Amygdala speaks: On algorithms for optimal distinctiveness

My fiery friend C, whose outbursts are poetic, recently wrote this on a social media site:

Another Brahmin acquaintance asked me today, ever so casually, if I were a Brahmin. This is the umpteenth time this has happened now. I am beyond immensely appalled at several micro issues this single question encompasses:
1. How 'easy' it is for a Brahmin to ask someone if they belong to the same caste as them. Imagine a person from any other caste asking this question. Go ahead, just imagine... I'll wait. What? You cannot even conjure up the hypothetical situation?
2. Why is it that a Brahmin needs validation? So is it like, if someone outsmarts them in any respect, they would be unbothered and at peace only if the other person also happens to be a Brahmin? Because anything else simply doesn't make any sense?
3. And why is it that anybody who is a Tamilian and a Vegetarian can only be a 'Tam-Brahm'? I am truly surprised at this existential generalization:
> A Tam-Brahm is a Tamilian & Vegetarian
> Therefore, All Tamilians & Vegetarians are Tam-Brahms!
So let me spell it out: I am a Tamilian. I am a Vegetarian. I am NOT a Brahmin. Thank you 🙏
Disclaimer: I don't practice Periyarism (at least not consciously) and I don't have anything against my Brahmin friends and acquaintances 🙏. Except for the above 3 issues

Since then, I have been wanting to write about this. As C points out, there are several issues at play here- I'm picking out a couple: One, stereotyping people. Second, a feeling of community and belonging that one group(Brahmins in this case) seems to develop for their own kind. While I originally intended to focus more on the stereotyping aspects of human behavior, I've always been curious about the latter too. Therefore, I've changed the outline of this article into putting forth an explanation for the latter and how it contributes to stereotyping.

Disclaimer: I write this from a psychology perspective- why we do what we do. I do not intend on conveying the opinion that the cause justifies the action and any resulting race/caste/language/community biased behavior.With all that out of the way, let me start this post.

Have you noticed the way people form groups or communities? It's mostly based on common background, shared quirks and interests. This kind of segregation is entirely logical in that human beings would want to associate more with people they can relate to. Yet, I've found this behavior immensely fascinating: Put a group of random people in a room, and you can watch this magic happen. It reminds me of the k means clustering algorithm demo they show in ML tutorials :P


First question: Why does this clustering happen?

Psychology has an answer ready. The phenomenon is called 'optimal distinctiveness'. Without getting into a lot of jargon, I can define it as the sweet spot where you belong and yet can be distinctive. Basically, it says that human beings have an innate desire to be unique. They also have an equal desire to belong and be accepted in some sort of a group. Now these are opposite forces: If you're too unique, you'd stand out so much that you'd never truly belong in any group-you'd become a loner and an outcast. On the other hand, if you belong and fit in perfectly well with everyone else, you lose out on your distinctiveness. Both these states are undesirable. Human beings want to have their cake and eat it too- therefore, these two forces have a tug of war till an optimum is established. We form groups which are distinct in terms of certain characteristics- this satisfies our craving for uniqueness while also making us feel included.
Fascinating, isn't it? I have no problems on this one-it's a beautiful theory that's true to reality.

Second question: What do we choose as the basis for this clustering?

If you closely notice this initial clustering, it's almost always based on region, language or community. At school/college, the first iteration would be something of this sort. People from a state would form a group and then start speaking their language. Anyone who fell into that bucket would be warmly welcomed into the group as 'their kind'. In a typical Indian scenario, you'd come across the 'North Indian' 'South Indian' divide. In a South Indian group, you'd again find language sub-groups. I'm sure there are equivalents in other cultures. And this feeling of community is so strong that people would go up to a stranger and say: Hey, you're a XYZ!(substitute language, region or caste) And they instantly go from stranger to being accepted into that group.


I just cannot tell you how frustrating I find this--it is one of my top grievances with human behavior!
Why on earth would you go and form a group based on any of these? Do people really think they would achieve optimal distinctiveness by language/caste based divides? It reminds me of those random starting points we choose for numerical methods. Choose a bad one, and the algo takes forever to converge. Yet, we human beings, who are supposed to be the pinnacle of evolution, choose such naive initial values when it comes to real-life.

And then we persist in sticking to these criteria our whole lives, trying to find more people who are 'our kind' just because they satisfy that one stupid condition on caste/region/language and ignoring everyone else. How on earth does that work? Do we really expect to find those uniquely common characteristics with these people? I definitely don't think so-- the gene pool in any community would be diverse enough to produce outliers that defy these stereotypes. Forget outliers, I don't even know if I subscribe to the idea that the majority of people in that community would obey the stereotype.

This might just be me, because most people find this entirely normal. I have discussed this with quite a few people, and they talk about how wonderful it is to find someone who speaks the same language as them in a foreign country, or about how they feel a warmth towards people of their city. I am one of those weirdos who entirely does not get this at all, and these attributes are just other data points for me. Sure it gives me something in common with people, but they seem too superficial to form the basis for clustering.

Do you wonder what I do differently? Let me tell you my algorithm: I start off assuming I belong with everyone or noone, and watch people from the fringes. I then find interests or personality traits that I can really relate to, and use that to form a group. These groups have been very rich in cultural diversity, and very fulfilling in that they give me an absolute sense of belonging and inclusiveness by feeding me those quirks that I care about. What I really enjoy is that mix of diversity and commonality- I do not want a group where we share too many things, and especially not those things that don't matter to me - language, caste, region, whatever. That would result in gene mutations and stunt diversity in a biological environment, and I believe the same holds true for socializing! :P To be honest though, maybe I wasn't always this way. It's a practice I have ingrained consciously, because I don't want to inadvertently discriminate against someone and contribute to an experience like my friend's.

We talk so much about inclusiveness and diversity, yet when it comes to actually applying these principles, we stand rigid in our archaic notions of what a community should be based upon. I am not entirely sure if I should be advocating this-because people do have a right to choose their groups on whatever basis they deem right. You could also point out to me that the basis I use for clustering is a different kind of discrimination. For example, if I'm choosing to socialise with someone witty, I'm discriminating against everyone who's not by excluding them from my circle. After all, they're as good as the next person, aren't they? I don't have a good answer to that. It just doesn't feel like blatant discrimination to me. Maybe someday, I'll have a more detailed justification to give you. :)

Would love to hear thoughts on this!