'- A) T —

An". 1-,...

‘Lfi-m “v...

 

l.||

n 4
H

H

w. a
L
l 7|"

". w
H'

.l

I?
|

I
&

l

;"
3|

 

 

EXCHANGE MD 30051“. STRUCTURE

“masts for the Degree 0f M. A.
35%le STATE URE‘JERSE‘E‘Y
MICME; M. LOUKiNEN
1971

 

ABSTRACT

EXCHANGE AND SOCIAL STRUCTURE
BY

MICHAEL M. LOUKINEN

In the sociological and social anthropological literature, one can
discern two models of the exchange process. Neither of these models
are incorrect, but they are applicable to somewhat different empirical
situations. The author has constructed a theoretical argument incor-
porating scope conditions which cover the significant features of these
contexts.

The discussion begins with a brief description of two models of
the exchange process. Exchange phenomena will then be considered within
the context of a social structure. The focus will be on two basic pro—
cesses: the impact of social structure upon the ongoing transactional
sequences; and the role of the exchange process in shaping the form of
a social network.

A pivotal concept in this discussion is the idea of exchange con-
tent. It is defined as that which the transactors themselves perceive
as being "given" and "received". Social roles are considered as
channels which guide, restrict, and facilitate the flow of exchange
contents. Role expectations function to govern the appropriateness
of certain content in an exchange relation. An important issue is the

problem of combining different kinds of exchange content within

multiplex role relations.

Another important social structural dimension is the problem of
the transactor's social location. Transacting incumbents are treated
as if they are embedded in a larger social network that conditions
their exchange options. The conceptual portrait of social reality is
one of a collection of transactors welded into a much larger social
network. Exchange content is continually circulating throughout the

network in different kinds of "channels".

EXCHANGE AND SOCIAL STRUCTURE

By

Michael M. Loukinen

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Sociology

1971

.. ,\'
;p
TABLE OF CONTENTS
CHAPTER
I. Theoretical Perspectives of the Exchange Process .
II. A Theoretical Problem. . . . . . . . . . . . . . .
III. The Social Context of Exchange . . . . . . . . . .
A. Transactions Within Roles:
B. Transactions Within A Social Network . . . .
IV. Levels of Role Summation and Modes of Exchange . .
A. The Concept of Role Summation:
B. Exchange Content - A Conceptual Pivot: . . .
C. Transactions Within Low-Summated Relations:
A Maximizing Model:. . . . . . . . . . . . .
D. Transactions Within High—Summated Relations:
"A Total Social Fact Model": . . . . . . . .
V. Structural Consequences of Summated Exchange . . .
A. The Process of Exchange Summation
and Cognitive Similarity:
B. Summated Exchange and the Formation of
Structural Density:. . . . . . . . . . . . .
VI. Structural Consequences of Desummation . . . . . .
A. The Process of Desummation:
B. Specific Exchange and Structural Density: .
VII. Summated Exchange and the Formation of

Local Moral Systems. . . . . . . . . . . . . . . .

ii

10

13

16

2O

22

25

28

Social Boundaries and the Formation and
Persistence of Local Moral Systems:
Encapsulating Density and the Monitoring

of Self Interest: . . . . . . . . . .

iii

31

TABLE OF FIGURES
Page

Figure l . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figlme 3 O O O O O O O O O O O O O O I O O O O O O O O O 0 O O 0

iv

EXCHANGE AND SOCIAL STRUCTURE

I. Theoretical Perspectives 2£_the Exchange Process

 

 

Underlying the complex phenomena of social interaction is a
basic process of social exchange. Individuals are continually per-
forming transactions in relation to one another. In the sociological
and anthropological literature one can discern two different models
of the social exchange process. One model is based upon the work of
George Homans and it has been further elaborated by Peter Blau. This
theoretical apparatus was constructed from and checked with obser-
vations of social interaction in Western urbanized settings. Its intel-
lectual heritage lies in the very roots of Anglo-American cultural
history. The philosophical underpinnings are best described as "unre-
strained utilitarianism".

This particular model assumes that man's social behavior is de-
termined by a calculus geared toward considering cost—gain outcomes,
in the continuous pursuit of "profit". The components of this calculus
make homo sociologicus simply a slightly different version of the clas-
sical economist's homo economicus. All social activity is reduced to
nothing more than a set of acts performed in order to gain advantages.
As Blau states it,

"Social exchange is. . . limited to actions that

are contingent on rewarding reactions from others

and that cease when these expected reactions are

not forthcoming." (Blau, 196uz6)
In this statement we can find the distinguishing feature of this model.
The reciprocity process consists of a "quid pro quo" or "unconditional

one—for-one" (Sahlins, 1965: 1am) comparison of what is "given and

received". Social interaction is perceived as a game where players
always follow a maximizing strategy.

An alternative model of the exchange process was developed in
the French sociological tradition. We find its most thorough expression
in Marcel Mauss' Ih£_§iftfi and in the writings of Claude Levi-Strauss,

especially in The Elementary Structures 2£_Kinship.

 

Social exchange from the French perspective is not just a simple
maximizing process, but far more complex phenomenon. According to Mauss,
it is not so much an economic phenomenon, but is what he calls a "total
social fact". It is an event which simultaneously carries a social,
religious, magic, economic, utilitarian, sentimental, legal and moral
significance. (Mauss, 1925: 1) Such a "loaded" social significance
is not confined to ceremonial exchange. As Levi-Strauss notes,

"We must insist that this primitive conception
of the exchange of goods is not only expressed
in well defined and localized situations. It
permeates all transactions, ritual or secular,
in the course of which objects or produce are
given or received." (Levi-Strauss, 19H9: 53)

One can delineate several features which distinguish this model.
Firstly, reciprocity is not governed by short term or very explicit
"profit" considerations. Secondly, transactors are not self-maximizers,
but mérél_persons. Thirdly, a transacted item often communicates a
special/message. That is, it may contain semantic markings giving it
meanings far different from those falling within the scope of a utili-
tarian framework. Here we must include the handmade gift a child gives
her parent that communicates affection, a "joint" of marijuana commu—

nicating "I can trust you", and many other presentations having sym—

bolic character.

