Institutionalizing Violence - the case of prison gangs

Since I'm doing my research one section at a time, the latter parts of my thesis are still subject to change. Right now, this is what I'm thinking.

I was originally going to do violence in general. That seemed too ambitious for such a short amount of time for research and, I'm still young, I can go there later. So, I decided to focus on prison violence. I wanted to look at the general forms of physical violence that have become "standards" in prison - rape and sexual abuse, officer beatings (both directions), etc. I thought I could take these from a general perspective and talk about how they became institutionalized. It's not that I thought this project impossible so much as disingenuous. Each of these has very different histories, serve different functions, and operate on different mechanisms. I still believe I can abstract enough away to encompass them into a single schema; however, this wouldn't really demonstrate the proper role of my theory of institutions. I think, if my theory is workable, that it's main role in explanation would not be for these general practices, but very specific practices with their own narratives. It's really a question of specificity - at what level of focus will my theory really be useful in effecting change. I don't believe that talking about general forms of prison violence will contribute much substantive help to resolving each of them (where they should be resolved).

What I want to do is to look at a more pressing and specific institution of violence in prisons - gang violence. Gang violence has become increasingly predominant in prisons over the past twenty-odd years as gang social structure was transplanted into prisons and gangs themselves became more violent. Applying my theory to this specific problem should produce a very grounded and detailed analysis of gangs and gang violence from which we can hopefully predict the emergence and development of gangs and find ways to prevent and deinstitutionalize their escalation to violence. This is one of the major thrusts that's so far sold me on this approach. I can give a very detailed analysis because I'm looking at something relatively specific and I believe I'll be able to offer germane suggestions for preventing and stopping this violence.

I lose the general analysis and the ability to transpose the analysis to other types of institutionalized violence; however, that's what theory is for. The theory itself should be clear enough that it can be taken from one instance of institutional behavior to the next. I shouldn't need to apply it to a general type of phenomenon to show that it can be applied to a range of cases. The application component, gang violence in this case, should merely show how the theory works empirically.

If I'm right in seeing the situation this way, then I've lost nothing in sharpening my focus.