Relational Models

Text-only Preview

Relational Models 1


Relational Models and Horizontal and Vertical Individualism/Collectivism:
A Cross-Cultural Comparison of Americans and Japanese

Ascan F. Koerner

Manako Fujiwara

Dept. of Speech Communication

University of Minnesota

Minneapolis, MN 55455

Paper presented at the annual NCA convention in Seattle, WA, November 9-12, 2000.

Relational Models 2
Various researchers (i.e., Singelis, Triandis, Bhawuk, & Gelfand,1995; Triandis &
Gelfand, 1998) have suggested that the cultural variables of horizontal and vertical
individualism and collectivism correspond to Fiske’s (1991, 1992) relational models of
collectivism, authority ranking, equality matching, and market pricing. In this study, we
tested this claim in a cross-cultural comparison between American and Japanese
respondents’ use of relational models in three different relationships. Results provide
evidence that cultural differences are reflected in relational model use such that communal
sharing was associated with a horizontal cultural orientation, authority ranking was
associated with a vertical cultural orientation, equality matching was associated with
horizontal individualism, and market pricing was associated with horizontal individualism
and vertical collectivism. In addition, increased use of communal sharing was also
associated with increased intimacy in respondents’ relationships.

Relational Models 3

There are two recurring problems cross-cultural researchers have to confront in
their research. First, much past research has focused on cultural differences using a single
dimension, most often individualism-collectivism. Analyzing cultures along only one
dimension, however, is problematic because cultures are far too complex to be accurately
described by a single dimension. Second, culture does not exist independently from the
individuals that constitute them. That is, culture is a psychological as well as an
sociological phenomenon and descriptions that do not take both perspectives into
consideration are necessarily limited in their validity.

The first problem has been addressed by researchers using additional dimensions to
conceptualize cultural differences, such as the typology of horizontal and vertical
individualism and collectivism proposed by Singelis and Triandis and their colleagues
(Singelis, Triandis, Bhawuk, & Gelfand, 1995; Triandis, 1995; Triandis & Gelfand, 1998).
Similarly, the second problem has been addressed by researchers using individual level
conceptualizations of individualism and collectivism, such as the concepts of ideocentrism
and allocentrism proposed by Triandis, Leung, Villareal, and Clack (1985).
To date, however, there is no research that has addressed both these problems
simultaneously, presumably for lack of an adequate theoretical model. In this article, we
test whether Fiske’s (1991, 1992) relational model theory can be used to adequately
conceptualize Singelis’s and Triandis’s cultural typology on the level of individual
psychology. Specifically, we hypothesized that Fiske’s communal sharing corresponds
with horizontal collectivism, Fiske’s authority ranking corresponds with vertical
individualism, equality matching corresponds with horizontal individualism, and market
pricing corresponds with vertical individualism.

Relational Models 4
Individualism and Collectivism
In cross-cultural research, the variable that has been most thoroughly studied is
probably the individualistic/collectivistic dimension (Triandis, 1989). Individualistic
cultures focus on independence and personal identity whereas collectivistic cultures focus
on interdependence and group harmony (Hofstede, 1984; Triandis, 1995). Bravery,
creativity, self-reliance, solitude, and frugality are valued in individualistic cultures,
whereas reciprocity, obligation, duty security, tradition, dependence, harmony, obedience
to authority, equilibrium, and proper action are valued in collectivistic cultures (Triandis,
The study of collectivism and individualism has been extraordinary fruitful becasue
the two dimension have shown to have a strong impact on a wide range of behaviors in
many cultures. For example, a study by Hui and Villareal (1989) explored the differences in
psychological needs between people in individualistic and collectivistic cultures. People in
individualistic cultures value self-reliance, and autonomy, whereas people in collectivistic
cultures value interdependence, affiliation, succorance, abasement and nurturance. In
conflict situations, Ting-Toomey and Kurogi (1998) found that people in individualistic
cultures used direct and face-threatening strategies whereas people in collectivistic cultures
used indirect and face-saving strategies. The two cultures differ in perception on conflict
management styles as well. In individualistic cultures, individual goals are addressed and
self-face is enhanced, whereas in collectivistic cultures group needs are addressed and
mutual and in-group face are enhanced (Ting-Toomey & Kurogi, 1998).
Even though individualism and collectivism are often discussed as dichotomous
categories, most researchers would argue that individualism and collectivism are polar