II. A_Theoretical Problem

 

These are the two different models of exchange behavior per-
meating contemporary social anthropological and sociological thought.
Current discussions about social exchange have reached a theoretical
impasse. We seem to be a bit stifled by our recognition of the two
different and seemingly contradictory positions. For some, the debate
has stopped right here. Others simply continue to impose a crude cost-
gain analysis upon all social phenomena regardless of their qualities.
This is absurd: consider a mother calculating the benefits she will
gain by feeding her child, and witholding the nourishment pending
certain reactions.

It is my contention that both of these models are heuristically
useful in research, but they refer to different kind§_of social relations.
The profit maximization model is based upon empirical evidence from ur-
ban areas where there is in fact a high distribution of these inter—
action patterns. And, it is not by accident that the natives partici-
pating in such social environments define it often as a "rat race".

The "total social fact" model was derived from data pertaining to a much
different kind of social relations. But it is important to remember
that both Mauss and Levi—Strauss claim that "primitive" exchanges do
occur in modern societies, (Levi—Strauss, 19u9: 55-56, and Mauss, 1925:
72—76) and that many economic anthropologists have found maximizing be-
havior in ”primitive" societies. (Burling, 1962: 80u, and Herskovits,
19uo: 3-2‘4).

Since both kinds of exchange behavior are found in both "primitive"
and "modern" societies, the next theoretical step should be obvious:

What are the general conditions producing these different forms of exchange

behavior? I intend to address myself to this problem, but in order
to do this it will be necessary to develop a new set of concepts to

interpret exchange relations within the context of social structure.

III. The Social Context 9f_Exchange

 

A. Transactions within roles:

When the concept of "role" entered the American sociological
vocabulary, several manifestos expressed hope that we had finally
found a concept to "link" the individual to society. (Linton, 19H5)
However, not only was the theoretical "bridge" never constructed,
but there has been surprisingly little contributed to our understanding
of how the concept can help us sift through the chaotic phenomena of
social structure. Nevertheless, I cannot think of a more frequently
used concept in sociological discourse. Yet it appears as if this
concept has not helped us understand the processes involved in social
interaction. Most of this stunted theoretical development is due to
the unfortunate semantic drift from the realm of the theater to actual
social life. The theatrical analog has cast a misleading phony or in-
sincere shadow on social interaction. Although alienated sociologists
find it appealing, there is still a large part of social life that is
not all pretense for an audience, but very serious activity.

In this discussion, I will avoid the theatrical imagery by con-
centrating on that intellectual tradition of role theory coming from
early treatises on legal theory. (Banton, 1965: ) The idea of "rights"
and "duties” associated with a particular role is a part of this heri~

tage. These components should be very useful in an exchange paradigm.

A transactor as a role incumbent must execute certain "prestations"

in fulfilling his duties, and on this basis has a right to claim cer-
tain benefits as reciprocation. The basic exchange consists of duties
(obligations) performed in return for rights (privileges).

A reciprocity process is implicit in the "mutual implication"
(Nadel, 1954: 50) of role relations. Dualities such as "ego-alterego"
(Parsons, 1951: 58) and "self—other" (Mead, 1956: 199) are isomorphic
with social interaction treated as a series of exchanges between two
transactors. Ideally, the duties executed by one incumbent are the
rights claimed by another, and vice versa. This conceptual "merger"
permits one to interpret the acts of "offering" and "reception" in
the language of both theoretical traditions.

By placing exchange within the context of normative culture
(Parsons, 1965: #3), one can avoid problems inherent in the maximization
model. Homans assumes the existence of isolated, non—socialized indiv—
iduals and ignores society. A crude utilitarianism is the only culture
he assumes man possesses. On the other hand, the French tradition is
dominated by the Durkhemian dictum: To begin the analysis one must
first assume the existence of society as a social fact. A social fact
central to this discussion is a role structure.

"the role concept is not an invention of anthro-
pologists or sociologists but is employed by the
very people they study. No society exists which
does not in this sense classify its population -
into fathers, priests, servants, doctors. . .
that is, in accordance with the jobs, offices or
functions which individuals assume and the entit—
lements and responsibilities which fall to them."
(Nadel, 1957: 20).

A conception of role structure is necessary because the execution

of certain transactions are required for any role incumbent. Roles

function to channel the regular transactions in a social structure.

Nadel's schema for classifying role components will be useful
when considering role performance as required transactions. In order
to avoid a static conceptualization, transactions can be graded into
a hierarchy of relevance (Nadel, 1957: 31) Let us distinguish three
main grades. (a) Certain transactions are peripheral or optional
because their omission or variation does not effect perceptions of
role performance. For instance, a milkman may purchase stock in a
soup company, but a judge ruling on the company's health standards
may not. (b) Some transactions are "sufficiently relevant" because
their execution influences the perception of role performance. If
these transactions are not properly accomplished, the incumbent's be-
havior will be perceived as deviating from the "normal" and will evoke
sanctions. (c) Other transactions are simply mandatory for the in-
cumbent because there is no normative tolerance regarding their exe—
cution. They are basic or "pivotal" to the role and "govern" of
imply the other transactions. Their (its) absence alters the whole
identity of the role. Take for example a salesman who simply cannot
sell a product.

The necessity for thinking of transactions occurring within role
expectations is apparent when studying the patterns of exchanges in
southern, rural Ireland. Arensberg and Kimball found a very sharp
role segregation based upon sex. Only the mother could prepare butter
and eggs and sell them at the market; and only she might purchase
commodities for home use. Similarly, the father was the only transactor
who could purchase tools and equipment for use in the fields and with

the animals; and only he could sell the cattle. (Arensberg 8 Kimball,

1940: 49). Virtually all interpersonal transactions were influenced
by the sex role of the transactor.

"That a man should concern himself with a

woman's work, such as the sale of eggs or

the making of butter is the subject of der—

isive laughter. . . Ridicule and graver

charges of unmanliness and unwomanliness

can be brought to reaffirm the conventions

against the offender in either sphere."

