A mix of three ancient haplogroups

Background to the Reimagined Mahabharata

First (Oldest) Previous  Next Top (Latest)

Genetics and Jaati

South Asia contains one of the most diverse populations in the world (possibly even more diverse than Africa).  Any discussion of the history of South Asia must address the reasons for this diversity and that makes it important to understand the genetics of the populations.  It is generally accepted that the diversity now is the similar to the diversity in 2000 B.C.E. -- the few anthropocentric measurements that have been made in the Indus valley show a range similar to what exists.  My contention is that this was true for the preceding 10,000 years and forms the background for the Reimagined Mahabharata.  This is a controversial position -- the controversy arises from claims about the origins of "Aryans" and "Dravidians", a racial/ethnic distinction that has driven expert opinion since the mid 1850s.  Aryans are supposed to have come in around 1500 B.C.E., destroying an existing urban Dravidian culture. The evidence for an Aryan invasion is scanty and requires that one discount local history and traditions as recorded in ancient Indian texts. Dravidians are supposed to have come in around 4000 B.C.E. -- this was based on physiological similarities between Dravidians and some pre-historic Mediterranean peoples (i.e., with even less cultural or local data than for the Aryan hypothesis).  Both these invasions supposedly occurred through the northwestern passes (the Khyber for one) through the Himalayas from Afghanistan..

Recent work by geneticists based on SNP correlation suggests something entirely different.  A better solution to the origins question is to push the most recent "invasion" back by another 10000 years or so.  The following excerpt builds to that.

Caste-based exclusivity in South Asia

I grew up as a child in Bombay (now Mumbai).  One of the things I remember vividly is my strong belief that I could determine the ethnicity of another Indian by their appearance.  I do not think that I came to this belief by myself – I think it was the general assumption, not only in my family, but in the surrounding community. I believed that I could identify the state that an individual was from and for some people in some states, the caste of the person. This was never tested with any rigor, and I no longer believe the strong version of this thesis, that I could do it for every state or ethnic group in India.  I certainly cannot do it now.  Coupled with a changing South Asian demographic, my own absence from India, and the creation of new states in India by carving up old states, it is simply no longer tenable. The weakest version is that I can recognize a South Asian; the next stronger belief is that I can identify broad “regions” (West, East, North, and South).

To a significant extent, the strong hypothesis confounds nature (i.e. genetic inheritance of traits) and culture (practices of dress, eating, behavior, and so on).  In Bombay (i.e., Mumbai of the 1960s), one could always identify a Sikh, for they were tall, wore a particular style of turban, had beards, were intimidating and scared most of the kids, but were considered extraordinarily honest, and drove taxis; I could recognize UP “bhayyas” for they wore white dhotis, a cap, had had round faces; a Sindhi home always smelled awful to my Tamizh nose (probably garlic, but I did not know that); and you could identify two Bengalis because as soon as they saw each other, they would start spouting Bengali rudely ignoring everybody else; the “Anglo-Indian” from Goa spoke “kasha-masha” English that was impossible to understand.  The list goes on. These cultural stereotypes probably overwhelmed any sharp distinctions that nature could have produced, but there were weaker distinctions to be made.

The reason of these distinctions in India can be credited to the millennia-old “caste” system. Jaati, as it is more correctly termed, divides the polity into endogamous units, supposedly based on the “purity” of their labor (inevitably, perhaps, the plush tasks score high on purity and the hard jobs that keep the system clean and healthy score low on purity).  People rarely married outside caste boundaries.  This is not to say that inter-caste sexual relationships, including marriage, did not occur – they were punished by a mix of religious and economic sanctions, which militated against expressions of interest  not leading to anything.  But the result is that  most Indian jaatis, (of which there are over four thousand) show, when analyzed with DNA-SNP analysis, a distinct founder effect dating from 100 to 200 generations ago. That is to say, Indian jaati endogamy began over 2000 years ago and maybe as late as 4000 years ago.

