yes. african shows up in muslims quite often. clear if you look at the pakistanis individual in HGDP.

But what about the E. Asian component? History and geography can explain its presence in Razib’s parents and in Bengali individuals. According to Reich et al. South Asians can be modeled as a linear combination of ANI and ASI or in this model a linear combination of S. Asian and European. The ASI is quite distinct from E. Asian. Why is there some E. Asian in most individuals? Is this real or is it an artifact of the ADMIXTURE analysis?

first, reich et al. know that there are other minor elements. that’s why they discard outliers which have these minor elements to a great enough extent. second, i assume that it is noise in most of the cases. but some of the south indian samples (tribes) look way too admixed to be noise. the slave trade probably came from southeast asia too. in sri lanka malays have preserved their cultural identity, just as they have to some extent in south africa (less so), but probably this percolated in parts of south india. we aren’t surprised when we see west asian among kerala christians. we should be less surprised if we see some southeast asian in the tamil country.

But what about the E. Asian component? History and geography can explain its presence in Razib’s parents and in Bengali individuals. According to Reich et al. South Asians can be modeled as a linear combination of ANI and ASI or in this model a linear combination of S. Asian and European. The ASI is quite distinct from E. Asian. Why is there some E. Asian in most individuals? Is this real or is it an artifact of the ADMIXTURE analysis?

The Saurashtrians do not look like Tamil Artisan castes

What is happening is Sanskritisation, wherein a caste group outside Arya-varta, tries to raise its status by claiming the same caste name as a similar level-type of work caste group in Arya-varta

In South India, the dominant Y lineage, about 50% is L1

Vishwakarma is a Sanskritic word , refers to the artisan god Vishwakarma.

It is an attempt by South Indian Dravidian artisan castes to raise their status by Sanskritisation.

Similarly, several South Indian Dravidian shepherd castes have started to call themselves as Yadavas , another attempt at Sanskritisation, since Krishna was a Yadava

Similarly the Vanniyars, another Tamil Dravidian backward caste have attempted to Sanskritise themselves by calling themselves Vanniya Kula Kshatriya. Kshatriya being a Sanskrit word for the second ranking caste.

Similarly, the Tamil Chakkilis ( leather workers ) have Sanskritised themselves by calling themselves as Arundhatiyars after Arundhati, a minor Vedic goddess

(Oh and btw, Rajasuya is a royal coronation ritual.)

What would be interesting to know is if these Tamil Vishwakarmas had any genetic relationship to the Gujarati artisanal castes who also adopt the name Vishwakarma and have similar origin myths, claims to status etc. Do they have some common ancestry and separated at some point in the past or did they just coincidentally find the same scriptural peg to hang their geneological hats on?

“South Asian ethno-linguistic categories and hierarchies are notoriously Byzantine, and I have difficulty making sense of them.”

It would be very helpful to have some summary that strikes middle ground between the very course grained descriptions based on Indo-European v. Dravidian v. other, or on Brahmin v. non-Brahmin caste v. dalit/tribal, or Hindu v. Muslim v. other on one hand, and the “there are thousands of sub-castes and each village is different” desciption on the other. Is it possible to provide a meaningful explanation of ethnicity diversity in South Asia that looks at dozens or scores of categories, rather than just a few or many hundreds, that would be practical to apply?

It would also be very useful to see included in that middle ground distinction a greater sense of what these categories mean in daily life in some sort of narrative context.

I realize that this is far more than a blog post of effort. The kind of summary I envision might run to thirty to sixty pages with a few inset illustrations and maps for each group. But, one would think that this level of detail has been undertaken sometime by someone in a civilization of more than a billion people that has been in regular contact with the rest of the world for thousands of years, and perhaps even on the net given the large number of South Asian, English speaking computer wizzes that there are in the world.

*Sort of like the way every American seems to descend from a “Cherokee princess”. While sometimes this was true – Gokool Meah, as a famous example – often a “Kashmiri” great-grandparent was actually a Scottish overseer. Makes me wonder about my own allegedly Kashmiri great-great-grandfather.