Start with How Civil Wars Start by Barbara Walter for the accessible version of what political scientists currently measure, and Civil Wars - a History in Ideas by David Armitage for the longer question of what the term has meant since Rome. Between them you get the empirical literature and the conceptual history without opening a journal.
Then be clear about what you are walking into. This field has competing explanatory frameworks, not a settled answer, and the disagreement is substantive rather than a matter of emphasis. Roughly: grievance accounts hold that people rebel when conditions become intolerable relative to expectations; state-capacity accounts hold that rebellion happens where states are too weak to prevent it, more or less regardless of grievance; ethnic-structure accounts hold that the composition and organisation of groups shapes both whether conflict occurs and what form it takes. Real cases can be read through all three. Nothing below adjudicates between them, and neither should you until you have heard each argued at length.
These are books about massacre, torture, forced displacement and genocide, described in detail. That is what the evidence consists of.
The trade books
How Civil Wars Start is written for a general audience and applies the comparative findings — anocracy, factionalism, demographic reversal — to the contemporary United States. That application is contested by other specialists, and it is the part of the book to hold most loosely. Civil Wars - a History in Ideas is intellectual history, tracing how Rome's civil wars framed every later use of the term. Wars, Guns, and Votes by Paul Collier is the popular version of his economic research programme, which argues that opportunity and financing predict rebellion better than stated grievances do. Collier's work is influential and heavily criticised on both data and framing, so read it as one strong position.
The founding academic statements
Each framework has a classic, and reading them in sequence is the point of this path. Why men rebel by Ted Robert Gurr, published in 1970, is the grievance case: relative deprivation, the gap between what people expect and what they get, as the driver of political violence. Political order in changing societies by Samuel Huntington, from 1968, is the institutional case: disorder follows when social mobilisation outruns the capacity of institutions to absorb it. Ethnic groups in conflict by Donald Horowitz, from 1985, is the structural case, and remains the most detailed comparative account of how ethnic party systems and group ranking produce conflict.
All three are dense academic books of their era, and each has been criticised at length by the generation that followed. Read them for the arguments, not as current findings.
The microfoundations turn
The last quarter-century of research moved from why wars start to how violence actually works inside them, and this is where the field is strongest. The logic of violence in civil war by Stathis Kalyvas is the pivotal book: it argues that much wartime violence is locally produced, driven by denunciation and private score-settling under conditions of contested control, rather than by the ideological cleavage the war is named after. It is a graduate text with formal models and demands patience.
Inside Rebellion by Jeremy Weinstein asks why some rebel groups discipline their fighters and others prey on civilians, and answers it through the resources available at recruitment. Insurgent Collective Action and Civil War in El Salvador by Elisabeth Wood is field-based, built on interviews with campesinos, and takes seriously that people joined for reasons of dignity and agency that the economic models do not capture. Killing neighbors by Lee Ann Fujii is the Rwanda study that most directly undercuts the ethnic-hatred explanation, showing local social ties determining who killed whom.
Rebel Rulers by Zachariah Mampilly examines rebel governance — insurgents running courts, schools and taxation — which reframes armed groups as would-be states rather than as spoilers.
Wider and older frames
The Dark Side of Democracy by Michael Mann argues that murderous ethnic cleansing is a byproduct of democratisation in multi-ethnic settings rather than an aberration from it; it is a large comparative work and its thesis is disputed. The Deadly Ethnic Riot is Horowitz again, on the riot specifically as a form with its own logic. On violence by Hannah Arendt is a short 1970 essay separating power from violence, and it is philosophy rather than social science — a useful corrective at the end.
Follow the full path and you will be able to say which framework a given account is using.
Follow the full ordered path here: Civil Wars and Political Violence: The Best Books to Read, in Order.