(Arensberg 8 Kimball, 1940: 49).

The utility of this approach is very evident when one considers
episodes where transactions are incongruent with role expectations I
am not referring to blatant violations, but the subtle "mixing” of
transactions in different role contexts. For example, the door to door
salesman who immediately calls you by your first name in order to in-
ject a friendship facade into the transaction setting. All pseudo-
gemeinschaft attempts fall into this category. Another example would
be the employee and supervisor who are trout fishing comrades. The
employee may sincerely give his fishing pal a new fly rod on his birth-
day, but not within a role context where other employees are present.

B. Transactions within a Social Network:

It is not enough to consider exchanges only within role expec-
tations confined to the self-other duality. In both their theoretical
formulations, Homans and Blau dwell upon a diadic model, and most of
their examples are drawn from two—person transaction episodes. But
transactors simply do not exchange in isolation, but within a larger
society. Person X may alter his mode of exchange with Person Y, not

only in response to Y's sanctioning reactions, but because a third

party "intervenes" in some fashion. (Nadel, 1954: 51). Who is this

intervening agent on behalf of normative culture? One need not claim,

a la Durkheim, that "society" or the conscience collective conditions

 

exchange processes; it is possible to be much more precise.

In this discussion, the concept of social network (Mitchell, 1969)
will extend the analysis beyond the restricted scope of the diad. A
social network may be defined as a non-bounded set of individuals, some_
of which transact with each other, but seldom in the presence of the
complete set. These two conditions are included in the definition to
distinguish a network from a social group. In groups, almost all in-
dividuals interact to some extent with everyone else, and usually when
the entire aggregate is mobilized into a common situation.

The portrait of social structure presented thus far is one where
individuals are transacting with other individuals. Exchanges are
channeled throughout a social network by a role structure. Transactions
in one role relation are conditioned by the other role relations con—
necting the two transactors, and by the exchange relations each parti-

cipant has in the wider social network.

IV. Levels 2£_Role Summation and Modes 2f_Exchange

 

A. The Concept of Role Summation:

Most of the social anthropological literature dealing with de-
scriptions of "primitive" or rural social structure contains remarks
about the multifaceted character of the social relations. This feature
has been discussed in terms of a "simplex-multiplex" dimension (Gluckman,
1955: 19 6 1962: 24), as a "specific-diffuse" pattern variable (Parsons,
1951: 86), and as "summated and non-summated" roles (Nadel, 1954: 63).
That persons perform a varying number of different roles in relation to
the same set of persons is the central theme of these discussions. A
typical instance is where two individuals interact as kinsmen, super-
visor-subordinate, employer-employee, and fellow congregation members.
In this discussion, the focus will be upon the number of different role
relations connecting any two incumbents. The greater the number of
roles two individuals perform in relation to one another, the more sum-
mated is their relationship. (For expository reasons, highly summated
role relations will often be called "diffuse", while the term "specific"
will refer to low levels of summation.)

B. Exchange Content — A Conceptual Pivot:

Now I shall attempt to demonstrate that the two different models
of exchange processes are associated with different levels of role sum—
mation. The crucial concept in this theoretical shift is the "content"
of exchange. Exchange content may be defined as the set of material or
non—material media items that incumbents themselves perceive as being
transacted. The actual media may be different kinds of information,
material commodities, services of various kinds, and so forth. The im-

portant point is that the concept refers to the cognitive systems whose

lO

phenomenal distinctions or semantic categories are derived from con-
trasts and discriminations regarded as appropriate by the transactors
themselves. Hence, a description of exchange content would consist

of a set of emic statements containing the transactor's natural lexemes.
(Kay, 1969: 10) Such statements can be falsified only if they contra-
dict the cognitive calculus by which the relevant transactors judge
items to be similar or different (Frake, 1962: 76).

It seems to me that a particular set of role expectations es-
tablishes boundaries or places limits for the appropriate semantic do-
main of exchange content. Some kinds of information, services or mater-
ial may be exchanged within a particular role (with varying degrees of
legitimacy), while other kinds are not permissible. The basic premise
of this discussion is cast in the following argument: If role expec-
tations limit transactions to certain appropriate kinds of content,
then the range of exchange content (total number of different kinds of
content) transmitted between two incumbents is proportional to the
level of role summation. That is, the higher (lower) the level of sum-
mation in an exchange relation, the wider (narrower) is the range of
transacted content.

C. Transactions Within Low—Summated Relations:

A Maximizing Model.

It is my contention that the maximizing model of exchange behavior
is appropriate only for minimally summated role relations. Since the
relation is specific, a narrow range of content is transacted between
the incumbents. The restricted domain of exchanged content is impor-
tant because this provides the transactors with an opportunity to com-

pare what was "given" and "received” during an exchange. For this very

11

specific exchange relations will usually have a built—in equivalency
standard for recurring transactions. A labor service exchange of only
one kind (very specific) can very easily be categorized into commen-
surable units based upon some such criteria as time input, or the num-
ber of tasks completed. If a scientist must seek the advice of a col-
legue on a particular "problem" (task unit), he may be obliged to re-
ciprocate by offering his advice when his collegue needs help with a
"problem". Or, a mother may babysit for a neighbor for "three hours"
or "while she goes shopping" and expect that neighbor to reciprocate
for an equivalent number of comparable units. When a narrow range of
content is transacted, it is simply that much easier for the individuals
to break it down into units for quantifiable comparisons.

The facility of making content comparisons produces a marked con-
cern for equivalence and equity. One of the most distinguishing fea-
tures of specific exchange relations is the extent to which attention
is focused upon the "balance sheet". The Homans and Blau model describes
this type of reciprocity process where the transactors continually try
to assure that the value gained for them is greater than or equal to
the amount "invested".

Such a concern for cost-gain ratios produces a marked concern for
immediate reciprocation. Obligations are discharged as soon as possible,
thereby shortening the temporal dimension between prestation and recip-
rocation. The duration of obligation is so brief that both transactors
have a distinct conception of a completed exchange cycle. At the lower
levels of role summation, it is not uncommon for the transactors to
make explicit stipulations regarding the proper content for reciprocation,

the proper number of units, and possibly even the appropriate time for

12

reciprocation.

