Content uploaded by Michael Robert Solomon
Author content
All content in this area was uploaded by Michael Robert Solomon on Nov 27, 2018
Content may be subject to copyright.
Michael
R.
Solomon, Carol Surprenant, John A. Czepiel,
&Evelyn G. Gutman
ARole Theory Perspective
on
Dyadic Interactions:
The Service Encounter
This article proposes that the dyadic interaction between a service provider and a customer is an im-
portant determinant of the customer's global satisfaction
with
the service. Based on role theory, a the-
oretical framework is presented which abstracts some of the critical components of service encounters
across industries.
R
ESEARCHERS interested in service marketing
are beginning to understand
what
they are study-
ing, but they are not yet clear
how
to study it. As
Bateson (1977) said, "The service marketing litera-
ture generally has been concerned with listing the dif-
ferences
between services and products. There has been
little attempt to point out the implications for mar-
keters in service companies and even less of an at-
tempt to propose new concepts or approaches" (p. 14).
Service marketing refers to the marketing of ac-
tivities and processes (health care, entertainment, air
travel) rather than objects (soap powder, cars). Rath-
mell (1966) made a similar, fundamental distinction
in defining goods as objects and services as deeds or
efforts. There are still considerable differences of
opinion within the marketing discipline as to whether
products and services are fundamentally distinct
(Bateson 1977; Judd 1964; Lovelock 1980; Uhl and
Upah 1983; Wyckham, Fitzroy, and Mandry 1975).
Michael
R.
Solomon
is
Associate
Director,
Institute
of
Retail
Manage-
ment,
and
Assistant
Professor
of
Marketing;
Carol
Surprenant
is
Assis-
tant
Professor
of
Marketing;
John
A.
Czepiel
is
Associate
Professor
of
Marketing;
and
Evelyn
G.
Gutman
is a
doctoral
student;
all
are
at
the
Graduate
School
of
Business
Administration,
New
York
University.
Journal
of
Marketing
Vol. 49 (Winter 1985), 99-111.
Most attempts to differentiate the two on one or more
dimensions ultimately arrive at a continuum (Bell 1981,
Liechty and Churchill 1979, Rathmell 1966); products
are arrayed at one end, services at another, and there
is considerable overlap between the two. This would
seem to indicate that while services marketing may
not be unique, a focus on the marketing problems pre-
dominantly present in this sector may enable us to
broaden our horizons and, in fact, contribute to im-
proved marketing concepts applicable to both goods
and
services.
One of the consequences of the recent interest in
service marketing is the increased recognition of the
importance of the person-to-person encounter between
buyer and
seller-client
and
provider-to
the overall
success of the marketing effort. Many service situa-
tions, especially those termed "pure" services, are
characterized by a high degree of person-to-person in-
teraction: consulting services, hairdressing, and med-
ical services, to name a few. Recognition of the im-
portance of the encounter is especially relevant in those
situations where the service component of the total
offering is a major element of that offering. This is
so regardless of whether the core element of the of-
fering is a material good or a service.
In mixed product/service offerings, the impor-
tance of the
encounter-the
person-to-person inter-
A
Role
Theory
Perspective
on
Dyadic
Interactions
/ 99
action between buyer and
seller-is
often overshad-
owed by a focus on the more tangible product attributes,
and the customer's more enduring interaction with the
product itself. In pure service situations where no tan-
gible object is exchanged, and the service quality it-
self is difficult to measure (financial planning, health
care), customer satisfaction and repeat patronage may
be determined solely by the quality of the personal
encounter. That is not to say that the encounter be-
tween an industrial salesperson and industrial pur-
chaser is not similar to personal service encounters; it
is. To the extent that the interaction with the sales-
person is an element in the total offering, the en-
counter is important and, in fact, constitutes a service
encounter.
However, it is primarily in the service sector that
the dyadic encounter! has generated a great deal
of
managerial concern. The ideas presented are partic-
ularly relevantfor people-based services (legal and other
professional services), rather than equipment-based
services (automatic teller machines, direct mail, in-
surance).
It
can, of course, be generalized to any mar-
keting situation in which personal interaction is an im-
portant element of the total offering. For our pwposes,
service encounter will be used to indicate face-to-face
interactions between a buyer and a seller in a service
setting.
Two quotations by service marketing managers re-
ported in Advertising
Age
(Knisely 1979) highlight the
managerial importance of service encounters.this way:
In a service business, you're dealing with something
that is primarily delivered by
people-to
people. Your
people are as much of your product in the consumer's
mind as any other attribute of that service. People's
performance day in and day out fluctuates up and
down. Therefore, the level of consistency that you
can count on and
try
to communicate to the consumer
is not a certain thing.
The real intangible is the human element which, with
the best will in the world, most of us cannot control
to anywhere near the same degree that a product
manager controls the formulation of a beauty soap,
for example (pp. 47-51).
Thus, the service encounter itself occupies a central
place in much of service marketing.
It
impacts on ser-
vice differentiation, quality control, delivery systems,
and customer satisfaction. The centrality of this ser-
vice component warrants a substantial theoretical
focus.
To an extent, all service personnel involved in
customer contact are marketers. Each individual rep-
resents the firm, defines the product, and promotes it
directly to the consumer (Shostack 1977). Given its
'Not all service encounters are simple dyadic relationships. Some
may involve a series of dyadic interactions, while others are still more
complex and involve a number of different actors.
100 /
Journal
of
Marketing,
Winter
1985
centrality to the service offering and its variable na-
ture, the service encounter is a worthy
if
difficult topic
to study. This paper presents a conceptual framework
by which person-to-person service encounters, whether
delivered in conjunction with a tangible product or not,
can be understood and analyzed.
The conceptual framework presented is adapted
from a social psychological perspective on human in-
teraction. The service encounter is approached as a
special case of the more general class of goal-oriented
dyadic interactions. Any encounter is assumed to con-
tain learned and consistent behavior patterns; each
participant should enact certain behaviors in order for
the transaction to proceed smoothly.
It
will be argued
that the degree of congruence with this learned pattern
or "script" by both the service provider and customer
is an important determinant of satisfaction with the
encounter. Thus, the focus is on the interdependence
of both individuals. Each depends upon the other to
make the interaction run smoothly (cf. Kelley and
Thibaut 1978).
Relevant Perspectives
Service Encounters Are Dyadic
The transaction or exchange is a cornerstone of any
marketing activity, yet few researchers have adopted
it as the basic unit of analysis. As noted by Penning-
ton (1968), one party to the transaction tends to be
studied in isolation. The traditional approach to the
salesperson-customer exchange, for example, is to view
sales success as determined by the seller's job satis-
faction (Cotham 1968), motivation (Oliver 1974), or
ability (Ghiselli 1978). Some workers in this area have
questioned the simplistic nature of a model based on
only one participant, and have recognized the dyadic
quality of personal selling (Sheth 1975, Webster 1968,
Weitz 1981). This belated recognition was presaged
20 years ago by Evans, who maintained that "The sale
is a social situation involving two persons. The inter-
action of the two persons, in tum, depends upon the
economic, social, and personal characteristics of each
of them. To understand the process, however, it is
necessary to look at both parts of the sale as a dyad,
not individually" (Evans 1963, p. 76).
The interactional emphasis employed here high-
lights the overlooked importance of the service en-
counter as a psychological phenomenon that exerts a
major impact upon outcomes. This is consistent with
the approach of Lutz and Kakkar (1976) who have
recognized the importance of the psychological situ-
ation and the adoption of a process model to under-
stand behavior. They propose a model incorporating
situation, decision processes, and social exchange.
