This short presentation aims to delineate the basic features of an alternative theoretical framework for the comparative analysis of separatist mobilisation. This framework is used in my dissertation, which is a small-N comparison of separatist mobilisation. The separatist movements under consideration are the West Papuans in Indonesia, the Oromo in Ethiopia and the Abkhaz in Georgia. Apart from a set of features common to the selected movements and states, I argue that comparative analysis is feasible due to the delimited character of the research question itself: In a relatively recent review article, James Fearon and David Laitin have found that a major puzzle of studies concerning ethnic violence is “why publics follow elites.” Similarly, this study aims to explain primarily how – and subsequently why- publics follow elites in instances of separatist mobilisation. It is moreover examined what is the role and impact of ethnic identities in such instances. The core research question leads us to focus on social interactions and collective processes rather than on macro-structural determinants, causal variables and necessary conditions of separatism and ethnic mobilisation. Typically, small-N comparisons are considered by the bibliography to be data-driven rather than theory-driven. Researchers of small-N case-oriented studies “examine cases as wholes, not as collections of variables” (Ragin 1987: 52) and are usually concerned with identifying “regularities, patterns, mechanisms and reccurent processes […] in inductive fashion through the systematic analysis of empirical data” (Brubaker & Laitin 1998: 430). Still, one needs theoretical guidelines in order to define what sort of mechanisms should be employed, what type of regularities are we searching for and what is the scope of possible relations of causality. As I will argue in the next section, traditional theories of nationalism, ethnic conflict and ethnic mobilisation are not sufficiently capable of providing such answers. I argue that it is more fruitful to combine some elements of the constructivist theories of ethnicity with the literature of social movements and contentious politics. Particularly I follow the theoretical proposal of Charles Tilly, especially as it is set out in the book Dynamics of Contention (DOC) which co-wrote with Doug McAdam and Sidney Tarrow. The core preposition of this framework is that episodes of contentious politics should be explained through the search for recurring causal mechanisms and processes. The various sequences and combinations of common mechanisms produce different outcomes of contentious episodes. In this analysis I employ four mechanisms or sets of mechanisms that appear capable of explaining how separatist mobilisation takes place in three different settings. The proposed mechanisms are 1) brokerage, 2) social appropriation of pre-existing structures and frame-bridging, 3) boundary activation/deactivation and 4) collective attribution of threat and opportunity. The combination of the proposed mechanisms seeks to capture the dynamic interplay among ethnic identities, separatist goals, institutions and organisational resources in a more satisfying way than the standard theories of ethnic mobilisation and separatism.