The primary purpose in low summated exchange relations is the
realization of "profit". Using Nadel's schema, one could say the
resources gained constitutes the pivotal transaction condition. Thus,
interaction is only a means to an extrinsic specific end: collection
of resource units in amounts beyond that invested. Profit is the nec-
essary condition of the relation; if it is not forthcoming, the re-
lation will terminate.

It is not surprising that incumbents of specific exchange re-
lations will perceive each other as resources. Thus, it would not mat-
ter to a transactor wh2_the individual happens to be, as long as he
is instrumental to the acheivement of a selected goal. Because his
value is only related to his ability to foster the other's goal, his
unique characteristics are irrelevant to the relation. He may be
substituted without significantly impairing the flow of transactions.
Under such conditions, the social bond is not very intense and is
easily terminated.

Specific exchange relations were what Max Weber defined as
zweckrationale action. He characterized it in terms of an orientation
toward a

"system of discrete individual ends, that is
through expectations as the behavior of objects
in the external situation and of other indiv-
iduals, making use of these expectations as
conditions or means for the successful attain-
ment of the actor's own rationally chosen ends."
(Weber, 1964: 115).
This type of action involves the "weighing of the relative importance

of ends", "a consideration of the benefits in relation to the undes-

irable consequences." Efficiency, expediency, and the "amount of

13

return" are the dominating considerations.
Toennies captured the essential features of minimally summated
exchange relations in his portrait of gesellschaft society.
"Nobody wants to grant and produce anything for
another individual, if it be not for a gift of
labor that he considers at least equal to what
he has given. . . In Gesellschaft, every person
strives for that which is to his own advantage
and affirms the actions of others' only in so
far and as long as they can further his inter—
est". (Toennies, 1957: 65).
D. Transactions Within High-Summated Relations-
A "Total Social Fact" Model:

As the level of role summation increases, the incumbents engage
in a wider variety of activities together, and consequently, the range
of exchange content expands. When a considerable array of content is
transacted through summated role relations, it is difficult, if not im-
possible, for the transactors to cognitively break it down into com-
mensurable units. The sheer diversity and volume of transacted content
prevents opportunities for them to construct quantifiable comparisons
across the content range. One can imagine a situation where two brothers
are always helping each other with work tasks, loaning each other tools
and equipment, making purchases for the other while in town, and whose
wives usually give each other baked goods. After a while it becomes
meaningless to compare apple pies with chain saws, and rides to church
with repaired roofs. This does not mean that there can be no specific
transactions with commensurable content units within such a relation:
but that within the context of the total content, such comparisons are

very difficult. While Firth noted that certain commodities exchanged

by the Tikopians are regarded as "reciprocally equivalent" and there is

14

a "general rating of objects on the basis of their respective util—

ities", he noted:
"there cannot be said to be a comprehensive scale
on which every type of good is placed and bears a
measurable relation to the type of good above and
below it. As far as material goods are concerned
there is no class of object which serves as a unit
of measurement for all the others, and there is no
conversion of the worth of different classes of
object into one another in a systematic way.
(Firth, 1939: 339).

In such diffuse exchange relations, the transactors will not
focus as much attention toward a balance sheet, so that one does not
find a regulating concern for equivalence and equity. Cost—gain con-
siderations are not an integral part of the relation, and if they do
occur at all, one cannot understand them solely in terms of the quid—
pro-quo mode of reciprocity. A much closer approximation of this re-
ciprocity process is found in Sahlin's concept of "generalized recip-
rocity" (Sahlins, 1965: 144). The main difference is in the length of
the time interval between prestation and reciprocation; in summated
exchange relations it is much longer. The social obligations incurred
by a prestation are so long lasting that the transactors do not enter-
tain conceptions of distinct exchange cycles. Stipulations concerning
time, content or units of the reciprocation are seldom encountered.

One of the most central characteristics of summated exchange is
the scope of moral regulation. These relations are usually the product
of a long history of transactions, and this produces a rather uncondi-
tional trust for the partner. Each transaction has a considerable moral
complexion because injustices and conflicts appear to spread from the

role in which they occur, to all the other roles connecting the incum-

bents. (Gluckman, 1962: 50). This potential "spreading” quality of

15

conflicts prevents the activation of blatant self interest in any

specific role. Incumbents are very cautious not to make any drastic
social mistakes for fear of the mushrooming consequences. One often
finds an excess of ritualization employed by the incumbents in order
to semantically mark appropriate expectations. (Gluckman, 1962: 24).

Social exchange in these relations is not determined by an ex-
trinsic set of goals, but is often a desired end in and of itself.
Summated transactors perceive each other as unique individuals, and
not as instruments at attain goals. In fact, the persons are not
usually selected for exchange in order to achieve goals; but goals
are often altered in order to maintain contact with valued individuals.
The incumbent is valued for special qualities and is not interchange-
able.

It is difficult to specify a pivotal condition (8) for a summated
exchange relation. The incumbents know each other very well because
much of the information exchanged is intimate. But such multibonded
relations are often abrubtly terminated, and usually with intense em-
otional consequences. Perhaps the pivotal condition might be best de-

scribed as a generalized loyalty to the other incumbent. The violation

 

of a well known rule, like not aiding in a crisis situation might be re-
garded as a breach of loyalty. Violations must be severe to rupture the
social bonds. They are the product of a history of good faith and trust
and the incumbents entertain expectations of continuity in their asso-

ciation.

16

V. Structural Consequences 2£_Summated Exchange

 

 

A. The Process of Exchange Summation and
Cognitive Similarity:

Homans asserted that as the frequency of interaction increases,
similarity and liking will follow. One can think of many instances
where this proposition is sound, but the simplicity of its formulation
obscures important social processes. One might ask, what conditions
prevent similarity and liking when associated with a substantial amount
of interaction? A counter-example might be a relation regulated by
status inequality. Some scholars think this factor may prevent liking,
or may even generate enough hostility to eventually terminate the re-
lationship. Supervisors encounter intense conflicts when friendship
relations develop between themselves and their subordinates. Elabor-
ate ritual procedures are activated to segregate these role relations.
Military officers must segregate their leisure activities from non-
officers by confining them to Officer's Clubs. Factory supervisors
usually sit at different lunch tables than workers.