Similarly, we propose that the fusion of two people
in a service setting is greater than the sum of its parts.
Given such a conceptual framework, it will be pos-
sible to design research by which the elements of sat-
isfying and nonsatisfying encounters can be identi-
fied. It is proposed that these elements will overlap
with dimensions of human social interaction that have
been shown to exert influence in other classes of goal-
oriented behavior.
This approach is influenced by work on the dy-
namics of both face-to-face encounters and group ac-
tivity.
It
stresses the mutuality of behaviors (cf. Thi-
baut and Kelley 1959) and acknowledges that a service
encounter is a form of social exchange in which par-
ticipants normally seek to maximize the rewards and
minimize the costs of the transaction (cf. Homans
1961). It is also assumed that it is at some point fea-
sible and desirable to measure units of behavior, and
to assess their contribution to the quality of eventual
outcomes (cf. Bales 1950).
The study of the service encounter is also influ-
enced by prior theoretical developments on dyadic in-
teractions in the marketplace. These perspectives have
usually centered upon the personal selling process
(Evans
1963, Sheth 1975, Willett and Pennington 1966,
Wilson 1977). For example, Sheth makes a relevant
distinction between two interaction dimensions: the
content versus the style of communication. The latter
dimension recognizes the centrality of ritualistic be-
havior patterns in shaping the outcome of the buyer /
seller interaction.
The ability to identify mutually satisfying factors
in encounters will be helpful in the design of services,
in the setting of service level standards, in the design
of service environments, in the selection, training, and
motivation of service providers, and in guiding cus-
tomer behaviors. This approach suggests that the
manager look to find ways to channel both provider
and customer behavior if satisfaction with the en-
counter is to be maximized.
Service Encounters
Are
Human Interactions
At a superficial level the acts of ordering a meal, ob-
taining a car loan, making plane reservations, or pick-
ing up a suit at the dry cleaners appear to have little
in common. At a social psychological level, however,
all of these incidents are conceptually similar: Each
act is a purposive transaction whose outcome is de-
pendent upon the coordinated actions of both partic-
ipants. As is the case in many types of dyadic inter-
actions, one cannot predict the quality of outcomes
with knowledge of only one actor's behavior. Instead,
much of social behavior consists of joint
activity-a
major task for the interacting person is the mutual co-
ordination of appropriate behavior vis-a-vis the other
person (Thibaut and Kelley 1959).
Communication between a service provider and a
customer is interactive; it is a reciprocal process rather
than a linear one. The service experience which dis-
tinguishes one service organization from another is a
result of the unique interaction between the experi-
encer and the contact person (Booms and Nyquist
1981). Since the success of a particular service vendor
rests on the quality of the subjective experience, the
nature of this experience is the critical determinant
of
long run market success. Facilities may be spotless
and the service delivered on time as
ordered-but
if
the customer leaves with a negative impression from
the attitude of an employee, other efforts may be
overlooked.
Although providers often behave as if they act on
a static consumer, it is imperative to understand the
consumer's participatory role in assigning meaning to
marketing stimuli (Booms and Nyquist 1981). Indeed,
it is more accurate to think of the service provider as
acting with the customer. While marketers often em-
phasize short run indices of seller effectiveness, this
perspective may be especially myopic in the service
sector (Czepiel 1980, Schneider 1980). Schneider cites
three examples of this myopia: Bank tellers' evalua-
tions depend more on how they "prove out" at the end
of the day than on the courtesy they display, airline
reservation clerks are judged more on paperwork
errors than on the goodwill they generate, and the
short-term dollar volume generated by insurance
salespeople is weighed more heavily than is success
in establishing long-term interpersonal relationships.
While indices such as accuracy and sales are clearly
important, the point here is that other criteria relating
to personal service should also be included in evalu-
ating service personnel. To reiterate, the quality of the
subjective
product-the
service
experience-is
the true
outcome of a service interaction. This product is man-
ufactured by both parties and must be approached as
such. To paraphrase an old Zen saying, we know the
sound of two hands clapping; what is the sound of one
hand clapping?
Service Encounters
Are
Role Performances
Adistinguishing feature of service encounters as a class
of human interaction is the purposive, task oriented
nature of the interaction. Specific short-term goals are
clearly defined and agreed upon by society (procuring
airline tickets, depositing a check). Due to this con-
sensus, ritualized behavior patterns evolve which gov-
ern the course of the encounter. Each party to the
transaction has learned (albeit with differing degrees
of facility) a set of behaviors that are appropriate for
the situation and will increase the probability of goal
attainment. Each participant has a role to play; the
script from which
he/she
reads is often strictly de-
fined.
This socially-defined structure renders provider/
A
Role
Theory
Perspective
on
Dyadic
Interactions
/ 101
client interchanges especially amenable to a role the-
oretical analysis of the service encounter. Role the-
ory, of course, is not new to marketing. Constructs
adapted from role theory have been used to explain
consumer behavior, especially with regard to expec-
tationformation (Sheth 1967). The major areas in which
the theory has been applied are in personal selling,
role portrayals in advertising, and husband/wife de-
cision making (see Wilson and Bozinoff 1980 for a
comprehensive review of role theory in marketing).
A role theoretic approach emphasizes the nature
of people as social actors who learn behaviors appro-
priate to the positions they occupy in society. Al-
though the "actors" in a service setting may be very
different individuals in their leisure time, they must
adopt a relatively standardized set of behaviors (i.e.,
read from a common script) when they come to work
or enter the marketplace. In fact, people are often de-
fined by the service roles they play. When an indi-
vidual is labelled nurse, clerk, or cab driver, one is
able to generate a profile of this person based on the
characteristics which are believed to covary with this
title. The pervasive tendency to "fill out" one's
knowledge of a person, given observation of reli-
gious, political, or occupational characteristics, is well-
documented in the literature on person perception and
"implicit personality theory" (cf. Tagiuri 1969).
This implicit structure is not confined to the ser-
vice provider. The recipient of the service also plays
a role. The customer/client role is composed of a set
of learned behaviors, a repertoire of roles; the partic-
ular script which is read depends upon the demands
of the specific service environment and other situa-
tional cues (Lutz and Kakkar 1976). The customer role
in an elegant restaurant involves very different actions
than an appropriate role in a fast-food setting.
It
will
be argued at a later point that the root cause of many
provider/client interface problems is the failure of
participants to read from a common script. First, it is
necessary to briefly introduce some of the basic con-
cepts of role theory as they are relevant to an analysis
of the dyadic service encounter.
An Overview of Role Theory
Role theory is based on a dramaturgical metaphor. The
study of a
role-a
cluster of social cues that guide
and direct an individual's behavior in a given set-
ting-is
the study of the conduct associated with cer-
tain socially defined positions rather than of the par-
ticular individuals who occupy these positions. It is
the study of the degree to which a particular part is
acted appropriately (role enactment) as determined by
the reactions of fellow actors and observers (the au-
dience). Since one aim in the intangible service en-
vironment is to provide consistent service at an ac-
102 /
Journal
of
Marketing,
Winter
1985
ceptable level across individual service providers, this
perspective seems particularly fitting (Grove and Fisk
1983).
Each role that one plays is learned. One's confi-
dence that one is doing the right thing leads to sat-
isfaction with a performance (termed role validation)
and success in interacting with others who are, of
course, also playing their respective roles. One's role
specific self-concept is formed by the reactions of oth-
ers to the quality of one's role enactment. Some roles
are more central to the individual than others (e.g.,
Lover versus Golfer). The self also can be thought
of
as a system of identities to which one is more or less
committed (Jackson 1981), where commitment to a
role implies a concern that one's role enactment be
convincing (Sarbin and Allen 1968). Self-concept re-
lated to a role constitutes a role identity (McCall and
Simmons 1966).
