NECSI Salon: Ethnic Violence

On January 28th, the monthly salon gathered at NECSI to discuss ethnic violence from the lens of complex science. Yaneer Bar-Yam, president of NECSI, gave a brief talk about NECSI’s paper about modeling violence. Marshall Wallace, past director of the Listening Project, also gave a quick talk about his field experience with communities who opt out of violence. Again on Feb 4th, NECSI hosted an informal discussion around the case study of Libya. What follows are my big take aways and Sam’s asides, embedded into the fairly rough live notes from the salon. I call out these take aways and asides specifically because note takers often are lost in the notes, just as a photographer is never in the picture.
We hope you’ll join us on Wednesdays of this month to begin exploring medical systems, on ensuing fourth Wednesdays for structured discussion, or on other Wednesdays for more informal times.
Register for this fourth Wednesday here.

I am primarily left with a sense of purpose towards fostering collective intent towards alleviating suffering. In this entry, you’ll see a few ways large-scale violence is posited to be avoided. It is my personal opinion (of which I will opine at the end) that diversity is the key to equality as well as dignity, based on both the complex systems modeling and field experience framing these discussions.

But first, what do we even mean by “violence”? We’re referring to violent events occurring at level of massacres or bombing. These levels do seem to be slightly contextual based upon general violence levels in the area.

When doesn’t violence happen?

Violence doesn’t happen when

  1. all your neighbors are like you, or
  2. all of your neighbors are varied (integration of diversity).

The space between these is where difficulty lies, when not all of your neighbors are like you, but not so consistently unlike you that diversity is the norm. Well then, what do we mean by “neighbors”? It ends up this is very geographically based, and roughly the distance you can traverse in a day by foot or horse (20-40 km), which leads us to believe these tendencies might skew with new travel abilities. Running these models matched up to actualities in Yugoslavia, India.

Some examples

Can either impose integration or separation, intentionally or subconsciously:

  • Singapore forces integration, “to prevent sectarianism.”
  • Greece and Turkey offered to send each other their people.
  • NYC has small patch sizes geographically. But we don’t know how things scaled, the density of people.

A third way to avoid violence: boundaries

The distance you might walk in a day-if there are boundaries around this, unlikely to have violence. These boundaries can be either political or physical. The idea behind this is that if there’s a grouping of a people of a given size in a certain space, they start to impose their values on that space. Others coming through, or “encroaching” who don’t hold the same values would seem to be intruding in a way which violates those values.

If we have patches of that size, can we create peace? Tested the models in Switzerland, which both has diversity of values as well as a general lack of violence. It ends up the mountain rages act as boundaries. Where these boundaries are insufficient, as represented in the models, is where the people do enact violence. Switzerland’s response to these rare pockets was to place a canton, which reduced that violence. Cantons throughout Switzerland are thoroughly mixed by demographic or not at all, which coincides with no violence in these models.

Field Perspective

The clashes in India indicated by census-data overlay are also the area called the Red Corridor, where the Maoist insurgency takes action. That doesn’t show up on the census – and so doesn’t show up on the map models. Gujarat in the West has occasional flareups. There patch sizes may be 20-40km, a lot has intense mixing as well.  But there’s a lot of political effort to create separation. It’s intriguing to me that patch size [and equivalents, in other jargon] comes up again and again. Population of Kenya is really dispersed, but the patches of conflict is indeed on the 30/40 km range already referenced. In short, there are alternative explanations for what’s going on – which significantly match up with the models, which is interesting.

from Wikimedia Commons

from Wikimedia Commons

Creating boundaries

In Rwanda, there were many separations as well as integration. What was happening there? There were boundaries – were these simply in the wrong places? In India, Pakistan, Afghanistan, the borders are in the wrong places. The people who live there will tell you that. It’s nice to have models which indicate the same.

Creating new provinces: Kenya has also tried this. [aside: contrast w/ gerrymandering? -s] Power dynamics come out in these processes – people aren’t putting borders in right places. What are their intentions?

Lots of Marshall’s work is on the unintended consequences of delivering aid. Interventions often see the arbitrary boundaries which exist, and work with those boundaries. As an example, one hotbed of violence in Kenya had a mass migration. Those migrating clustered as the villages they had previously been in, but settled across a border between two states was. One state side was Anglican, the other side was Protestant. As the aid agencies only worked on one side of the border or the other, religious conflict emerged as one area simply got more attention and aid than the other. Religiously-aligned people previously uninvolved due to geographic distance got caught up in the conflict because of the way resources were brought into the space.