Most of the literature concerning this issue focuses on friend-
ship relations as an impediment to the effective functioning of author-
ity. But is it the authority relation per se that prevents the devel—
opment of similarity and liking? It seems to me that one can find a
more abstract scope condition to include this factor. Let us consider
an example where the authority relation is not quite so salient.

Mr. Jones interacts very frequently with his secretary, Mary,
but exchanges are limited to matters concerning their occupational
duties. He is very brief and directly to the point in all his commun—

ication: Never does he allow conversation to stray toward content

17

unrelated to their work. Once when Mary asked him about his children
he promptly reminded her that there was work to be done, and that was
their reason for being there. Afterwards he would simply ignore any
extraneous conversation she might attempt. This made Mary uncomfor-
table. Now she tells the other secretaries that it is horrible to
work with Mr. Jones not because he is authoritarian, but because he
is a robot. The efficient executive quickly gained a reputation for
being unfriendly. Is it the authority relation that prevents liking
and similarity? Let us look at one more example.

A large office building was unoccupied at nights, except for a
three man janitorial crew and a night watchman. The task requirements
for the janitorial crew were minimal and they usually spent the latter
half of the night drinking beer and "shootin' the bull". The watchman
was not their superior, in fact, he worked for a different organization.
He limited his transactions with the janitors to strictly task content,
like asking them to turn off the lights and lock the doors. After
asking him to join them for a few drinks, the janitors began to dis-
like the watchman because he always refused. He was no longer a "good
guy", they thought him unfriendly, and worst of all, he might even be
a "squealer". They ceased their informal cooperation with him, leaving
doors unlocked, turning fans on and failing to turn them off, and
leaving all the lights on after they finished cleaning. The watchman
decided to drink beer and "shoot the bull" with them. Shortly after-
wards, the janitors resumed their informal cooperation.

There seems to be something more general going on in these ex-
amples than conflicts due to authority. It was not the authority
factor disturbing the secretary and the janitors, but a person's re-

fusal to expand the range of exchange content. I think that before

18

the mere increasing frequency of interaction can produce similarity
and liking, specific exchange relations must break down to let sum-
mation occur. It seems as if during the course of recurring specific
transactions, the incumbents acquire an expectation that role sum—
mation 22gh£_to occur. Both the secretary and the janitors made offers
to broaden the base of exchange, and they were very upset when their
offers were rejected. It would appear that summation is a natural
social process.

There are several plausible explanations of why transactors
would appear to desire summated exchange relations. Summation is a
necessary condition for the development of cognitive similarity since
it permits a wider range of information content to enter the flow of
message exchanges. Cognitive similarity cannot develop if the trans-
acted message content is restricted.

Homans would solve the problem by defining similarity as a
rewarding experience. (Homans, 1961: 103). The social psychological
balance theorists would argue that uncertainty is unpleasant and pro-
duces tension, therefore man seeks certainty to escape this tension.
Since man seeks certainty, he will attempt to compare his opinions
with those in his social space. Specific exchange relations prevent
one from comparing a comfortable range of beliefs, and summation a1-
leviates tension by expanding the permissible range of compared beliefs.
(Festinger, 1968).

Another plausible explanation may be constructed using the con-
cept of role distance. (Goffman, 1961). Evidently recurring trans-

actions in specific role relations puts a strain upon the incumbent.

19

This strain might be produced because such relations prevent a person
from presenting his entire self into the exchange process. The more
specific a role, the smaller the segment of self that may be presented.
Role distance refers to the discrepancy or gap between an actor's real
self and the self presented in role activity. Perhaps the role dis—
tance is greater in specific relations, and incumbents might try to
broaden the transaction base in order to present a complete self.

All of these explanations are plausible, but they do not fit so
well into an exchange paradigm. On the other hand, summation might occur
simply because it bestows certain advantages to the incumbents. Since
these relations are regulated by expectations of continuity and trust,
and not in terms of immediate gains in short-term cycles, they repre—
sent a form of social "insurance" for the incumbents. The expectations
of continuity and trust give each incumbent good "credit-ratings" with
each other so that resources are readily available for immediate needs.
Such exchange relations are highly dependable and provide transactors
with a social base of security. Regardless of the reasons why summated
exchange evolves out of specific transactional relations, it is a nec-
essary condition for the development of cognitive similarity.

As summation occurs, the range of exchanged messages expands and
the number of beliefs compared between two transactors increases dras—
tically. Transmitted messages focus upon similar beliefs and avoid dis-
similarities. Person perceptions shift from a configuration of the
incumbent as a means to obtain goal objects to an individual with unique
characteristics and strongly held convictions. Expectations become

diffusely oriented toward the personal beliefs of incumbents. At this

point, when liking has occurred, disagreements become increasingly

2O

painful and the discovery of similarities more rewarding. In this
way, summation promotes normative cultural uniformity.

B. Summated Exchange and the Formation of

Structural Density:

The concept of structural density (Barnes, 1968) refers to an
important morphological property of social networks. Density may be
defined as the ratio of the actual number of social relations con-
necting a set of persons to the theoretically possible number (Barnes,
1968: 61). In other words, it measures the extent to which everyone
knows everyone else. Peasant societies, small stable communities, and
marginal enclaves usually are very high in structural density. One of
the most important consequences of summated exchange is its tendency to
produce structural density. Let us see how this occurs.

Assume two actors are engaged in a summated exchange relation,
and one of these incumbents maintains a summated relation with a third

person. This may be represented as follows:

 

 

 

 

 

 

 

Figure 1

Let A and B, and B and X represent a set of transactors engaged in sum-
mated exchange relations. Let B and Y represent a set of transactors
involved in a specific transactional relation. I would like to suggest

that because both A and X hold summated relations with B, there are

21

psychological pressures upon them to like each other. But neither is
obliged to like Y who only maintains a specific relation with B. One
simply feels obligated to like and speak well of the friends of our
friends. This is reflected in folk statements such as: "Any friend of
yours is a friend of mine;" and "We consider him just one of the fam—
ily." Norms prohibit one from saying derogatory things about friends
of your friend while in his presense.