The concern that one be competent in performing
a role is applicable in a service setting.
If
the service
provider's occupational role is relatively salient in the
self-concept, commitment to the effective perfor-
mance of that role should be strong; giving good ser-
vice will matter. In addition to the basic motivation
to perform competently (White 1959), desire to per-
form a service role well should also be a function of
group (i.e., company) cohesiveness. Service person-
nel are members of an organization; group member-
ship brings with it a responsibility to act in accord
with prescriptions that define one's position in the group
(McCall and Simmons 1966), especially when this
position is valued by the group member. For both of
these reasons, the issue of morale as it impacts upon
service can be viewed as a question of commitment
to a role identity.
For both provider and consumer, the successful
enactment of even the most basic service scenario in-
volves the mastery of a wide range of behaviors. On
the provider side, this learning process is often ex-
plicit.
It
may take the form of a training program or
perhaps an apprenticeship to an accomplished role
player. On the other hand, the consumer's burden is
at times explicit (e.g., the elaborate directions for fill-
ing out forms and reporting to various offices when
dealing with bureaucracies such as the Department of
Motor Vehicles), but it is more typically implicit. Re-
gardless, there are always behaviors which may come
automatically to a veteran but which demand great
cognitive activity by the novice. Anyone who recalls
the adolescent apprehension over "doing the right
things" on one's first solo outing to an expensive res-
taurant can attest to this (how to talk to the waiter,
what to order when the menu is written in another
language, how much to tip). The heuristic value of
this approach is probably maximized in situations in-
volving the execution of well-learned behaviors that
possess a high degree of social consensus as to ap-
propriate and expected actions. Many routine service
transactions fall in this category.
In the case of a person's encounter with either a
novel situation or one perceived as warranting active
problem solving (i.e., a high involvement situation),
this perspective may still be useful. Though its ex-
planatory power at the molecular level of behavior may
be diminished, role theory can still be employed to
specify molar orientations.
It
seems likely that the oc-
cupant of a novel role who has not yet developed a
script specific to that role (promotion to a new job,
one's first experience in dealing with an interior dec-
orator or a lawyer) will attempt to employ an existing
script similarly structured. Alternatively,
he/she
will
use an idealized script that has been internalized through
vicarious socialization. Thus, one may assign the role
of Lawyer to the larger category of Authority Figure,
and act
deferentially-much
as one would act with
other known representatives of this category, such as
Physician or Teacher. Or one might approximate be-
havior based upon expectations derived from TV,
movies, or books regarding how lawyers and their
clients are supposed to act.
While the gaps in a novel script are filled in by
accumulated experience, reference to an idealized script
at early stages delimits behavioral options by estab-
lishing the parameters of possible activity; a subset is
created from one's total behavioral repertoire. Vague
role knowledge may not provide specific guidance for
appropriate role behavior, but it may go a long way
in advising one what not to do or say. For example,
a person may not know exactly what to talk about
with a member of the clergy, but one is probably aware
of some subjects which should not be discussed.
Role Expectations
Role expectations are comprised of the privileges, du-
ties, and obligations of any occupant of a social po-
sition (Sarbin and Allen 1968). These expected be-
haviors must always be defined in relation to those
occupying the other positions in the social structure.
It is important to remember that a role player's be-
havior is interdependent with the behavior of those in
complementary positions. One's role conduct must take
into account the role behavior of others. The totality
of complementary
roles-to
a bank teller, for exam-
ple, there may be customers, co-workers, head teller,
and branch
manager-is
a role set (Merton 1957).
One important result of proper role socialization
is the acquired ability to predict the behavior of other
role players. In role theory terms, this is known as
"taking the role of the other" (Mead 1935). This em-
pathic process whereby the actor anticipates the oth-
er's expected role behavior allows the actor to gauge
his/her own behavior to the predicted behavior of oth-
ers (Rose 1962). Research in personal selling has
demonstrated that the salesperson whose behavior is
contingent upon the behavior of the customer is more
effective than one who does not adjust behavior to
meet the customer's speciftc needs (Weitz 1981).
These predictions are based on expectations for
behavior implied in common meanings. For example,
a customer who walks into a clothing store is com-
municating consideration of a buying transaction or at
least wants to browse. This behavior allows the sales-
person to initiate the actions which correspond to a
sales role. The salesperson's approach would not have
the same meaning outside the store as inside, though
in both cases a stranger is initiating conversation and
perhaps asking questions
of
a somewhat personal na-
ture. Once the shopper enters the store,
he/she
adopts
the role of customer and a role-defined dyadic inter-
action familiar to both parties may begin.
While consumers and providers have common ex-
pectations about appropriate role behaviors, these ex-
pectations differ among encounters and are moderated
by provider/consumer characteristics and percep-
tions, and by production realities (Czepiel et al. 1982).
Provider/consumer characteristics and perceptions
about the encounter dictate which behaviors comprise
a satisfactory interaction and can serve to differentiate
offerings of the same type of service. For example,
within a class of service encounters (e.g., buying
clothing) the behaviors of the role players will differ
as a function of the learned expectations within that
speciftc setting; different behaviors are expected in J.
C. Penney stores compared to Brooks Brothers stores.
Production realities refer to the set of dimensions as-
sociated with production characteristics such as time
factors, technology, location, content, and complex-
ity, that constrain the encounter and help determine
which role behaviors are appropriate.
The expectations held by each role player about
appropriate behaviors are multidimensional. The con-
cept of "bridging the gap
...
between the producers
and the consumers with respect to values, percep-
tions, possession, time, and place dimensions of ex-
change" [transaction] (Sheth 1982) is relevant here.
If the salient dimensions
of
the encounter are clear,
appropriate role behaviors can be identified and eval-
uated (Czepiel et al. 1982).
Role Expectations Affect Performance
The concept of expectations is not new to marketing.
The consumer satisfaction literature deftnes product
satisfaction as a function of consumer expectations and
perceived performances. The greater the negative dis-
crepancy between expectations and performance, the
greater the corresponding dissatisfaction (Churchill and
Surprenant 1982; Czepiel, Rosenberg, and Akerele
1975; Swan and Coombs 1976). The view adopted
A
Role
Theory
Perspective
on
Dyadic
Interactions
/ 103
here is similar: Satisfaction with a service encounter
is seen as a function of the congruence between per-
ceived behavior and the behavior expected by role
players.
It should be noted that this relationship has re-
ceived empirical support in some studies on personal
selling. In a study of differences between sold and
unsold prospects for insurance, Riordan, Oliver, and
Donnelly (1977) found that role
congruence-the
ab-
solute difference between acustomer's perceptions of
actual and ideal insurance
agents-emerged
as a par-
simonious discriminator between those who pur-
chased a policy and those who did not. An earlier study
which used a somewhat similar population found that
successful agents fulfilled expectations concerning
similarity, expertise, friendliness, and personal inter-
est (Evans 1963). In addition, a study of interactions
between wholesale drug salespeople and retail phar-
macists showed that the degree to which seller be-
havior differed from buyers' role expectations of a drug
salesperson was related to degree of supplier loyalty
(Tosi 1966).
Aconsideration of role congruence in a service
setting is actually a two-dimensional issue of intra-
role and inter-role congruence. Intra-role congruence
reflects the degree to which the service provider's
conception of
his/her
own role is concordant with the
organization's conception of that role." Inter-role con-
gruence is the degree to which provider and client share
a common definition of service roles.
