This study responds to the endemic lack of clarity and consensus afflicting academic and policy discussions on the causes of ideological violence and, by extension, the appropriate means for preventing/containing it. I trace, conceptualise, and problematise the long-standing debate between two deeply entrenched oppositional camps or ‘paradigms’ – heuristically dubbed the ‘conservative’ and ‘progressive’ paradigms of ideological violence, respectively – that propose competing explanations for the causation of ideological violence; the former considering it a product of irrational individual dysfunction, the latter viewing it as a rational (if often misguided) response to societal dysfunction. Further, I show that extant attempts at reconciling/synthesising these paradigms have, to date, proven problematic. I explore how and why these opposing paradigms emerged and why debate between them persists. I argue that they are shaped, perpetuated and marred by multiple extra-academic dynamics and naturalised assumptions and conclude that clarity and consensus is unlikely unless we can ‘reset’ the debate, making a conscious decision to ‘step back’ from our extant paradigms/assumptions and approach the phenomenon with fresh eyes. I propose and demonstrate two methodological approaches that – used in conjunction – can contribute towards this end. Firstly, I propose that – and demonstrate how - Genealogical Analysis can aid in this ‘stepping back’ by denaturalising our entrenched assumptions on the causes of ideological violence (i.e., our extant paradigms) by uncovering how and why those assumptions came to be held and reified. Secondly, I propose and demonstrate Comparative Historical Analysis’ utility as a tool that can aid in re-approaching the phenomena with fresh eyes by helping - gradually and collaboratively - to construct a new set of more methodologically-rigorous assumptions (i.e., a new paradigm) upon which extant research built upon either extant paradigm can be resituated, reinterpreted, de-limited, and synthesised, and further research can be premised.