The question of whether Islam causes violence is a complicated one, because so many of the reasons for why human beings engage in violence emanate via shared universal tendencies. In fact, more often than not, we share such tendencies not only with our fellow human beings but also with our extended primate cousins.
EO Wilson, the founder of Sociobiology, remarked how many of such violent tendencies are context dependent and probabilistic. Hence, we can go from subdued and peaceful to bloodthirsty maniacs in a very short period of time so long as the right buttons are pushed, but even if the right buttons are pushed, it doesn’t mean that we absolutely will flip out, it just means that in such a context we would be more prone to flipping out.
For me, I am prone to violent outbursts when people cut me off in traffic, but I’ve never actually been caught up in a violent conflict because of it. Furthermore, on some occasions I don’t flip out, but my prefrontal cortex prevails over my amygdala and I stay calm, cool and collected. In other words, different parts of my brain battle it out and my ultimate behavior is the manifestation of which part wins. In cases where I flip out, violent conflict is more likely than in cases where I don’t (I’m not too worried that violence will ever happen, but it is more likely in when my middle finger is waving out the window).
So as it pertains to violence and Islam, the question is complicated by the fact that there are shared tendencies towards violence in all human groups, and by the fact that other variables matter too, such as group differences in behavior that exist apart from belief, political environment, proximity to mountains (which apparently increases the odds for conflict), culture, etc. But due to our shared tendencies, humans act according to somewhat predictable patterns of behavior depending on circumstances. I say “somewhat predictable” because the patterns of behavior are probabilistic, but still, under similar circumstances, Muslims, Christians and Jains will all act in similar ways.
But to me there is a difference between acknowledging this and engaging in a form of denialism that pretends as though belief has no bearing on behavior. This can’t possibly be the case, because human beings act in accordance with belief every day. If you see a giant chasm and you believe what you see, you walk around it (hopefully). So our behavior is not random and uncorrelated with what we say we believe. Instead, we tend to behave in accordance with how we see the world. Hence, imams will be imams, priests will be priests, and monks will be monks, day in and day out.
When it comes to overarching sets of beliefs, such as ideologies and/or religions, it seems likely to me that belief functions in the same way, such that they result in behavioral tendencies that are correlated with the belief. On some occasions, belief x might manifest in behavior y (peace), while on other occasions, belief x might manifest in behavior z (violence). But there are probably tendencies toward one manifestation over another, such that, even under similar circumstances, the Muslims will be Muslims, the Christians will be Christians, and the Jains will be Jains. The differences in how the groups manifest will often, though not always, come down to belief. The question as to whether one doctrine is prone to more violence than others, then, is a valid one that we should take seriously.