Intra-role congruence. The first part of the issue
has been addressed by some workers in the area of
personal selling. In this domain it comes under the
rubric of role ambiguity. A lack of role clarity has
been shown to be a major source of
job
tension, dis-
satisfaction, and reduced innovativeness as workers
are unsure of the exact nature of role expectations (cf.
Kahn et al. 1964). This factor has been demonstrated
to mediate satisfaction for industrial salespeople (Ford,
Walker, and Churchill 1976), managerial personnel
(Oliver and Brief 1977-78), and retail salespeople
(Dubinsky and Mattson 1979), as well as committee
members (Bible and Brown 1963) and teachers (Bible
and McComas 1963).
2An additional variant of this issue centers on the problems that arise
when the role conceptions of co-workers do not overlap with those
of management. For example, the informal peer group may be highly
cohesive yet devote the bulk of its attention to social interaction rather
than goal performance (Davis 1969). This situation is exemplified by
piece-workoperations where overzealous workers who exceed the quota
are branded as rate-busters and ostracized by peers. While such dis-
parities in role concepts possess important implications for productiv-
ity and morale issues, further delineation of what constitutes the or-
ganization's conception of a role is a complex matter and beyond the
scope of this article. For our purposes management's role-definition
is assumed to be dominant.
104 /
Journal
of
Marketing,
Winter
1985
Role congruence is partly determined by disposi-
tional characteristics (i.e., some people are just not
suited to certain roles) and by past experience and
amount of interaction in that role (Sarbin and Allen
1968). This may explain why there is less incongru-
ence found in friend roles than in occupational roles
(Block 1952).3
The amount of overt communication about role
expectations is obviously an important mediator (e.g.,
feedback from sales managers to floor personnel re-
garding criteria for advancement, or the formulation
of explicit criteria in academe for promotion and ten-
ure). Since such communication flows are facilitated
in cohesive groups, it is not surprising that greater role
consensus is found in small organizations (Thomas
1959). Schneider (1980) has proposed that incongru-
ence between the service orientation of employees who
are probably self-selected to be "service enthusiasts,"
and the perceived orientation of management as "ser-
vice bureaucrats" who care only about maintaining the
system, engenders role ambiguity and conflict. This
process in turn translates into dissatisfaction, frustra-
tion, and intentions to quit.
Inter-role congruence. The second type of role
congruence is equally crucial: the degree of agreement
between both parties involved in the service transac-
tion regarding the appropriate roles to be played. A
lack of clarity is likely to influence the efficacy of
group or dyadic performance. At the least, the neces-
sity of expending effort to predict an individual's be-
havior (which is obviated by congruent role enact-
ment) decreases the time available to expend on task
activities. Early group dynamics studies demonstrated
this quite clearly. For example, the existence of an
unclear group structure was shown to impede the abil-
ity to survive of Air Force crews under stress (Tor-
rance 1954). In another study, confederates under in-
structions to remain silent in a group problem solving
session decreased productivity, as other members were
hampered by ambiguous role expectations. These det-
rimental effects were eliminated if the confederate was
identifiedas a listener at the outset of the session (Smith
1957). In other words, group effectiveness depends
upon each member understanding the role expecta-
tions of the other members so that each is clear about
his/her own role expectations.
It
seems likely that accurate mutual comprehen-
sion of role expectations is a prerequisite for a satis-
fying service experience. This joint assignation of roles
probably occurs during the initial encounter and per-
sists throughout subsequent encounters in the service
'The amount of experience in playing friendship roles which leads
to greater role congruity may be confounded by the large degree of
self-selection exerted in such roles relative to other roles.
environment. As in everyday person perception, the
first impression is a pervasive one. The service cus-
tomer seeks to reduce risk by looking for tangible signs
of capability to deliver the service. As a result, the
first time customer will be especially vigilant as
he/
she assimilates such environmental clues as the ap-
pearance and demeanor of the service provider (Booms
and Nyquist 1981).
This initial labelling process can be thought of as
role assignment or "altercasting" (Weinstein and
Deutschberger 1963). The outcome of this process
drastically affects the subsequent tone and content of
the interaction. As one partner identifies a salient role
(e.g., a friend prefaces advice with "as your lawyer
...
"), the complementary role of the other partner
is simultaneously defined. A pompous suit salesper-
son calls forth a different customer role than the ob-
sequious haberdasher who is eager to please. For in-
stance, it seems likely that the customer will be more
assertive about any idiosyncratic preferences in style
or tailoring in the second case than in the first, where
he/she may be more intimidated
and/or
submissive.
It
is important to note that the long-term effectiveness
of each type depends upon the customer's expecta-
tions. To paraphrase a hotel chain's motto, there will
be no surprises.
Role discrepancies. Problems arise when there is
a discrepancy somewhere in the system. This incon-
sistency with expectations may be exhibited in one of
two ways: (1) the employee's perception of job duties
or qualifications differs from the customer's expec-
tations of those duties, or (2) the customer's concep-
tion of the customer role differs from the employee's
notion of that role.
This proposition can be illustrated by considering
either of two extant taxonomies of marketing inter-
actions. Both McMurry's (1961) classic continuum of
personal selling and the breakdown in terms of em-
ployees' communication functions by Booms and
Nyquist (1981) seem to share the recognition that the
role requirements of employees can range from those
of a virtual automaton to those of an equal partner vis-
a-vis the customer. 4
Consider the situation where the employee's role
concept is at odds with that projected by the customer.
An independent clothes shopper may resent the intru-
sion caused by the unwanted advice of a clothing
salesperson who is regarded as a mere order taker. A
counter clerk at McDonald's may not be prepared to
"Mclvlurry's work is centered on personal selling, while Booms and
Nyquist work in the area of service marketing. Though beyond the
scope of this paper, it is interesting to speculate on where one area
ends and the other begins. As an illustration, both approaches begin
with order takers and end with positions involving the creative seIling
of intangibles (insurance salesperson).
make menu recommendations, or a waiter may brush
off a patron with "not my station." A patient may
resent an overly familiar manner in a doctor, and a
doctor in tum may bristle at the patient who performs
self-diagnosis.
In these examples it is clear that the role players
are not reading from the same script. If the structure
of service scripts is better understood, the transaction
can be engineered for congruency, and there is a greater
probability that a climate for service (Schneider 1980)
will prevail.
The Service Script
A more precise analysis
of
the service script can per-
haps be aided by recent developments in cognitive
psychology. Although this area is in some ways far
removed from the theoretical sociology of role theory,
some cognitive psychologists also (perhaps coinci-
dentally) rely on the heuristic value of the dramaturg-
ical metaphor. Investigations of schematic informa-
tion processing contain the assumption that much of
social interaction is governed by learned assumptions
regarding the course the interaction should take. The
interface of psychology and artificial intelligence re-
search has produced work on a priori plans or scripts.
Abelson (1976) defines a script in this context as "a
coherent sequence of events expected by the individ-
ual, involving him either as a participant or as an ob-
server" (p. 33). Despite the differences in intellectual
origin, this definition is quite compatible with the above
discussion of role expectations.
If
a script is thought
of
as a learned sequence
of
causal chains (cf. Schank 1980), it seems plausible to
make the leap to service scripts (Smith and Houston
1983). This script would contain information about
the role
set-one's
own expected
behavior-plus
the
expected complementary behavior of others, and would
reflect the individual's learned (or imagined) concep-
tion of the prototypical service experience. Informa-
tion about a service encounter would be stored in dif-
ferent levels of memory" as a function of its degree
of abstraction (Schank 1980). An illustration is pro-
vided by Schank's example
of
a "dentist script." Re-
membrances of specific visits to a dentist would be
stored in Event Memory. Unless some truly unique
event occurred ("I got a date with the receptionist"),
this recollection would gradually be incorporated into
Generalized Event Memory, a collection
of
events
whose common features have been abstracted.
