Great many violent events happened during 1991-2005 in the 12 states that emerged after the collapse of the USSR but only a few civil wars are registered in the major datasets. That brings up a number of questions about the operational definitions of civil war that generally point in the direction of shifting the research attention from refining the quantitative parameters to grasping the essense of the phenomena in question. It is proposed that civil war partially overlaps with several other type of violent crisis: inter-state wars, civil unrest and revolutions, internal repression, military coups and mutinies, banditry and organized crime, and terrorism. These overlaps create six 'gray zones' where only very nuanced examination rather than application of rigid criteria could help in distinguishing civil wars from other crises. Therefore, data collection based on a single "robust" definition, which incorporates several verifiable parameters, is not necessarily the only path to scientific knowledge about civil wars.