Quality & Quantity 38: 547–557, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
Classiﬁcation and the Relations of Meaning
ANDREAS SCHNEIDER and ALDEN E. ROBERTS
Department of Sociology, Antropology and Social Work, Texas Tech University, P.O. Box 41012,
Lubbock TX 79409-1012, U.S.A.
Abstract. This paper intends to demonstrate the parallels between a qualitative methodology, com-
ponent analysis, which is predominantly used in cognitive anthropology and linguistics, and the
quantitative explorative method-cluster analysis. Social identities and their related structural categor-
ies serve as examples of the method. In the methodology and logic involved in the categorization of
meaning, abstraction is the key difference between connotative and structural meaning. Abstraction
and the identiﬁcation of higher order categories in cluster analysis are compatible with the extraction
of structural meaning from the semantic differential ratings of the affective meanings of identities.
The dichotomy of exclusion and inclusion is the most relevant relation for the qualitative analysis
of meaning and can be mathematically operationalized by Euclidean squared distance in k-means
This paper intends to demonstrate the parallels between a qualitative methodology,
component analysis, which is predominantly used in cognitive anthropology and
linguistics, and the quantitative explorative method-cluster analysis. Component
analysis is a qualitative method that was developed by Ward Goodenough (1956,
1981), an anthropologist and ethnolinguist. He subjected kinship terminology to
component analysis. The method involves what Nida (1975) terms the vertical
analysis of meaning. More inclusive meanings of categories are compared with
less inclusive meanings. One of the problems faced by component analysis is that
meaning and relationships between meanings are multidimensional.
The multidimensionality of component analysis can be demonstrated using an
example of social identities. Identities are individual representations of social roles
(Darendorf 1965, 1972; Strauss 1994; Stryker 1980; Turner 1962). According to
Cooley’s (1922) concept of looking-glass self, people observe themselves in the
eye of the other. Mead (1934) regarded this reﬂexivity as a key to developing a self.
Internal imaginary interaction between the societal “me” and individual “I” leads
to the self-conscious ego called self. Mead distinguished between signiﬁcant others
(those whose opinions are important to a person) and generalized others (the “av-
erage” person). The interaction of a person with others, and internal conversations
with previous stages of oneself, lead to the development of the self.
However, there are limits to the plasticity of the self. Building the self (Burke
1980; Stryker 1980; Stryker and Burke 2000) through social histories of interaction
(Strauss 1994), social identities reﬂect the structure of society. Social structure
has been conceptualized as social class, exchange networks, role and status sets,