'The notion of different levels of memory is no longer considered
accurate. More properly, one would speak of differences in activation.
However, as Bettman (1979) notes, a liberal interpretation of Craik
and Lockhart (1972) can encompass an activation model and, using
the terminology already in place, avoids the problem of defining new
terms at length.
A
Role
Theory
Perspective
on
Dyadic
Interactions
/ 105
Knowledge like "going to a health professional's of-
fice," which is information about specific situations
in general, would be stored in Situational Memory.
Finally, goal-based
information-a
major component
of a service
encounter-would
be stored in Intention
Memory. This encoding process is depicted schemat-
ically in Table 1.
It
should be kept in mind that the
dentist also possesses scripts corresponding to patient
types (new patient, anxious patient).
Consumers can be thought of as possessing cog-
nitive scripts for a wide variety
of
service encounters.
Although a high degree of consensus can be expected
across people regarding script components, a process-
oriented approach must acknowledge the fluid nature
of such a construct. A variety of variables will me-
diate the idiographic content of scripts. Most cultural
expectations are for ranges of behavior rather than for
specific microbehaviors. Some mandate variation rather
than conformity, as with scientists or designers (Rose
1962). Expectations may change over time as a script
becomes redefined, though acceptance of a new service
script is probably facilitated by integration with the
old one. As an example, the traditional gas station
script included having one's car windows wiped. The
revised version frequently no longer includes this act
but retains other elements. A related example is con-
sumer resistance to such new forms of transaction as
the Universal Product Code, bank machines, and self-
service gas stations (Lovelock and Young 1979). These
changes involve the sudden learning of radical script
changes.
Service
Transaction-Mindless
Behavior?
For the most part, routine service encounters take place
in an almost automatic style with a minimum of cog-
nitive activity. As an illustration, it seems likely that
most people cannot accurately recall details of all the
service interactions they experience in the course of a
day (what a salesperson looked like, what happened
when one bought a pack of gum). Like actors in a
long running play, people in familiar situations often
interact by rote with little conscious attention at the
time and even less recall later. This tendency has been
termed mindlessness; a person interacts with the en-
vironment in a passive fashion with a minimum of
cognitive activity (Langer 1978). Research supports
the idea that adults tend to spend a significant portion
of time in a mindless state unless they are provoked
into mindfulness (Langer, Blank, and Chanowitz 1978;
Langer and Imber 1979). This provocation usually takes
the form of an unfamiliar situation or perhaps embar-
rassment, a jolt back to reality. As long as the struc-
ture of a communication is familiar, regardless of its
content, mindlessness appears to be the norm.
As long as the structure of a service script is fol-
lowed, it may be deduced that the encounter is char-
acterized by mindlessness. Any experience in a re-
petitive job, where one seems to lose time perspective
and operate in a partial trance, will serve to illustrate
this proposition.
It
can be argued that it is only when
the experience somehow deviates from the service script
that the participants are individuated and the situation
takes on an affective valence. Suddenly, one must ex-
pend cognitive effort to orient behavior as the pre-
dictability of the role enactment is diminished. The
result of this deviation may be either positive or neg-
ative.
It
is proposed that the jolt from mindlessness
puts the customer in an evaluative set. Without this
evaluation may not occur at all. Swan and Trawick
(1978) found that for low involvement, frequently
TABLE
1
An Artificial Intelligence Approach
to
the Structure of a Dentist Scripta
INTENTION
MEMORY
HEALTH
PROBLEM
FIND
PROFESSIONAL
+MAKE CONTACT +
PROFOFFICEVISIT
SITUATIONAL MEMORY GO TO
OFFICE
+WAITING ROOM +
ENTER
OFFICE
+
HELP
+
LEAVE
+
BILLSENT
GENERALIZED
EVENT MEMORY
EVENT
MEMORY
"Adapted from Schank 1980, p. 263.
106 /
Journal
of
Marketing,
Winter
1985
Dentist visits include:
-getting
teeth
cleaned-dentist
puts funny tooth paste on teeth, turns on
machine, etc.
-getting
teeth
drilled-D
does x-ray, D gives shot of novocaine, D drills,
etc.
also: Dentists fill the health care professional role in
HEALTHCAREVISIT.
The time I went to the dentist last week:
-I
drove to the dentist.
-I
read Newsweek. There were holes in all the pictures.
-I
entered.
-He/she
cleaned my teeth.
-He/she
poked me in the eye with the drill.
-I
yelled.
-The
dentist didn't charge me.
purchased goods, more than 50% of respondents did
not recall forming any opinion at all about the prod-
uct. There was no deviation from expectations, thus
no reason to engage in the cognitive effort necessary
to form an evaluation. When some deviation occurs,
the customer attempts to discover the reason for the
deviation. Sometimes the customer is pleasantly sur-
prised (a bus driver may be unexpectedly courteous).
Other events may be experienced negatively ("that
woman didn't have to snarl at me when I asked her
to wrap my package"). In general, it is proposed that
extremes in evaluation (whether positive or negative)
will only be experienced when some departure from
expected role behavior is encountered and the abrupt
cessation of mindless behavior necessitates active pro-
cessing.
Predictability and Personalization
of Service
Service marketers often find themselves on the horns
of a dilemma: How to provide efficient, standardized
service at some acceptable quality level, while treat-
ing each customer as a unique person? Paradoxically
it may be argued that the customer often faces a sim-
ilar conflict. There is a trade-off between the gain in
personalization when one is treated as an individual
and the loss in predictability as the guidance provided
by role expectationsdissolves. The circumstances which
may give rise to both instances will now be briefly
considered.
Negative Discrepancies
Under what conditions will a disruption of scripted
behavior result in a subjectively negative experience?
An answer may be found when a key dimension of
role behavior is
considered-level
of involvement.
Sarbin and Allen (1968) identify eight levels of role
involvement along a continuum of self-role differen-
tiation. At the low end is noninvolvement (a lapsed
club membership) and casual role enactment (a cus-
tomer in a supermarket). The level of ritual acting fol-
lows. This is a stage of relatively mechanical behav-
ior, where the need to maintain behavioral consistency
requires some involvement of the self. Examples in-
clude the waitress who puts on a big smile and the
bank teller who inquires, "Is it hot enough today for
you?" Engrossed ("heated") acting is the next level.
The continuum ascends all the way to ecstasy and be-
witchment, which might characterize situations where
a consumer becomes totally engrossed in a product.
Such extreme involvement may be found during the
consumption of aesthetic experiences imparted by art,
music, etc. (cf. Holbrook 1980; Levy, Czepiel, and
Rook 1980).
It
seems likely, however, that most
(though certainly not all) service encounters are rooted
in the lower regions of involvement.
At the low end, minimal visceral participation is
involved.
It
seems likely that a high premium will be
placed upon efficiency and predictability. As anyone
who has ever been frustrated by the blundering of an
inexperienced fast-food employee knows, disruption
of the routine, leading to slower service, usually re-
sults in a negative experience. Other workers have noted
the tendency of clients to become aggressive toward
a contact person when dissatisfied (Eiglier and Lan-
geard 1977).
During low end service encounters, the treatment
of customers as individuals and not as role occupants
will probably not pay. Despite occasional promises to
"have it your way," consistency and speed are the
scripted attributes which are important and desired by
the customer.