One must remember that summated relations require a great deal
of time, energy and resources to develop and maintain. Summated in-
cumbents are in rather constant contact with one another. Since an in-
dividual cannot be in two places at one time, it is difficult to main-
tain separate summated exchange relations. The time, energy, and re-

source requirements for segregated summated relations increases in

 

proportion to the number of such relations a person tries to maintain.
It follows that if the number of summated exchange relations for any
incumbent increases, there is a greater probability that he will have
to interact with more than one of these persons at the same time.
Hence, as summation occurs between any two transactors, the incumbents
of their respective summated role sets will probably develop exchange
relations with each other. This process is the formation of density

in a social network.

22

VI. Structural Consequences gf_Desummation

 

 

A. The Process of Desummation:

Desummation refers to the process by which summated exchange re-
lations are rendered more specific. It involves disengaging certain
content from the exchange relation. Such a decrease in the range of
transacted content increases the probability of the incumbents breaking
down the remaining content into commensurable units. One would then
expect a rise in the frequency of cost—gain considerations and inequity
perceptions. Summated exchange relations are regulated by a diffuse
balance conception covering the total range of content. The trans-
ferring of exchange content to another relation upsets this balance.
Sudden shifts toward specific exchange can generate incredible hostil—
ity. When this does occur, the transactors will break down the general-
ized reciprocal pattern into a reconstructed history of the relation in
terms of inequities associated with a series of distinct transaction
cycles. Let us look at a few empirical instances of desummation.

A marriage relation is extremely summated. When the incumbents
love each other, the reciprocity process is very generalized. The
transactors do not strive to maximize their own profit, but are rewarded
when the other benefits. Each does so much for the other, they think
nothing of it and are unaware of giving something in return for some-
thing else.

When desummation begins, the couple does fewer activities together
and their relation gradually becomes more specific. As the range trans-
acted content shrinks, unit comparisons become more feasible. Inequity
perceptions become increasingly frequent. Arguments serve to redefine

the history of their marriage in terms of a sequence of injustices.

23

These inequities are conceived in terms of a series of distinct spe—
cific transactions: "Last week you used the car while I had to stay
home and watch the kids."

If the marriage collapses and the incumbents decide to termi-
nate their relation and divide their property, hostility overwhelms
them. The process of deciding who deserves what marks the bottom level
of specific exchange. Interaction is reduced to the lowest level of
forced comparison as they recall the entire history of their asso—
ciation in order to maximize their perceived deserved share. Most
divorce lawyers agree that if a couple do not hate each other when
they file for a divorce, they will once it gets started.

The problematic consequences of desummation are evident when
cash transactions occur in summated relation. A cash—based transfer
gives a brief but very uncomfortable taste of desummation. Such
transactions are very problematic with kinsmen. There are folk
warnings against making economic deals with relatives: One should
never buy a car from an in-law, or rent an apartment from an uncle.
The negotiation of debts with friends is also a painful transaction
that is very ritualized. Friends attempt to azgid_concern for the
equity dimensions, and try to get such unpleasant activity finished
as soon as possible.

Another instance of a highly summated exchange relation is the
”patron-client" relation that social anthropologists have analyzed in
precommercial agrarian societies. (Paine, 1971). This particular
summated relation often provides a vital interclass link between
the landed aristrocracy and peasantry populations (Moore, 1966). A

very similar relation is found in the landlord—sharecropper bond in

24

the Deep South (Davis 8 Gardner, 1941).

The content of this relation is fairly standardized. Usually,
the landlord would provide the land, tools, housing, supervision, pro—
tection, and the administration of local justice. (Moore, 1966 8
Silverman, 1970). As time passed and summation increased, he would
help the peasant during unexpected crises, contribute advice and in—
formation about the outside world, and sponsor festive occasions. He
was the protector upon whom the peasant could depend (Paine, 1971). On
the other hand the peasant would be obliged to contribute many kinds of
labor throughout the year, and to give the landlord the major share of
the surplus (Moore, 1966). As the level of summation increased the
peasant would be expected to give his loyalty. He would support his
master politically in war and peace, and would support him socially by
speaking well of him and enhancing his reputation and prestige (Silver—
man, 1970).

As the urban market expanded and the scope of the state's author-
ity spread into the countryside, this complicated exchange relation was
desummated (Moore, 1966). As Karl Marx noted:

"The bourgeoisie, where ever it has got the upper

hand, has put an end to all feudal, patriarchial,

idyllic relations. It has pitilessly torn assun-

der the motley feudal ties that bound man to his

"natural superiors", and has left no other bond

between man and man than naked self-interest, than

callous "cash payment". (Marx, 1848: 11).
This relation was stripped to the "cold cash nexus" as commercialized
agriculture grew in response to the demands for food in the swelling
urban areas. (Moore, 1966). The emerging nation—state assumed the

lord's tasks by providing protection and administering justice as it

gained a monopoly of force. Thus, content from the lord—peasant ex-

25

change relation was transferred to an emerging state apparatus, and
the former relation became more specific. As the range of transacted
content decreased, all that was left was a specific monetary relation
between lord and peasant. As opportunities for content comparisons
increased, the social attention turned toward equity considerations
and evoked cries of "exploitation" (Moore, 1966 8 Silverman, 1970).
Such sudden desummation might very well have been the Spark igniting
the great rural rebellions in twentieth century China and the Soviet
Union (Moore, 1966). It is now happening in rural Italy, and in some
parts of Latin America (Silverman, 1970).

I think these illustrations demonstrate the hazardous conse-
quences of shifts from summated exchange to more specific transactions.
Even in the marriage relation there is some kind of an accounting pro-
cess going on, but it has a very latent quality. When desummation sets
in, it is as if all the repressed hostility and aggression had been
accumulating beneath a veneer of affection, and were suddenly released
as the incumbents were forced to settle accounts.

