. Lewis, H. S. 1974. Leaders and fol­
cans in Honolulu. Am. Ethnol. 1(2): lowers: some anthropological perspec­
295-308 tives. Addison- Wesley Module An­
65. Kapferer, B. 1972. Strategy and Trans­ thropol. 50
action in an African Factory: African 82. Lombardi, J. R. Reciprocity and Sur­
Workers and Indian Management in a vival. Unpublished manuscript
Zambian Town. Manchester: Univ.
83. Lomnitz, L. 197 1 . Reciprocity offavors
Press in the urban middle class of Chide. In
66. Kapferer, B. 1976. Introduction: trans­ Studies in Economic Anthropology, ed.
actional models reconsidered. See Ref. G. Dalton, pp. 93-106. Washington
67, pp. 1-22
67. Kapferer, B., ed. 1976. Transaction and DC: Am. Anthrol. Assoc.
Meaning: Directions in the Anthropology
84. Lomnitz, L. 1974. The social and eco­
of Exchange and Symbolic Behavior.
nomic organization of a Mexican shan­
Assoc. Soc. Anthropol. Common­ tytown. In Anthropological Perspectives
on Latin American Urbanization, Vol.
wealth, Essays in Soc. Anthropol., Vol.
4, Chap. 6. LAUR: Sage Publications
1 . Philadelphia: Inst. for Study Human
Issues 85. Lomnitz, L. 1977. Survival and reci­
68. Kaut, C. 196 1 . Utang no loob, a system procity-the case of urban. marginality
III Mexico. In Extinction and Survival in
of contractual obligation among Taga­
logs. Southwest. J. Anthropol. 17: Human Populations, ed. C. D. Laugh­
2SCr-72 lin, A. I. Brady. New York: Columbia
69. Kiefer, T. M. 1972. The Tausug: Vio­ Univ. Press
lence and Law in 0 Philippine Moslem 86. Lomnitz, L. 1977. The Survival of the
Society. New York: Holt, Rinehart & Unfittest. In press
Winston 87. Malinowski, B. 1922. Argonauts of the
70. Kolenda, P. M. 1963. Toward a model Western Pacific. New York: Dutton
of the Hindu Jajmani system. Hum. Or­ 88. Marriott, M. 1976. Hindu transactions:
gan. 22(1): 1 1-31 diversity without dualism. See Ref. 67,
71. Kuhn, A. 1963. The Study ofSociety, A pp. 109-42
Unified Approach. Homewood, Ill: Ir­ 89. Mauss, M. 1954. The Gift: Forms and
win-Dorsey Functions of Exchange in Archaic Soci­
72. Lande, C. H. 1973. Networks and eties. Glencoe: Free Press
groups in Southeast Asia: some obser­ 90. Mauss, M. 1925. Essai sur Ie don: forme
vations on the group theory of politics. et raison de I'echange dans les societes
Am. Polit. Sci. Rev. 67:103-27 archaiques. Annee Sociol. n.S. 1 :30- 1 86
73. Leach, E. R. 195 1 . The structural impli­ 91. Meeker, B. F. 1971. Decisions and ex­
cations of matrilateral cross-cousin change. Am. Sociol. Rev. 36:485-95
marriage. J. R. AnthropoL Inst. 8 1 : 92. Meggitt, M. 1977. Blood is Their Ar­
24-53 gument: Warfare Among the Mae Enga .
74. Lebra, T. S. 1969. Reciprocity and the Tribesmen of the New Guinea High­
asymmetric principle: an analytical lands. Palo Alto: Mayfield
 280 BEFU

93. Mulkay, M. J. 1971. Functionalism, Ex­ 109. Sahlins, M. D. 1963. Poor man, rich
change and Theoretical Strategy. New man, big-man, chief, political types in
York: Schoken. 220 pp. Melanesia and Polynesia. Compo Stud.
94. Newman, P. L. 1965. Knowing the Soc. Hist. 5:285-303
Gururumba. New York: Holt, Rinehart 1 10. Sahlins, M. D. 1965. Exchange value
& Winston and the diplomacy of primitive trade.
95. Nord, W. R. 1969. Social exchange the­ Proc. Am. Ethnol Soc., pp. 95-129
ory: an integrative approach to social 1 1 1 . Sahlins, M. D. 1965. On the sociology
conformity. Psychol Bull 7 1 (3):1 74- ofprimitive exchange. In The Relevance
208 of Models for Social Anthropology, ed.
96. Olson, M. 1965. The Logic of Collective M. Banton, pp. 1 39-236. New . York:
Action: Public Goods and the Theory of Praeger
Groups. Cambridge: Harvard Univ. 1 12. Salisbury, R. F. 1962. From Stone to
Press Steel Cambridge Univ. Press
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org

97. Orans, M. 1968. Maximizing in Jaj­ 1 1 3. Salisbury, R. F. 1976. Transactions or

Access provided by McMaster University on 01/30/15. For personal use only.