Positive Discrepancies
In contrast, the personal touch is desirable in other
service situations. The assembly line nature of some
medical clinics is seen as a major drawback, and one
would certainly hope that a hair stylist would not use
a prepackaged mold to cut one's hair. These represent
situations of higher ego involvement; their greater
centrality to the self results in the high intensity role
enactment described as engrossment (Goffman 1961).
In such situations we hope that the service provider
treats us as a person instead of a number.
More Isn't Always Better
It may be postulated that satisfaction is positively re-
lated to predictability for low involvement services,
and positively related to flexibility/personalization in
the case of high involvement services. The point here
is that greater personalization
of
services does not
necessarily result in a more positive service experi-
ence. Instead the subjective outcome depends upon
the unique demands of the situation. This differential
also functions within a class of services. While we
expect attentive service at a high priced department
store, such attention would seem incongrous in a bar-
gain basement. In a similar vein, an early study of the
waitress-diner dyad showed that the optimal relation-
ship varied with the standing of the restaurant (Whyte
1948). In better eating places, waitresses suppressed
the desire to talk back to customers, and the formality
of their behavior was positively related to the per-
ceived status of the diner. On the other hand, in lower
standard restaurants it was the norm for a waitress to
put customers in their place. Waitresses who con-
formed to this norm actually received larger tips than
those waitresses who acted a middle-class script and
were respectful to their customers. Apparently con-
formity to the role expectations of the consumer is
A
Role
Theory
Perspective
on
Dyadic
Interactions
/ 107
rewarding, even if following the script results in ob-
jectively less desirable treatment. The crucial element
for improving routinized service transactions may be
to give the customer what
he/she
expects, with no
surprises.
Implications for Service Marketing
Viewing service transactions from a role theory per-
spective has a number of advantages. Role theory
compels us to adopt an interactive approach since roles
are defined in a social context. Furthermore, appro-
priate role enactment is determined by the reactions
of others. The quasi-ritualized nature of role behavior
makes it possible to examine the structure and content
of interacting roles apart from the specific individuals
occupying the roles. Thus individual difference vari-
ables are seen as moderating factors rather than as de-
terminants of behavior.
The concept of role expectations and predictability
is an especially powerful one for understanding the
nature of the service transaction. These expectations
form the basis for service scripts. Using these con-
cepts, deviations from scripted behavior can be ex-
amined for both positive and negative consequences.
Deviations may occur because one of the parties to
the interaction steps out of role, the participants do
not share common role definitions, or because the ac-
tors are not reading from a common script. Whatever
the reason, behavior that is unexpected reduces mind-
lessness and mandates increased cognitive activity,
which results in a closer scrutiny of the service situ-
ation-for
better or for worse.
Propositions
Using the concepts developed above, it is possible to
derive a set of propositions which can be used to ex-
amine service encounters.
PI: Service encounters can be characterized
as role performances.
The structure of the encounter is so-
cially defined with associated meanings
that guide and direct the behavior of the
interactants.
P2: Role behavior is ritualized, learned be-
havior.
a. The content
of
roles is relatively con-
sistent across actors.
This implies that a high degree of
consensus should exist, across individu-
als, regarding the content of the roles.
It
should be possible then to discover the
content of service provider roles, i.e., role
definitions, and to extract the key ele-
ments.
108 /
Journal
of
Marketing,
Winter
1985
b.Facility in role performance is a func-
tion
of
experience and communica-
tion.
From this proposition we would ex-
pect that novices in a service encounter
would expend more cognitive effort than
experienced role players. From an orga-
nizational point of view, we would also
expect communication about role expec-
tations to facilitate the learning of role
behaviors and to mediate experience.
c. Service scripts, containing informa-
tion about the role set, are learned by
both service providers and customers.
Experienced role players should have
more elaborate scripts than novices. Rad-
ical changes in the service script should
encounter greater resistance from expe-
rienced role performers, since this in-
volves discarding a reasonably efficient,
well-developed script and learning a new
one.
P3: Role similarity is a potential basis for
classifying services.
If
the key elements of service provid-
er roles are extracted, it should be pos-
sible to categorize services in terms of role
similarity rather than industry similarity.
For example, a bank teller's role may have
more in common with an airline reser-
vation clerk's role than with that of a bank
loan officer. Such a classification scheme
would facilitate the development of gen-
eral service principles underlying en-
counters.
P4: Role behaviors are interdependent. The
appropriateness
of
behavior is deter-
mined by others.
For a service provider in a service set-
ting' others include management, co-
workers, and customers. Thus, the role
player will adjust to the feedback re-
ceived from all members of the audience.
When these groups are not in agreement,
we would expect role ambiguity to be
high. Because role behaviors are inter-
dependent, each player attempts to iden-
tify the other's role early in the interac-
tion to facilitate prediction and also to
adjust personal behavior accordingly.
Thus, role assignment takes place early
in the encounter and influences subse-
quent interaction. We would expect that
the early stages of the encounter are more
important to the ultimate success of the
interchange than are the later stages.
P5: Congruent role expectations facilitate so-
cial interaction.
a. When customers and service provid-
ers read from a common script (high
inter-role congruence), the encounter
is more satisfying.
b. When service employees and the or-
ganization share common role expec-
tations, role clarity and job satisfac-
tion increase.
Both of these propositions reflect the
importance of the shared nature of the
experience. Congruent role expectations
enhance predictability and, hence, de-
crease the amount of effort which must
be expended to complete the transaction.
Predictability is most desired in services
with low involvement. However, even for
high involvement services such as edu-
cation and therapy, role expectations ex-
ist and form a basis for prediction. The
script for a visit to a therapist, for ex-
ample, may include a great deal of vari-
ation and the therapist's behavior may not
be totally predictable. This is because the
role of therapist includes variability rather
than conformity.
If
the patient's script calls
for a supportive, nondirective type of be-
havior and the therapist presents "canned"
solutions, inter-role congruence will be
low and dissatisfaction is likely to be high.
The degree of predictability needed will
vary across service encounters, not the
need for predictability. The subproposi-
tions above imply that management must
be wary of communicating inconsistent
or contradictory expectations to employ-
ees on the one hand and customers on the
other, regarding service levels, image,
customer base, etc.
P6: Discrepant role expectations decrease ef-
ficiency.
When role expectations are discrepant
from actual behavior, communication will
be inhibited and productivity reduced.
When role players read from different
scripts, considerable confusion is likely
to result and the encounter no longer fol-
lows a predictable sequence. This should
result in increased dissatisfaction.
Future Directions
These propositions represent only some of the more
fundamental relationships that can be derived from a
role playing perspective. Many others could be gen-
erated and tested.
In addition to empirically validating these propo-
sitions, the discussion of role theory and service en-
counters gives rise to numerous other questions. A brief
description may indicate some issues for future in-
vestigation.
At the beginning of this article it was necessary to
delimit service encounters, considering only face-to-
face interactions between a buyer and seller. To what
extent do role theory concepts hold when we relax our
definition of a service encounter? For example, do role
expectations operate similarly when the interaction takes
place over the telephone? Another interesting question
concerns man/machine interactions. Most automatic
teller machines appear to be programmed to simulate
the role of a bank teller. Some even call the customer
by name. Can we talk about a man/machine dyad?
An ATM role? To what extent is a theoretical struc-
ture like role theory applicable in this situation?
Another direction for future investigation might well
include polyadic interactions, those interactions which
include multiple service personnel. One might hy-
pothesize that interacting with several service provid-
ers (a maitre d'hotel, wine steward, waiter, and bus
boy) could be studied as a series of dyadic interac-
tions. An interesting question here would be how the
customer integrates the experience. One possibility
would be to use an averaging process. An alternative
might be a lowest common denominator or weakest
link model. Of course it is possible that the dyadic
model is not appropriate for analyzing polyadic inter-
actions at all. A group dynamic approach may be more
suitable, particularly when the customer interacts with
service providers as a group.
