Amartya Sen's "Identity and Violence"

We all agreed that the book was poorly written. However I think that it teased out some good ideas.

For one I agree with the premise that Sam Huntington does not know what he is talking about. First, his attempt at categorizing groups of people "civilizationally" is wrong. Second, but somewhat more importantly, the very approach to understanding human behavior by categorizing people into groups is flawed. Why?

1. Attempt at categorizing groups of people "civilizationally": I know that his basis for postulating a "Hindu" civilization is wrong. If he is wrong on one classification then he may be wrong on the others. The ignorance of some western (notice i do not capitalize the "w" -- this is merely a locational rather than an ideological classification) "scholars" is appalling. There is NO religion called "Hinduism" -- since all the other classifications are based on religion it is odd that he should choose one classification based on a non religion. This choice suggests a common (if ignorant) view of Hinduism as a religion with a shared belief system. Hinduism is more akin to a conversation that includes atheists and other approaches to solving the issue of existential dread. Doubt and discussion (sometimes argumentative and/or violent) is the only real thread of commonality among the various constituents resident in South Asia.

2. Understanding human behavior by categorizing people into groups: Huntington's methodological approach is fundamentally wrong as well. The approach stems from the desire to understand human behavior. A metatheory that explains behavior then would be useful not just as an interesting human exercise but also as a way of dealing with interactions between people and groups of people. Since humans are social animals (or beings if the word "animal" offends some religious notion) it is tempting to seek this understanding in social terms -- hence the focus on categorization. But contrary to the reams of psychobabble devoted to the notion of group behavior the main actors in the group are individuals. That is groups do not cause people to behave in a certain way. A group does not have a brain, or soul or whatever -- it cannot be an actor. Thus the whole focus on categorizing people into groups cannot yield any information on how people (and yes even groups) behave. The causal relationship must therefore run from individuals to groups. The valid questions are (a) why do people join groups? (b) How do people form groups? (c) How do these groups interact? -- in that order. Once we understand (a), (b) , and (c) we can begin to understand the concomitant policy, and therefore ethical, question -- Can we make this interaction "better"?

Incidentally the argument that it is impractical to focus on the individual as an unit of analysis (therefore requiring the focus on the group) is wrong -- the empirical testing of broad theory by focusing on the individual has a hallowed history in the science of statistics/econometrics and more recently in experimental and behavioral economics.