Reading maps another way

In my struggle to understand the thirty years' conflict in Sri Lanka it becomes more and more apparent I have to read maps a new way. 

At a macro level, the map of abandoned irrigation tanks shows the regions where war was its most destructive. Another map, at a smaller level, shows "Section C" of the Mahaveli Project, where contemporary tank landscapes were built to accommodate Sinhalese settlers in a Tamil region. This may not be a "war map" but it does frame the ethnic conflict. 

A less tangible map may include a region of the island (the Tamil east) where my Sinhalese colleague refused to bring his students. He's trained his students to lie about the conflict so he may have chosen to "forget" about his commitment to meet our mutual colleague and her students out there. 

Another map, a bit more tangible, may show confiscated houses along the Batticaloa-Trincomalee Road, each house the small locus of a military checkpoint now inactive but still present. 

Can I venture smaller-scale maps where people were kidnapped, apprehended, burnt to death, places where memory may not take us for the pain? 

And what about islands of safety. How small were these "safe havens" and how porous? Did they change in nature so that so that it's their unpredictability we might map?

It was my friend Gihan who suggested I could map Sri Lankan tank landscapes. Neither of us, I'm sure, thought the project could move in this direction.