It is a basic axiom in representation studies to understand the polis in terms of groups, which may be determined by geographical borders, ideology, class, gender, age or ethnicity. As a consequence representation of the polis also occurs group-wise. Political scientists are concerned with the composition of electoral lists, party cadres, executives and representative assemblies in terms of geographical spread or ethnic background. Concerning substantive representation, they analyze whether and to what extent women’s issues result in policy outcomes or whether policy congruence varies according to class. At a general level such research questions deal with the quality of democratic representation – it is generally accepted that no meaningful entity should be systematically excluded from processes of representation (descriptive nor substantive).
An intersectional perspective on representation poses a fundamental problem for our understanding of how representation works and should work. Intersectionality questions the groups that we have accepted to be the structuring principle of politics. The concept of intersectionality implies that the identity of a person is formed not by one single category but instead by intersections and interactions of a variety of identities such as gender, ethnicity, age, sexual orientation. The broad categories that we are used to apply in political and policy contexts have become acknowledged to be not fine grained enough to understand the complexity of society. Existing analyses of gender balance in political institutions may not inform us about the representation of specific groups (e.g. older Roma women). Acknowledging the reality of intersectionality seems thus to result in a dilemma: on the one hand, it is simply no longer desirable to keep conducting the traditional type of representation studies using the broad single categories, but, on the other hand, it is not feasible to extend the same approach to every intersectional subgroup imaginable.
This paper explores a possible solution to this ‘intersectionality dilemma’. It discusses the potential of taking symbolic representation as the starting point of studies on other dimensions of representation, i.e. descriptive and substantive representation. Symbolic representation refers to the way citizens feeling represented. Symbolic representation received far less scholarly attention than descriptive and substantive representation. Generally, empirical studies – thereby often based on ‘critical mass’- or ‘politics of presence’-theories- start with identifying the number of a particular group in legislatures (descriptive representation) and then study the extent to which their interests are represented (substantive representation). (Lack of) symbolic representation is often assumed to be an automatic result of (lack of) descriptive and substantive representation. There is thus a ‘logical sequencing’ of the dimensions of representation in research agenda’s running from descriptive to substantive to symbolic representation. This paper re-orients this sequencing and explores the pros and cons of taking symbolic representation as the starting point of representation studies. More precisely, the paper explores the potential of inductively determining by whom and by which political debates and discourses citizens feel represented (thereby avoiding the use of a priori defined group categories) for next determining the levels of descriptive and substantive representation.