These examples illustrate only a few of the many
areas in which the concepts presented could be ex-
tended. Other researchers could easily expand this list.
Regardless of the particular route chosen, it is hoped
that the concepts developed in this paper will stimu-
late further research.
Conclusion
Role theory and the related concepts developed here
make it possible to consider both customer-service
provider interactions as well as service provider-or-
ganization interactions. The emphasis is on the joint
behaviors of the actors. The setting the organization
provides, together with the implicit and explicit cues
it gives service employees, helps to determine the
content of the employee role which, in turn, has an
impact on and is affected by the customer role.
Using a mid-range theoretical structure such as role
theory to examine service encounters permits us to de-
A
Role
Theory
Perspective
on
Dyadic
Interactions
/ 109
velop general principles applicable in a range of ser-
vice settings across individual role performers.
It
min-
imizes the need to treat each service encounter as a
unique experience. Furthermore, the theoretical struc-
ture permits marketing researchers to integrate find-
REFERENCES
Abelson, Robert F. (1976), "Script Processing in Attitude For-
mation and Decision Making," in Cognition and Social Be-
havior, John S. Carroll and John S. Payne, eds., Hillsdale,
NJ: Erlbaum.
Bales,RobertF. (1950),InteractionProcessAnalysis:AMethod
for the Study
of
Small Groups, Reading, MA: Addison-
Wesley.
Bateson, John (1977), "Do We Need Service Marketing?," in
Marketing Consumer Services: New Insights, P. Eiglier et
al., eds., Cambridge, MA: Marketing Science Institute, re-
port no.
77-115,
1-30.
Bell, Martin (1981), "A Matrix Approach to the Classification
of Marketing Goods and Services," in Marketing
of
Ser-
vices, J. H. Donnelly and W. R. George, eds., Chicago:
American Marketing.
Bettman, James (1979), "Memory Factors in Consumer Choice:
AReview," Journal
of
Marketing, 43 (Spring),
37-53.
Bible, B. L. and E. J. Brown (1963), "Role Consensus and
Satisfaction of Extension Advisory Committee Members, "
Rural Sociology, 28,
81-90.
----
and J. D. McComas (1963), "Role Consensus and
Teacher Effectiveness," Social Forces, 42, 225-233.
Block, J. (1952), "The Assessment of Communication Role
Variations as a Function of Interactional Context," Journal
of
Personality, 21, 272-286.
Booms, Bernard H. and Jody Nyquist (1981), "Analyzing the
Customer/Firm Communication Component of the Ser-
vices Marketing Mix," in Marketing
of
Services, James
Donnelly and William George, eds., Chicago: American
Marketing.
Churchill, G. A. and C. F. Surprenant (1982), "An Investi-
gation into the Determinants of Customer Satisfaction,"
Journal
of
Marketing Research, 20 (November),
491-504.
Cotham, James
C.,
III (1968), "Job Attitudes and Sales Per-
formance of Major Appliance Salesmen," Journal
of
Mar-
keting Research, 5 (November),
370-375.
Craik, Fergus and Robert S. Lockhart (1972), "Levels of Pro-
cessing: A Framework for Memory Research," Journal
of
Verbal Learning and Verbal Behavior, 11,
671-684.
Czepiel, John A. (1980), Managing Customer Satisfaction in
Consumer Service Businesses, Cambridge, MA: Marketing
Science Institute.
----,
E. Gutman, M. Solomon, and C. F. Surprenant
(1982), "A Contingency Approach to Understanding Ser-
vice Encounters: Theoretical Inputs for Management Ac-
tion," paper presented at the Workshop on Research into
the Management of Service Businesses, London Business
School.
---,
L. Rosenberg, and A. Akerele (1975), "Perspec-
tives on Consumer Satisfaction," Proceedings, Chicago:
American Marketing, 119-123.
Davis, J. H. (1969), Group Performance, Reading, MA:
Addison-Wesley.
110 /
Journal
of
Marketing,
Winter
1985
ings from other social science disciplines and to apply
them in a service setting. Since control of the service
encounter is a crucial area of managerial concern and
a difficult task to accomplish, the perspective gained
is important.
Dubinsky, Alan J. and Bruce E. Mattson (1979), "Conse-
quences of Role Conflict and Ambiguity Experienced by
Retail Salespeople," Journal
of
Retailing, 55 (Winter),
70-
86.
Eiglier, Pierre and Eric Langeard (1977), "A New Approach
to Service Marketing," in Marketing Consumer Services:
New Insights, P. Eiglier et al., eds., Cambridge, MA: Mar-
keting Science Institute, report no.
77-115.
Evans, Franklin (1963), "Selling as a Dyadic
Relationship-
A New Approach," American Behavioral Scientist, 6 (May),
76.
Ford, N.
M.,
O. C. Walker, and G. A. Churchill, Jf. (1976),
"The Psychological Consequences of Role Conflict and
Ambiguity in the Industrial Sales Force," in Marketing:
/776-1976
and Beyond, K. L. Bernhardt, ed., Chicago:
American Marketing.
Ghiselli, Edwin E. (1978), "The Validity of Aptitude Tests in
Personnel Selection," Personnel Psychology, 26 (Winter),
461-477.
Goffman, Erving (1961), Encounters: Two Studies in the So-
ciology
of
Interaction, Indianapolis: Bobbs-Merrill.
Grove, Stephen J. and Raymond P. Fisk (1983), "The Dra-
maturgy of Services Exchange: An Analytical Framework
for Services Marketing," in Emerging Perspectives in Ser-
vices Marketing, Leonard L. Berry, G. Lynn Shostack, and
Gregory D. Upah, eds., Chicago: American Marketing.
Holbrook, Morris B. (1980), "Introduction: The Esthetic Im-
perative in Consumer Research," in Symbolic Consumer
Behavior, Elizabeth C. Hirschman and Morris B. Hol-
brook, eds., Ann Arbor: Association for Consumer Re-
search.
Homans, George C. (1961), Social Behavior: Its Elementary
Forms, New York: Harcourt.
Jackson, Susan E. (1981), "Measurement of Commitment to
Role Identities," Journal
of
Personality and Social Psy-
chology, 40, 138-146.
Judd, Robert C. (1964), "The Case for Redefining Services,"
Journal
of
Marketing, 28 (January),
58-59.
Kahn, Robert
L.,
Donald M. Wolfe, Robert P. Quinn, J.
Diedrick Snoek, and R. A. Rosenthal (1964), Organiza-
tional Stress: Studies in Role Conflict and Ambiguity, New
York: Wiley.
Kelley, Harold H. and John W. Thibaut (1978), Interpersonal
Relations: A Theory
of
Interdependence, New York: Wiley-
Interscience.
Knisely, G. (1979), "Greater Marketing Emphasis by Holiday
Inns Breaks Mold," Advertising Age (January
15),47-51.
Langer, E. (1978), "Rethinking the Role of Thought in Social
Interaction," in New Directions in Attribution Research, J.
Harvey, W. Ickes, and R. Kidd, eds., Hillsdale, NJ: Erl-
baum.
---,
A. Blank, and B. Chanowitz (1978), "The Mind-
lessness of Ostensibly Thoughtful Action: The Role of
Placebic Information in Interpersonal Interaction, " Journal
of
Personality and Social Psychology, 36, 635-642.