A hundred generations, and for that matter, two hundred generations are not enough for speciation – Indian jaatis are still fertile with each other (!) – but they are enough for some significant changes.  If the founder group was small enough and had some unusual genetic mutation, that would be more prevalent in that jaati than in the general population.  Some of these changes may be covert changes that are hard to detect in a deliberate manner, but recognizable unconsciously as a pattern.

The point is that by adopting the jaati system at some pre-historical point in the past, Indian civilization cast into stone the diversity of that era. The diversity of traits displayed by the various castes represent the diversity of traits that existed over two thousand years ago! Unlike Europe and China, where significant chunks of the population share a genetic heritage that can be traced to specific migratory events, the Indian population was already diverse at the beginning of recorded Indian history. If we are to find an explanation for the diversity of traits in South Asia, we have to begin with an explanation for these diverse traits being in place 2000-4000 years ago.

Sources of Diversity in South Asia

The analysis of correlation between SNPs between various ethnic and family groups has been used to build inheritance trees.  These attempt to explain how two groups split from a parent group and approximately when that might have happened (in terms of generations). When two populations meet and become one, it is possible to propose how these populations contributed to the pattern of SNPS in the current population – crudely, what proportion of traits were contributed by each of the parent populations.

Such analysis has been applied in a limited way to South Asian populations, across jaatis and within jaatis.  Much more work needs to be done and the work has to be extended to all jaatis if possible.  But based on current studies, the family tree for a subset of jaatis is best explained by positing an “Ancient North Indian” population and an “Ancient South Indian” population that merged.  Jaatis in the northwest tended to be as much as 90-95% ANI.  As one moves south and east towards the Deccan peninsula the percentage of ANI decreases and the percentage of ASI increases.  This continues south through the Deccan – unfortunately, this information is particularly scanty – and ends with the lowest ANI of 40% and ASI of about 60% in the deep south.

The source of the ASI and the ANI is not the proposed Dravidian invasion of 4000 B.C.E and the Aryan invasion of 1500 B.C.E.  If the Dravidian invasion was the source of the ASI, and it was followed by the Aryan source of the ANI, there would have been significant pockets of higher ASI percentage all over the north rather than in just one or two places. The ANI proportion being 95% in Afghanistan suggests that ALL north-western invasions of India came from groups that were ANI themselves (otherwise a 95% ANI would not happen). A recent invasion of ANI would have created a much more stepped transition from ANI to ASI, not the observed smooth transition from 95% ANI to 40% ANI.  Some of the oldest tribal populations in the south have substantial  ANI percentage.  

When the same comparisons are made with populations from East India (W. Bengal, Bangladesh, Assam, and the other Naga states), the split between ANI and ASI is harder to discern as one goes east – however, one finds a greater and greater admixture of variants similar to East Asian populations from Tiibet in the north and Burma in the east and other points further east.  I have not yet seen a report analyzing the east-west spread, though research has been in progress for some time.  This can be viewed as the source for an Ancient East Indian (AEI) which is not addressed by the  Dravidian-Aryan invasion theory.

I begin with the hypothesis that South Asian diversity derives from three ancient sources. (I've called them "haplotypes", though, technically they are not because we don't have a definite collection of SNPs to identify one, but have a derived factor in a multi-factor analysis of all SNPs -- but it makes the point that a certain pattern in the DNA continues to exist.) It is important to understand that the word “ancient” does not refer to 2000 BCE or even 4000 BCE.  All the genetic evidence indicates that the ANI is over 10,000 years old, the ASI is even older, probably as much as 50,000 years, and we do not have an estimate for an Ancient East Indian, if there is one.  The east-west change is sharper than the north-south change and therefore hints at a later time for the mingling of populations.  I hypothesize that migrations from Tibet started around 8000 B.C.E.

By 8000 BCE, the Indo-Gangetic plain and Deccan peninsula was already populated with descendants of these three lines – an ANI, an ASI, and an AEI.  The history of South Asia begins with the cultural evolution of these three populations.

Caveat emptor, for the thesis that three independent ancestral strands occupied the Indo-Gangetic plain is speculation. What follows is further speculation on how those three groups evolved, and their relations with each other before the crisis.

To be continued ... South Asia from 60,000 B.C.E, to 4000 B.C.E.