Relational Models 5
opposites along one continuum (Triandis, 1989). That is, cultures vary from one another
based on the degree of their individualism or collectivism rather than belonging into one of
two categories. For example, according to Hofstede (1984), the United States, a typical
individualistic culture had the highest Country Individualism Index Values (CIIV) (91)
among 39 countries, whereas CIIV of Venezuela was the lowest (12). Japan scored 46,
which was a little less than the mean (51). Overall, Western countries tend to be more
individualistic, whereas South American and Asian countries tend to be more collectivistic.
Vertical and Horizontal Individualism and Collectivism
The individualism/collectivism dimension alone, however, is insufficient in
explaining differences between cultures. For example, Triandis (1995) observed that there
were great differences among highly individualistic cultures as well as among highly
collectivist cultures. For example, American culture is very different from Swedish culture,
even though both are very individualistic, and similarly, Korean culture is very different
from the culture of an Israeli kibbutzim, even though both are very collectivist. Based on
such observations, Triandis and his colleagues (Singelis, Triandis, Bhawuk, & Gelfand,
1995; Triandis, 1995; Triandis & Gelfand, 1998) have argued that there is at least one other
important dimension that differentiates cultures: the linear ordering of relationships as
being either vertical or horizontal. Cultures that emphasize vertical relationships
differentiate persons from one another according to rank and create a strict social hierarchy.
Cultures that emphasize horizontal relationships, in contrast, stress the equality of all
persons and create a flat social hierarchy. Although Triandis and his colleagues were the
first to use the horizontal and vertical labels to describe this dimension, other scholars (e.g.,

Relational Models 6
Hofstede, 1984; Schwartz, 1994, 1999) have discussed very similar dimensions in their
work as well.
For example, the concept of vertical and horizontal relationships corresponds to
power-distance, one of the important cultural dimension identified in Hofstede’s (1984)
seminal work on cross-cultural differences. Like individualism and collectivism, power
distance is conceptualized to exist along a continuum from low to high. According to
Hofstede’s Power Distance Index (PDI), Asian and South American countries such as the
Philippines (94), Mexico (81), and Venezuela (81) are highest on power distance, whereas
European countries such as Austria (11), Israel (13) and Denmark (18) are lowest. Both the
United States (40) and Japan (54) are in the mid-range among the 39 countries investigated
by Hofstede (1984).
Dimensions similar to individualism and collectivism, and vertical and horizontal
relationships also have been identified by Schwartz (1994, 1999). Among the seven
cultural value types identified by Schwartz, the autonomy - conservatism dimension is
roughly equivalent to individualism and collectivism and Schwartz’s hierarchy - egalitarian
dimension is conceptually similar to horizontal and vertical relationships. Hierarchical
cultures value social power, authority, humility and wealth, whereas egalitarian cultures
value equality, social justice, freedom, responsibility and honesty (Schwartz, 1999).
Combining the two dimensions of individualism and collectivism with vertical and
horizontal relationship orientations allows researchers to distinguish between four different
cultural types (Singelis et al, 1995; Triandis, 1995; Triandis & Gelfand, 1998). They are
horizontal individualism (HI), vertical individualism (VI), horizontal collectivism (HC),
and vertical collectivism (VC). According to Singelis et al. (1995) and Triandis & Gelfand

Relational Models 7
(1998), in HI cultures people want to be unique and independent but they are not
particularly interested in being distinguished from others and do not seek high status.
People in VI cultures want to be distinguished from others and seek high status through
competition. In HC cultures equality, interdependence and sociability are important and
people see themselves as part of an in-group. People in VC cultures also see themselves as
members of an in-group, however, they see differences among group members, and status
inequality is expected. Of course, these cultural types not only describe different cultures
but sub-cultures as well. In their study of collectivism within the United States, for
example, Singelis et al. (1995) found that vertical collectivism is seen more often among
Asian Americans than European Americans.
Instantiations of Culture on the Level of the Individual
Although culture is often defined as shared norms and values, culture also exists on
an individual level. Culture, in other words, is also the product of the attitudes, believes,
and behaviors of individuals within a group of persons of similar values and norms.
Consequently, one can also describe culture from a psychological perspective, which
requires one to investigate how cultural norms and values are represented in the cognition
of individuals. Triandis, Leung, Villareal, and Clack (1985) did precisely that for the
cultural dimensions of individualism and collectivism. They argued that the psychology
and behaviors of individuals are never completely uniform within any given culture and it
is therefore useful to identify individualism and collectivism on the individual level as
idiocentrism and allocentrism, respectively. Likewise, Markus and Kitayama (1991)
developed the notion of independent and interdependent construals of the self to describe
cultural differences on the level of the individual.