B. Specific Exchange and Structural Density:

Specific exchange activity appears to involve the destruction of
structural density. Once desummation has set in, a smaller range of
information is exchanged and this restricts the development of similar-
ity and liking. Incumbents perceive each other as resources and often
dislike each other as they regard cooperative activity as formal and
strenuous. One encounters no expectations obliging incumbents to like
the strictly occupational contacts of their own collegues.

If a social network is characterized by a high distribution of

specific exchange relations, one can expect interpersonal competition

26

to occur when resources become a bit scarce. If each individual is
pursuing a strictly maximizing strategy in his own self-interest, one
should be able to pinpoint an optimal mode of rationality for their
conduct. Since the incumbents perceive each other as resources, they
may be unwilling to introduce incumbents of their direct exchange set
to each other. Persons might be perceived as scarce resources and
transactors would attempt to block access to their own resource supply
persons. Transactors protect their own resource base as an optimally
rational strategy.

Consider this illustration. Let person A have a specific ex—
change relation with B, and B engages in specific transactions with X.
B exchanges content k_with A, in return for content 1, and B also ex-

changes m_for £_with person X. This may be represented as follows:

8
A X

Figure 2

If B introduces X to A, X may provide A with another market for l, or
with a less costly supply of k) or with alternative contents substitu-
table for k_and l_at more attractive terms. It is also quite possible
for X to obtain m from A, or A might even provide X with a market for

n: It would be rationally disadvantageous for maximizing transactors

27

to introduce their first order exchange partners to each other. In-
stead, they seggegate contacts to insure a stable resource base.

Such a competative exchange strategy entails the dissolution of
structural density. Transactors know their minimally summated relations
are tenuous and subject to alteration upon the perception of new advan-
tages, because very little connects them with their partners. Trans-
actors try to prevent each other from cultivating opportunities that
might harm the current flow of transactions, while at the same time

searching for better exchange options for themselves.

28

VII. Summated Exchange and the Formation of Local

Moral Systems

 

A. Social Boundaries and the Formation and Persistence

of Moral Systems:

As mentioned earlier, normative similarity is a product of role
summation. And, it is recalled that summated exchange produces struc-
tural density in a social network. Once this happens, structural den-
sity has a "multiplier effect" on the formation of a uniform normative
culture. If the beliefs of any social aggregate are to be commonly
distributed, they must be communicated to those who will presumably
share them. If a particular social network is characterized by a high
level of structural density, it will be that much easier for the beliefs
to be distributed throughout the entire network. The information flow
covers a wider social domain more efficiently, thereby creating a more

uniform normative culture.

Social boundary mechanisms promote normative similarity by en-
couraging summation and density. Boundary mechanisms operate as "so-
cial facts” which create "total institutions" (Goffman, 1961). In
these situations individuals are often forced to interact with a limited
set of others. Isolated organizations like fishing crews, lumberjack
camps, construction crews in remote areas, combat units, and remote
villages have a high distribution of summation in their social struc-
tures. Other boundaries encourage summation and density not as a con-
sequence of ecological conditions but through social sanctions. Such
organizations as fraternities, religious sects, underground political

groups, mental health wards and secret cults develop density through the

activation of a rigid social boundary. Both ecologically and socially

29

determined total institutions force individuals to transact with a
limited set of others, leaving no other alternative but summated ex-
change and density formation.

Social boundaries maintain normative patterns by screening out
the "outsiders" who carry dissimilar cultural content. Peasant and
small scale rural communities are characterized by an inward orien-
tation toward familar relations and traditional culture. Anthropol-
ogists are well aware of the difficulties involved in gaining accep—
tance within such exclusive collectivities. But these boundaries
cannot be explained only in terms of external social facts because
the internal organization of transactions also plays a part.

If one is to account for the exclusivity and persistence of
traditional social structures in terms of internal processes, it is
necessary to return to the idea of an exchange network channeling
transactions through role relations of differing degrees of summation.

Consider the model represented below:

 

 

Figure 3

30

Let the solid lines between persons A and B represent a set of trans-
actions channeled through a summated relation. Let X represent a per—
son with whom A has just made an aquaintance, and who is unknown to B.
Furthermore, it is assumed that a substantial distribution of summation
and density characterize the total networks (not represented) of A and
B. Let us consider X as an "outsider" approaching a dense, summated
network.

If A wanted to exchange content k_with X, and to disengage this
transaction with B, what would be the consequences in terms of his re-
lation with B? His exchange relation with B would become less summated,
and consequently more Specific. This would expose their relation to
the potential hazzards of desummation. Since transactors encounter
problems when disengaging certain content from summated relations, an
outsider has difficulties in penetrating the exchange network because
this may involve a transfer of content.

Summated incumbents simply cannot spare very much additional time,
energy and resources required to enter new exchange relations with per-
sons unknown to their summated incumbent partners. Transactors can
develop them with friends of summated others because they are often in
the same situation. And if one did not have the necessary maintenance
resources for a particular summated relation, a legitimate excuse is
that they were allocated to a common friend. But such is not the case
with an unknown person. These factors make the cultivation of exchange
relations with "strangers" a choice with high costs.

We can now see why outsiders such as the social anthropologist
have a difficult time penetrating a summated network. There is a social

boundary based upon the character of summated exchange relations that

31

"welds" summated incumbents into their networks and keeps outsiders on
the periphery. But if the outsider happens to develop a summated re-
lation with one of the insiders, he will probably develop relations with
the other summated incumbents exchanging with that person. Outsiders
become insiders through the process of role summation.

B. Encapsulating Density and the Monitoring of

Self Interest:

What usually is referred to as gossip in folk language can be
thought of as communication which reinforces normative structure. Sum-
mated incumbents are great social gossipers due to the wide range of
permissible information exchange. Those who are a party to the gossip
clarify and reafirm normative culture amongst themselves. Summated
participants build solidarity in these private information exchanges.
As Gluckman noted:

The closer the bonds of a relationship, the more
intimate and esoteric the gossip, and the more
trite it will appear to outsiders. (Gluckman, 1963:
310).

