Introduction

Defining ‘region’ in IR is controversial. Some scholars emphasize geographical proximity while others give importance to the cognitive and ideational factors, yet a third group seeks to combine the two perspectives to define a region. Earlier studies on region and regionalism generally highlighted geographical proximity as an essential criterion of a region. Today, many scholars also posit that ‘a region is firmly rooted in territorial space: a group of people living in a geographically bounded community, controlling a certain set of natural resources, and united through a certain set of cultural values and common bonds of social order forged by history’. Region-ness, these scholars argue, is ‘the convergence of several dimensions’ such as ‘cultural affinity, political regimes, security arrangements and economic policies’, which generates ‘regional coherence within a particular geographic area’. The spatial factor is also considered important in contemporary studies on economic regionalism in which it is argued that economic regionalism primarily hinges ‘on the importance of geographic proximity’. Hence, ‘Geography’, as a scholar has concluded, ‘should not … be dismissed outright as a starting point for identification of regions’.

Many scholars, however, find the essentialist, conventional view of defining a region primarily based on the ‘geographic proximity’ factor problematic because it presupposes an objective, static phenomenon, which ignores the dynamics of social and political forces. Scholars belonging to the constructivist school of thought, in particular, dismiss geographic proximity as a critical factor and define region in nonphysical terms. They argue, countries sharing a communal identity comprise a region regardless of their location. According to Acharya, ‘regions are not a geographic given, but are socially constructed, made and remade through interactions’. Peter J. Katzenstein similarly maintains that regions’ ‘geographic designations are not “real”, “natural” or “essential”, they are socially constructed and politically contested and thus open to change’.

A third group of scholars seeks to combine the two perspectives and argues that members of a common region generally share cultural, economic, linguistic and political ties. Kym Anderson and Hege Norheim have noted that ‘while there is no ideal definition [of a region], pragmatism would suggest basing the definition on the major continents and subdividing them somewhat according to a combination of cultural, language, religious, and stage-ofdevelopment criteria’.