The goal of this project is to build an empirically-based classification system, a taxonomy of violent conflicts, analogous to Kingdom, Phylum, Class, Order,… in botany and zoology. I am doing this to enhance the ability of peace researchers to identify the causes of violent conflicts and, more precisely, to make it possible to achieve the capability to do medium-term (6-24 months) conflict early warning. A paper I presented at the 1997 Peace Science Society Meeting describes the motivation and design of this effort to build a taxonomy of violent conflicts.
At the core of this taxonomy-generating project is the assembly of a dataset of all recorded violent conflicts in the world since 1400 AD in which more than 32 people were killed. I have been assembling and putting into my computer all compilations of violent conflicts that I can find. The version of this dataset that has been released to the public contains 3,708 violent conflicts.
With the data I have in my office or at least identified but that is not yet in my computer, I expect the tally will exceed 5,000 but is not likely to reach 6,000. What that means is that on average, for each year since 1400 AD, approximately 8 to 10 new conflicts erupted each year. However, one thing I have found during this project is that the actual number has varied significantly over the six centuries. In particular, the 18th century was markedly less conflictual for a number of widely separated regions. I find that of interest because the existence of a global-scale phenomenon almost requires that the cause or explanation for that phenomenon be global in scale as well.
At this point in time I am putting into the dataset information about who, when, and where for each conflict and, whenever it can be ascertained, the number of military and civilian fatalities for each conflict. Future research will attempt to identify 19 additional characteristics (variables) about each conflict. With a dataset of conflicts and the collected information about each of them, I will then be able to use cluster analysis techniques to group the conflicts by their shared characteristics and with that information develop the taxonomy.
One thing I have been trying to do with this project is get beyond the Eurocentric bias of most existing compilations by accessing sources not in West European languages or not published in a North American or European press. I have found a few sources of that nature (Japanese, Chinese, Turkish, and Russian), but I am sure that more exist. If anyone reading this knows of a compilation that I might be able to fruitfully use, please let me know.