In "The Ontology of Political Violence" Stathis Kalyvas argues:
"Civil wars are typically described as binary conflicts, classified and understood on the basis of what is perceived to be their overarching issue dimension or cleavage: we thus speak of ideological, ethnic, religious, or class wars. Likewise, we label political actors in ethnic civil wars as ethnic actors, the violence of ethnic wars as ethnic violence, and so on. Yet such characterization turns out to be trickier than anticipated, because civil wars usually entail a perplexing combination of identities and actions.
Consider the following description of the American War of Independence in South Carolina: “There came with the true patriots a host of false friends and plunderers. And this was true of both sides in this terrible struggle. The outlaw Whig and the outlaw Tory, or rather the outlaws who were pretended Whigs and Tories as the occasion served, were laying waste the country almost as much as those who were fighting for the one side or the other.” Years later, Abraham Lincoln described the Civil War in the American West as a situation in which “murders for old grudges, and murders for pelf, proceed under any cloak that will best cover for the occasion.”
The Chinese Civil War was often fought by diverse and shifting coalitions of bandits and local militias; for a long time, the Communists were for the bandits “only one of several possible allies or temporary patrons.” In Manchuria, for instance, it was extremely difficult to differentiate between members of the Anti-Japanese Resistance and bandits because moving from one to another was very common: it is estimated that 140,000 of a total 300,000 resistance members had a bandit background. Common criminals were also used extensively during the Cultural Revolution. The determinants of violence in the province of Antioquia during the Colombian Violencia were “far more complex than any innate, unavoidable differences between monolithic groups of Liberals and Conservatives—the traditional explanation for la Violencia—might suggest”; in fact, “the point of la Violencia, even in supposed areas of ‘traditional settlement’ where partisan objectives were the guiding force behind armed insurrection, is that it was multifaceted and ambiguous, that politics and economic considerations can never be considered as discrete forces."
In short, ambiguity is endemic to civil wars; this turns their characterization into a quest for an ever-deeper “real” nature, presumably hidden underneath misleading facades—an exercise akin to uncovering Russian dolls. Thus, it is often argued that religious wars are really about class, or class wars are really about ethnicity, or ethnic wars are only about greed and looting, and so on. The difficulty of characterizing civil wars is a conceptual problem rather than one of measurement. If anything, the more detailed the facts, the bigger the difficulty in establishing the “true” motives and issues on the ground, as Paul Brass has nicely shown in the case of ethnic riots in India. An alternative is to recognize, instead, that the motives underlying action in civil war are inherently complex and ambiguous."