Relational Models 8
Idiocentrism is the expression of individualist values and norms on an individual
level. It is concerned with values such as a comfortable life, competition, pleasure and
social recognition. According to Triandis et al. (1985), it is related to need for achievement,
alienation, and anomie. Allocentrism is the expression of collectivist values on an
individual level. It is concerned with values such as cooperation, equality, and honesty. It is
strongly related to perceptions of social support received, and quality and satisfaction with
the support.
Markus and Kitayama’s (1991) interdependent self is similar to allocentrism,
whereas the independent self is similar to idiocentrism. Their notion of self is that of a
cognitive construct that others cannot know directly and that is beyond “a physical or
ecological sense of self […], and of the continuous flow of thoughts and feelings” (p. 225).
Markus and Kitayama argued that although some aspects of self are universal, most
important aspects of the self are determined by culture. According to their definitions, the
interdependent self is concerned with connection to others and social relationship, whereas
the independent self is concerned with separation from others, autonomy, and
independence. Marcus and Kitayama (1991) argued that individuals in Western cultures
hold more independent construals of the self and individuals in Asian and Hispanic cultures
hold more interdependent construals of the self.
One advantage of conceptualizing cultural variables as existing on an individual
level is that it explains differences that exist not only between individuals of different
cultures, but also offers an explanation for variations among individuals of the same
culture. For example, research found evidence for the coexistence of ideocentrism and
allocentrism in India (Sinha & Tripathi, 1994), China (Ho & Chiu, 1994), and Japan

Relational Models 9
(Yamaguchi, 1994). Triandis (1995) also suggested that Japan and China, both
collectivistic countries, are showing increasing evidence of idiocentric traits among their
members, especially among the younger generations. Similarly, people in Scandinavian
countries, which are usually described as individualistic, recognize the importance of
in-groups, which is indicative of the allocentrism of the individuals.
Relational Models and Cultural Differences

Allocentrism and idiocentrism, proposed by Triandis et al. (1985), are the
individual level instantiations of the cultural dimension of individualism and collectivism.
However, as we have argued above, individualism and collectivism alone are incomplete
descriptors of culture and their explanatory power is dramatically increased by adding the
cultural level dimension of vertical versus horizontal relationship orientation. Therefore, to
arrive at a better conceptualization of how culture is instantiated on the level of the
individual, idiocentrism and allocentrism have to be augmented by individual level
representations of horizontal and vertical relationship orientations as well.

One way to achieve this augmentation would be to design a new measure for
horizontal and vertical relationship orientation on the individual level. Although ostensibly
relatively simple, this method has the great disadvantage that even though strong
theoretical reasons suggest their existence, there are so far no data to empirically support
this theoretical claim and the psychometric properties of such newly designed measures are
obviously unknown. A better strategy, therefore, would be to identify an existing theory
that conceptualizes and measures horizontal and vertical relationships, much like Triandis
et al.’s (1985) and Markus and Kitayama’s (1991) individual level measurements of
individualism and collectivism. Fortunately, precisely such a theory exists with Fiske’s

Relational Models 10
(1991, 1992) relational model theory, which not only has concepts similar to horizontal and
vertical relationship orientations on the individual level, but concepts similar to
allocentrism and idiocentrism as well (Singelis et al., 1995).
Relational Model Theory
Fiske (1991, 1992) argued that people interact with one another to construct and to
participate in relationships that are based on one or more of only four fundamental
relational models. He further claimed that the impact of these basic models is: pervasive,
that is, governing all domains and aspects of social relationships; exhaustive, meaning no
other fundamental types of relating exist; and generative, meaning that all relationships are
constructed using the four basic relational models he proposed. The four relational models
defined by Fiske are communal sharing (CS), authority ranking (AR), equality matching
(EM) and market pricing (MP). These four models are types of relating rather than
relationship types. That is, persons in actual and ongoing interpersonal relationships can
and usually do relate to others in ways consistent with all four relational models within the
context of the same relationship. In some relationship domains an ongoing relationship
might be characterized by communal sharing, whereas in other relationship domains the
partners employ a hierarchy ranking and equality matching. An example would be a couple
that pools its income in one checking account (CS), where one partner takes an instructor
role in gardening and tells the other what to do (AR), and where both partners take turns
cleaning dishes (EM).
The model of CS is similar to Mills and Clark’s (1982) description of communal
relationships, where there are no differences between the individuals within the
relationship. That is, relating according to this model is based on the perception that the