This relates so painfully to my experience in humanitarian and disaster response, where funding and support are silo’d arbitrarily or based on arbitrary policy, rather than on actual need and people. Marshall’s experience is in the extreme of how this can play out, and we must be cognoscente of it as we build tools and engage with communities lest we fall to the same fate.

Field and Model

YBY: the largest Swiss canton is also separated into circles, separated by religion.

In Rwanda before the violence, the Muslim population was about 10% of the total. The Muslim community as a whole stayed out of the genocide, didn’t kill others. The Hutu Muslims didn’t kill Tutsis, et cetera. Some people suggested that they didn’t participate because they had clustered themselves. This model would say those patches should have been targets for others… note there was also class-based violence within whatever else was happening. Maoists would tell you this is what’s happening there as well, not linguistic or religious conflict.
Are these patches the right size? Too small? Too far apart? People tell stories afterwards, but want to see the system.

Questions and Answers

Can be found where we took live notes, to be aware of attention spans and word counts.
My favorite bits include:

  • Can we us stories to change a zero-sum mentality to a winning™ mentality?
  • MLK talking to white america about their myths of fairness and justice and equality.
  • Breaking points, with examples from peace demonstrations that devolved into violence after a rock was thrown through a window, often by non-participants.


Breakout Sessions

Conflict patch analysis: what’s needed next?

Many times people make up differences if they don’t exist. So to what degree is this inescapable? How to identify the identity the factors that matter the most at a given time?
Again, more in-depth notes exist on our live notes, with top-level thoughts as:

  • How do we get to predictions or global scope?
  • How do we refine the model / what might be changed?
  • What other data is possible?
  • How do we prioritize where to look?

Failed States

Belgium as a stable failed state: little violence, garbage still gets picked up. Libya as an unstable failed state: lots of violence, no civic infrastructure. Often places in distress are able to ask for, and hopefully receive, outside assistance. Bolstering from the UN, outside trade support, etc. Besides this large-scale response, local individuals can also receive training and credibility from well-respected and known entities such as UNHCR, Red Cross, or academic or medical institutions; then return home to increase the quality of life and stability of an area. Groups like ISIS break these models, as any external entity (and representatives of such entities) are met with dangerously high levels of threat.

I’d like to point out that the Kurds seem to be doing pretty well (not to diminish their substantial losses) at responding to these sorts of issues, as a distributed group as well. We just don’t like to pay attention to them, as they’re feminist Anarchists.

Closing thoughts from salon

There are a lot of people working on peace and negotiation in Boston (where the Salon took place). We (NECSI) don’t know them yet, we need to start reaching out to them and offering to engage with them. We have a simple model that provides a deep lens, and would like to help move beyond the meme of “all just get along” to an understanding that this may not always work.

NECSI takes part in ongoing conversations with military groups. Do they talk about these issues – ethnic violence? Yes, with a focus on anchoring discussions in the sense that we’re really an integrated world now. The US is no longer in “out-group” conflict: we are an integrated system. [socioeconomic; also environmental ++] Military is vey forward thinking, compared to healthcare.

Strategists tackling complexity are often still constrained by policy. We still believe we’re looking at a whole system. The approach of blaming individuals versus governments in recent conflict is related to this shift, represented in part by things like not being able to work with refugee groups in Syria because they’re classified as terrorists. This also relates to how difficult things are with Dadaab, the Somalian refugee camp in Kenya, based on migratory patterns and tribal tensions.

People also talked about scales and stories. A single person can do a thing that matters, and that’s exemplified in both the model and from the field. Does that make this a universality class? In a similar thread, we wondered what can be done in actuality, within people’s comfort levels? How do we encourage people to understand complexity so it’s in their intuition when they make choices?

Maryam and I joined remotely, and had a discussion in our chat about if it’s ethical to separate people by imposing boundaries, or by opting for the “my neighbors should be like me” model over avoiding violence through going for “all my neighbors are different from me (and each other), and that’s ok!” model.
In the social sciences, there’s a strong advocating for integration because of the empathy it forces. People tend to self-segregate (this is called homophily, or “birds of a feather”) as it’s easier to be around people who already “get it.” But it’s also easier to avoid noticing how badly others might have it, if you don’t have to see them on a regular basis. See also gentrification, gated communities, slums, etc. When schools in the US were being racially integrated, that was the point that was made (and what proved true) – ends up “separate but equal” does not, in fact, work. Maryam pointed out that the schools are still gender specific in Iran, and that she fears this adds to tension and lack of understanding on all sides. So, while separating out genders, or races, or religions might make things safer in the short term, wouldn’t this lack of empathy and understanding make things less stable and healthy (read: equal, just, dignity for people) in the long run?

My lingering question is about how to maintain important cultural practices while also advocating for enforced empathy.