maniland: a model of caste relations. transactors? an economic anthro­

Am. Anthropol 70:875-97 pologist's view. See Ref. 67, pp. 41-59
98. Paine, R., ed. 1 97 1 . Patrons and Brokers 1 14. Sansom, B. 1976. A signal transaction
in the East Arctic. St. John's Memorial and its currency. See Ref. 67, pp.
Univ. Newfoundland Inst. Soc. Beon. 143-61
Res. 1 1 5. Schneider, H. 1974. Economic Man:
99. Paine, R. 197 1 . A theory of patronage The Anthropology of Economics. New
and brokerage. See Ref. 98, 8-2 1
Paine, R. 1976. Two modes 0fP.
York: Free Press
100. exchange 1 1 6. Schwimmer, E. 1974. Exchange in the
and mediation. See Ref. 67, pp. 63-86 Social Structure ofthe Orokaiva. Tradi­
101. Parkin, D. 1976. Exchanging words. tional and Emergent Ideologies in the
See Ref. 67, pp. 163-90 Northern District ofPapua. New York:
102. Polanyi, K. 1957. The economy as in­ St. Martin's
stituted process. In Trade and Market 1 17. Scott, J. C. 1972. The erosion of patron­
in the Early Empires: Economics in His­
tory and Theory, ed. K. Poianyi, C. M. client bonds and social change in rural
Arensberg, H. W. Pearson, pp. 243-70. Southeast Asia. J. Asian Stud. 32(1):
5-37
Glencoe: Free Press 1 1 8. Scott, J. C. 1972. Patron-client politics
103. Powell, J. D. 1 970. Peasant society and
c1ientelist politics. Am. Polito Sci. Rev. and political change in Southeast Asia.
Am. Polito Sci. Rev. 66:91-1 1 3
64(2):41 1-25
104. Pryor, F. L., Graburn, N. H. 1977. The 1 19. Scott, J . c., Kerkvliet, B . J . 1973.
How
myth of reciprocity. In Social Ex­ traditional rural patrons lose legiti­
change: Advances in Theory and Re­
macy: a theory with special reference to
search, ed.
K. J. Gergem, M. S. Green­ lowland Southeast Asia. Compo Stud.
berg, R. Willis. New York: Wiley. In Soc. Hist. 5:501-40
press 120. Silverman, S. F. 1965. Patronage and
l04a. Rehfisch, F. 1962. Competitive gift ex­ community-nation relationships in cen­
change among the Mambila. Cah. Es­ tral Italy. Ethnology 4: 1 72-89
tud. Afr. 3(9):91-103 121. Simpson, R. L. 1972. Theories ofSocial
105. Rosman, A., Rubel, P. G. 1972. Feast­ Exchange. Morristown, NJ: General
ing With Mine Enemy: Rank and Ex­ Learning Press
change Among Northwest Coast Soci­ 1 22. Singlemann, P. 1972. Exchange as sym­
eties. New York: Columbia Univ. Press bolic interaction. Am. Social Rev.
·
1 06. Rosman, A., Rubel, P. G. 1972. The 37:414-24
potlatch: a structural analYSis. Am. An­ 123. Singlemann, P. 1973. On the reification
thropol 74(3):658-71 of paradigms: reply to Abbott, Brown,
1 07. Rubel, P. G., Rosman, A. 1970. Pot­ and Crosbie. Am. Social Rev. 38(4):
latch and Sagali: the structure of ex­ 506-9
change in Haida and Trobriand soci­ 124. Sociological Inquiry 1964. Research
eties. Trans. NY Acad. Sci. 32(6): and commentary on the theorems and
732-42 perspectives of George C. Homans.
108. Rubel, P. G., Rosman, A. 197 1 . Pot­ Sociol. lnq. 34:2
latch and hakari: an analysis of Maori 125. Spiro, M. E. 1966. Buddhism and eco- ·
society in terms of the Potlatch model. nomic action in Burma. Am. Anthropol
Man 6:4:660-73 68: 1 1 63-73
 SOCIAL EXCHANGE 28 1

126. Stack, C. 1974. All Our Kin: Strategies 1 30. Turk. H .•Simpson. R. L. 197 1 . In­
for Survival in a Black Community. stitutions and Social Exchange. The
New York: Harper & Row Sociologies of Talcott Parsons and
127. Strathem, A. 1 97 1 . The Rope ofMoka. George C. Homans. Indianapolis:
Big-men and Ceremonial Exchange in Bobbs-Merrill
Mount Hagen. New Guinea. London: 131. Wolf, E. 1966. Kinship, friendship, and
Cambridge Univ. Press patron-client relations in complex soci­
128. Strickon, A., Greenfield, S. M. 1972. eties. In The Social Anthropology of
Structure and Process in Latin America. Complex Societies. ed. M. Banton, pp.
Patronage. Clientage. and Power Sys­ 1-22. London: Tavistock
tems. Albuquerque: Univ. New Mexico 132. Young. M. W. 1971. Fighting with
Press Food: Leadership, Values and Social
129. Titmuss. R. M.1971. The Gift Relation­ Control in a Massim Society. Cam­
ship. From Human Blood to Social Pol­ bridge Univ. Press
icy. New York: Random House
Annu. Rev. Anthropol. 1977.6:255-281. Downloaded from www.annualreviews.org
Access provided by McMaster University on 01/30/15. For personal use only.