Except for crazed warmongers and those who profit from the war system, no one wants war, nor bloody attack. Thus, reason would suggest, we who wish to work toward a mitigation of such anthropogenic catastrophes would do well to pay attention to the research that suggests ways to manage conflict without bloodshed.
A popular idea is to arm one's self, one's protectors, and one's nation to the teeth, on the theory that doing so is the path to deterrence. The reasoning is that knowing that our enemies know that attacking us is going to lead to their own annhilation prevents them from doing so.
Fair enough. What if, however, there are other methods of deterrence, of assuring damage that doesn't kill or wound or even starve others? Would that also serve as a deterrent to any threatening adversary? What does the research tell us about this?
One interesting 2006 study by Jacob Bercovitch (New Zealand professor of International Relations) and Robert Trappl (head of the Austrian Research Institute for Artificial Intelligence in Vienna) concludes that conflicts that are shorter, that haven't yet produced many fatalities, and that are between parties of comparable power levels are far more amenable to mediation, that is, to nonviolent management.
This suggests that when we see conflict brewing, that is the time to jump in to promote discussion, before the positions are set in concrete. For many developing conflicts, this could lead to a good outcome. A sort of global conflict antennae detection system has been proposed since at least 1957, when it was theorized in the premier issue of the Journal of Conflict Resolution. Since we are so busy funding war and preparation for war, humanity has not gotten around to this yet, even though it would likely save enormous amounts of money and lives.
This research also would say that we might ask about the definition of power. If, for instance, an adversary were told that certain behaviors would result in the leadership of that party (nation, nation-state, insurgency, armed challenger group) being hurt--not physically, but in loss of income, resources, influence, ability to travel, possibly even freedom--that might change the power equation enough to produce a nonviolent negotiated path to better management of the conflict. In other words, threatening the leadership of, say, a genocidal government with incarceration, might work to slow and stop the genocide (unless, as in the example of Darfur, the African Union called for the International Criminal Court’s arrest warrants against Sudanese President Umar al-Bashir to be suspended while the continental body carries out a probe into alleged genocide in Darfur--thus dramatically ending the credibility of the International Criminal Court in Africa).
Or, in the case of deterring invasion, if the people of, say, the US were highly trained and committed to mass and virtually total noncooperation with any potential invading and occupying power, that would provide an interesting deterrent. Of course, for that threat to be effective, it would need to be credible, and our fat lifestyle and our civil society's sad history of failure to rise to the occasion (e.g. allowing George W. Bush and the US Supreme Court to blatantly steal the 2000 election) would make any potential invader scoff at such a threat. We would need to change our American way of life before that threat would hold any power.
But theoretically, we have all that power in civil society. Historically, we've seen the hints of that when civil society in South Africa decided to simply withhold cooperation from the apartheid government. As long as they were using insurgent violence with Umkonto we Siswe (the armed wing of the African National Congress, literally the Spear of the Nation), they were losing, the leaders (e.g. Nelson Mandela) stayed in prison, and the white minority ruled. When civil society stopped cooperating, apartheid crumbled. The same was true in India, when Gandhi called for hartal (cessation of societal activity, effectively a general strike) first in spring of 1919, and the British were terrified. Only when his own Indians committed violence did Gandhi call off the hartal. This was also the period when labor across the world was beginning to grasp that power and led to the formation of unions and collective bargaining. Many other examples, from Solidarity in Poland to the slowdowns in Chile as Pinochet's power was confronted, show that this power of withholding is power that might be organized for a substitute military.
Generating more research and making the real world examples manifest is a dual agenda for those who would like to see war begin to evaporate as a method of conflict management. The costs of war are so massive that no one has ever calculated them--the human lives, the national treasure, the environment, the reduction in all social services, the net loss of employment. Surely it's time to generate serious alternatives.
Bercovitch, J. and Trappl, R. (2006). Machine learning methods for better understanding, resolving, and preventing international conflict. In R. Trappl (Ed.) Programming for peace: Computer-aided methods for international conflict resolution and prevention. Dordrecht, The Netherlands: Springer. (Cited in Druckman, Daniel (2009). Doing conflict research through a multi-method lens. In Bercovitch, Jacob; Kremenyuk, Victor; & Zartman, I. William (Eds.). The Sage handbook of conflict. Thousand Oaks, CA: Sage. p.p. 119-142, p. 129).