---and
L. Imber (1979), "When Practice Makes Im-
perfect: Debilitating Effects of Overlearning," Journal
of
Personality and Social Psychology, 37, 2014-2024.
Levy, S.
J.,
J. A. Czepiel, and D. W. Rook (1980), "Social
Division and Aesthetic Specialization: The Middle Class and
Musical Events," in Symbolic Consumer Behavior, E. C.
Hirschman and M. Holbrook, eds., Ann Arbor: Association
for Consumer Research.
Liechty, M. and G. A. Churchill, Jr. (1979), "Conceptual In-
sights into Consumer Satisfaction with Services," in Pro-
ceedings, Chicago: American Marketing.
Lovelock, Christopher H. (1980), "Toward a Classification
of Services," in Theoretical Developments in Marketing,
C. W. Lamb and P. M. Dunne, eds., Chicago: American
Marketing,
72-78.
---
and Robert F. Young (1979), "Look to Consumers
to Increase Productivity," Harvard Business Review, 57
(May-June), 168-178.
Lutz, R. J. and P. Kakkar (1976), "Situational Influence in
Interpersonal Persuasion," in B. B. Anderson, ed., Ad-
vances in Consumer Research, V. 3, Cincinnati: Associa-
tion for Consumer Research, 370-378.
McCall, G. J. and J. L. Simmons (1966), Identities and In-
teractions, New York: Free Press.
McMurry, Robert (1961), "The Mystique of Supersalesman-
ship," Harvard Business Review, 39 (March-April), 114.
Mead, George H. (1935), "Mind, Self, and Society," in Works
of
Mead, Charles W. Morris, ed., Chicago: University of
Chicago Press.
Merton, Robert K. (1957), "The Role Set," British Journal
of
Sociology, 8, 106-120.
Oliver, Richard L. (1974), "Expectancy Theory Predictions of
Salesmen's Performance," Journal
of
Marketing Research,
11 (August), 243-253.
---
and A. P. Brief (1977-1978), "Determinants and
Consequences of Role Conflict and Ambiguity among Re-
tail Sales Managers," Journal
of
Retailing, 53 (Winter),
47-58.
Pennington, Allan L. (1968), "Customer-Salesman Bargaining
Behavior in Retail Transactions, "Journal
of
Marketing Re-
search, 5 (August), 255-262.
Rathmell, John M. (1966), "What Is Meant By Services?,"
Journal
of
Marketing, 30 (October),
32-36.
Riordan, Edward
A.,
Richard L. Oliver, and James H. Don-
nelly, Jr. (1977), "The Unsold Prospect: Dyadic and At-
titudinal Determinants," Journal
of
Marketing Research,
14 (November), 530-537.
Rose, Arnold M. (1962), "A Systematic Summary of Sym-
bolic Interaction Theory," in Human Behavior and Social
Processes: An Interactionist Approach, Arnold M. Rose,
ed., Boston: Houghton Mifflin.
Sarbin, Theodore R. and Vernon L. Allen (1968), "Role The-
ory," in The Handbook
of
Social Psychology, 2nd ed., V.
I, Gardner Lindzey and Elliott Aronson, eds., Reading, MA:
Addison-Wesley.
Schank, Roger C. (1980), "Language and Memory," Cogni-
tive Science, 4, 243-284.
Schneider, Benjamin (1980), "The Service Organization: Cli-
mate is Crucial," Organizational Dynamics, 9(Autumn),
52-65.
Sheth, J. N. (1967), "A Review of Buyer Behavior," Man-
agement Science, 13, 8718-8756.
----
(1975), "Buyer-Seller Interaction: A Conceptual
Framework," Proceedings, Sixth Annual Conference, As-
sociation for Consumer Research, 382-386.
---
(1982), "Toward An Integrated Theory of Market-
ing," presentation given at the AMA Doctoral Consortium,
University of Minnesota.
Shostack, G. Lynn (1977), "Human Evidence: A New Part
of
the Marketing Mix," Bank Marketing (March),
32-34.
Smith, E. E. (1957), "The Effects of Clear and Unclear Role
Expectations on Group Productivity and Defensiveness,"
Journal
of
Abnormal Sociology and Psychology, 55 (no. 4),
213-217.
Smith, Ruth A. and Michael J. Houston (1983), "Script-Based
Evaluations of Satisfaction with Services," in Emerging
Perspectives in Services Marketing, Leonard L. Berry, G.
Lynn Shostack, and Gregory D. Upah, eds., Chicago:
American Marketing.
Swan, John and Linda Coombs (1976), "Product Performance
and Consumer Satisfaction: A New Concept," Journal
of
Marketing Research, 13 (April),
25-33.
---
and I. Fredrick Trawick (1978), "Testing an Ex-
tended Concept of Consumer Satisfaction," in New Dimen-
sions
of
Consumer Satisfaction and Complaining Behavior,
Ralph L. Day and H. Keith Hunt, eds., Bloomington: In-
diana University Press,
56-61.
Tagiuri, Renato (1969), "Person Perception," in The Hand-
book
of
Social Psychology, 2nd
ed.,
V. 3, Gardner Lindzey
and Elliot Aronson, eds., Reading, MA: Addison-Wesley.
Thibaut, John W. and Harold H. Kelley (1959), The Social
Psychology
of
Groups, New York: Wiley.
Thomas, E. J. (1959), "Role Conceptions and Organizational
Size," American Sociological Review, 24,
30-37.
Torrance, E. P. (1954), "The Behavior of Small Groups under
the Stress of Conditions of 'Survival'," American Socio-
logical Review,
19,751-755.
Tosi, Henry L. (1966), "The Effects of Expectation Levels
and Role Consensus on the Buyer-Seller Dyad," Journal
of
Business, 39 (October),
516-529.
Uhl, K. P. and G. D. Upah (1983), "The Marketing of Ser-
vices: Why and How Is
It
Different," in Research in Mar-
keting, V. 6, J. N. Sheth, ed., New York: Elsevier.
Webster, Frederick E. (1968), "Interpersonal Communication
and Selling Effectiveness," Journal
of
Marketing, 32 (July),
7-13.
Weinstein, E. A. and P. Deutschberger (1963), "Some Di-
mensions of Altercasting," Sociometry, 26,
454-466.
Weitz, Barton A. (1981), "Effectiveness in Sales Interactions:
A Contingency Framework," Journal
of
Marketing, 45
(Winter), 85-103.
White, R. W. (1959), "Motivation Reconsidered: The Concept
of Competence," Psychological Review, 66, 297-333.
Whyte, William F. C. (1948), Human Relations in the Res-
taurant Industry, New York: McGraw-Hill.
Willett, Ronald P. and Allan L. Pennington (1966), "Cus-
tomer and Salesman: The Anatomy of Choice and Influence
in a Retail Setting," in Science, Technology
and
Market-
ing, R. M. Haas, ed., Chicago: American Marketing.
Wilson, David T. (1977), "Dyadic Interactions," in Consumer
and Industrial Buying Behavior, Arch Woodside, Jagdish
Sheth, and Peter Bennett, eds., New York: North-Holland,
355-365.
---
and Lome Bozinoff (1980), "Role Theory and Buy-
ing-Selling Negotiations: A Critical Overview," in Mar-
keting in the 80's, Richard Bagozzi, ed., Chicago: Amer-
ican Marketing, 118-121.
Wyckham, R.
G.,
P. T. Fitzroy, and G. D. Mandry (1975),
"Marketing of Services: An Evaluation of the Theory," Eu-
ropean Journal
of
Marketing, 9,
59-67.
A
Role
Theory
Perspective
on
Dyadic
Interactions
/ 111
