Arctic Geopolitical Configuration
107contemporary international relations.”^21 However, individual indicators are
listed and briefly described but their technical specification remained
missing.
At the same time, Wolfson, Madjd-Sadjadi, and James highlight the
relevance of a cluster analysis in understanding the full range of interactions
among political, economic, and conflict-related variables. The authors
argue that “it is appropriate to begin without imposing too many
restrictions on the analysis.”^22 Generated within several disciplines of
social research (reflecting the development in political, economic and
conflict environments) the variables are assumed to be part of a time-
dependent, nonlinear, interactive, and complex system that is yet to be
discovered using the technical capabilities of a cluster analysis.
Notwithstanding intrinsic paradoxes,^23 the approach is popular in
contemporary science.^24 It is not one specific algorithm, but a series of
exploratory, descriptive and multivariate statistical measurements to sort
objects (cases) into groups in a way that the degree of association between
two cases is maximal if they belong to the same group and minimal
otherwise.^25 In other words, a cluster analysis allows for revealing
geopolitical groups (“zones of consensus”) and intergroup buffer areas
(“zones of conflict”). It should be remembered, however, that it discovers
structures in data without explaining why they exist, hence a qualitatively
analytical assessment of results is necessary. We therefore extend the
results of clustering with an evaluation of the relationship between
temperature variation and group membership stability and discover the
probability of occurrence of intraregional conflict in the Arctic provinces.
The choice of variables in this study is justified by integrating the
interdisciplinary logic of these abovementioned theoretical approaches.
The geopolitical performance of each Arctic province is assumed to be
defined as the simultaneous interaction of geographic, economic,
demographic, military, and regional attributes. We operate on a sub-
national level of spatial analysis. Traditionally, geographic and
demographic factors have been considered strategically important due to
two primary conditions for a state’s sovereignty: territory and
(^21) Ibid.
(^22) Wolfson, Madjd-Sadjadi and James, “Identifying National Types: A Cluster
Analysis of Politics, Economics, and Conflict”, 608.
(^23) Among others, whilst cluster analysis is structure seeking, the procedure itself is
structure imposing.
(^24) See Hartigan (1975) for a summary of the many published studies reporting the
results of cluster analyses.
(^25) See Kendall (1973).