These common evaluative agreements are, for my purposes, the most
important subset of transacted information content. Not only do the
gossiping persons solidify normative culture in relation to each other,
but in such situations extremely important bits of information are
transmitted through the network: the evaluation of a transactor's ex—
change performance. Gossip participants are in the process of evalu-
ating transactors either positively or negatively in terms of their per-
ceived congruency or discrepancy between the reported behavior and the

accepted rules of conducting exchanges.

Information content exchanged during gossip sessions provides an

32

opportunity for the incumbents of an actor's network to develop a
common stereotype of his exchange performance. The social source of
a person's self concept lies in his immediate network. Incumbents

of such a network entertain a common stereotype of the transactor and
exchange with him on the basis of this shared image. The higher the
density of a transactor's immediate network, the more control these
incumbents have over his self concept, and the greater are the pres-
sures for him to conform.

If all the incumbents an individual exchanges with tell him he
is "selfish" (maximizes too much) and that he should not be that way,
he will suffer great tension, and either leave his social area, or
curtail his self interest. But, if a few of these incumbents are
segregated from the rest, and some of these condone his exchange be-
havior, he will not fell such intense pressures to conform. The com-
raderie of deviants thwarts social pressures (Asch, 1955). If an
exposed self—maximizer cannot find a neutral social niche in the social
environment, and wants to have a high self concept he will conform to
the expectations of his social network.

Homans and Blau ignore the social processes monitoring self in-
terest by focusing their attention upon the diad exchange process. A
person must not over-maximize only because his actions will not be re-
warding to the other person. But exchange must be thought of as a
social process involving portions of a network. If a person over-max—
imizes his tarnished reputation spreads throughout the entire community.
Arensberg and Kimball find that these persons do not simply loose one

exchange relation, but they may have a "social death" and be condemned

to isolation.

33

The author's note:

Ordinarily, of course, failure to fulfill such

obligations either in directly social situations

or in agricultural cooperation does not lead to

violent conflict, but rather to an estrangement.

The "niggardly friend", especially the one who

fails to return obligation, is held up to general

condemnation, not only by his kindred but by the

whole community. (Arensberg 8 Kimball, 1949: 74).
Thus, by placing exchange behavior within a social structural context,
one sees not only utilitarian calculation, but a moral phenomenon. The
consequences of excessive self interest spread from one role to the en-
tire summated set connecting two persons; and from the confines of their
relation to the wider social network. In summated and dense social

networks exchange behavior is a "total social fact", while in non-sum-

mated, less dense networks, it is more utilitarian.

BIBLIOGRAPHY

34

BIBLIOGRAPHY

Arensberg, Conrad M.
1950 The Irish Countryman. New York: Macmillan Co.

 

Arensberg, Conrad M. and Solon T. Kimball.
1968 Family and Community i2_lreland. Cambridge: Harvard
University Press.

 

Asch, Solomon.
1955 "Opinions and Social Pressure." Scientific American 193 (5).

Banton, Michael P.
1965 Roles: A2_Introduction :2_the Study 2f_Social Relations.

 

 

Davis, A., B. Gardner and M. Gardner.
1941 Deep South. Chicago: The University of Chicago Press.

 

Durkheim, Emile.
1933 The Division gf_Labor. New York: The Free Press.

 

Firth, Raymond.
l939 Primitive Polynesian Economy. Hamden, Conn: Archon Books.

 

Gans, Herbert
1962 The Urban Villagers. New York: The Free Press.

 

Gluckman, Max.
1955b The Judicial Process Among the Barotse of Northern
Rhodesia. London: Manchester University Press.

 

1962 Essays 92_the Ritual gf_Social Relations. London: Man-
chester University Press.

 

 

Goffman, Erving.
1961 Encounters. Indianapolis, Indiana: Bobs-Merrill Co.

 

1961 Asylums. New York: Doubleday and Co.

Homans, George
1950 The Human Group. New York: Harcourt, Brace and World.

 

35

Levi-Strauss, Claude.
1949 The Elementary Structures 9f_Kinship. Glencoe, Illinois:
The Free Press.

 

Linton, Ralph.
1945 The Cultural Background 2£_Personality. New York: Appleton-
Century-Crofts, Inc.

 

 

Marx, Karl.
1888 The Manifesto 2f_the Communist Party. Bobbs-Merrill
reprint series #S-455.

 

 

Mead, George Herbert.
1956 92_Social Psychology. Chicago: University of Chicago Press.

 

Mitchell, Clyde.
1969 Social Networks iE_Urban Situations. London: Manchester
University Press.

 

 

Moore, Barrington.
1966 The Social Origins 2f_Dictatorshipand Democragy, Boston:
Beacon Press.

 

 

Paine, Robert.
"A Theory of Patronage and Brokerage," Forthcoming in
Patronage, Brokerage, and Clientship ig_the Arctic.
Memorial University of Newfoundland Press.

 

 

Parsons, Talcott.

1951 The Social System. New York: The Free Press.

 

Parsons, Talcott.
1965 In Parsons, Shils, Naegele, and Pitts (eds.). Theories gf_
SocieEy. New York: The Free Press.

Sahlins, Marshall D.
1965 "On the Sociology of Primitive Exchange." In The Relevance
of Models for Social Anthropology. A.S.A. Monograph #2.
London: TaViStock Publications.

 

 

Silverman, Sydel F.
1970 "Exploitation in Rural Central Italy: Structure and Ideology
in Stratification Study." Comparative Studies iE_Society
and History, II—

 

 

Sorokin, Pirtrim.
1929 Principles 9£_Rural—Urban Sociology. New York: Holt.

 

 

Srabin, T.
1953 "Role Theory". In Gardner Lindzesy. Handbook 2: Social
Psychology. Cambridge: Addison-Wesley.

 

36

Toennies, Ferdinand.
1957 Gemeingphaft und Gesellschaft. (Trans. by C. Loomis).
New York: Harper and Row.

 

Weber, Max.
1964 The Theory 9f_Social and Economic Organization. New York:
The Free Press.

 

 

HICHIGRN STATE UNIV. LIBRQRIES

I ll” IHI ||||||| \IIIH I

31293103499004