But, the genetic argument that I have made, which is to associate Y-DNA T with the Afro-Dravidians may be possible to show in a suitably designed autosomal study. But, since Y-DNA T is not uniform across Dravidian speakers (one presumes based on Y-DNA data that there was an initial demic expansion in the core proto-Dravidian area followed by expansions that is proportionately more culturally transmitted in the wider Dravidian area), in a way that goes beyond the ANI v. ASI cline, one would expect a search looking for Dravidian v. non-Dravidian components to have a hard time parsing out this subpart of a proto-Dravidian population.

Most autosomal studies have offered inputs in the form of linguistic groups or macrolinguistic groups, with an aim to discerning if autosomal genetic clusters form along those lines and they do. The autosomal data, while not unequivocal, does show a distinction between ANI and ASI. And, with enough clusters permitted (as Dienekes does in his recent analysis) almost every subpopulation can be distinguished from almost every other subpopulation as a distinct autosomal population that matched closely the input groupings without using a process that relies on the labels assigned to the input genomes. The hard part is to discern larger dimensional trends within an ASI component.

But, the fact that a population is distinct from other populations autosomally doesn’t tell you much about where its autosomal mix came from. Pure cluster analysis doesn’t itself distinguish between admixed and non-admixed populations, or determine how many components went into that population’s mix when (except in cases where recent merger of populations leaves mostly unadmixed as cryptic subpopulations).

To test the genetic part of an Afro-Dravidian hypothesis you would need to take a fine grained set of samples from populations that are Dravidian and are hypothesized to be Dravidian and look to see which autosomal markers, if any, corrolate to the prevalence of Y-DNA T in that population, and then compare those autosomal markers to autosomal markers that corrolate with Y-DNA T that is phylogenically similar to the Y-DNA T found in South Indian populations.

We certainly aren’t there yet and I don’t claim that this is anything more than a hypothesis. Y-DNA T was just recognized as a distinct haplogroup at all in the last few years, and I haven’t been able to locate any sources that I can read (i.e. that aren’t behind paywalls) that even do a phylogenic comparison of Y-DNA T subhaplogroups by geography (which could itself make or break the genetic part of the hypothesis), and since there are almost no pure Y-DNA T populations, providing pure types to use as comparisons in first order autosomal analysis, it isn’t easy work.

But, it is a hypothesis with considerable supporting evidence from multiple different disciplines including the Y-DNA studies, so it is worth the effort to investigate.

While there may be cases where autosomal studies reveal connections that patrilines and matrilines fail to reveal (for example, in cases where one instance of patriline dominated replacement by invaders to an existing population is followed by a second such instance that obliterates the Y-DNA traces of the first invasion wave), there shouldn’t be any cases where strong Y-DNA or mtDNA traces are entirely invisible in autosomal DNA evidence, even though the autosomal DNA evidence is much more formidable to work with in multiple loci at the same time (if you use individual allels and genes, it is only modestly more difficult). Autosomal data don’t make uniparental market data obsolete, it only adds nuance to it.

1) i followed some of your links. they’re old. that doesn’t invalidate them, but the genetic stuff i know has been superseded.

2) there’s no autosomal data to corroborate the hypothesis. in fact, lots of to refute it. that’s more for the people who read your comment. don’t care what you think about the issue, it’s pretty clear to me that the overwhelming preponderance of evidence weighs against you in the ‘afro-dravidian’ hypothesis.

I am inclined to agree based on linguistic evidence of time-depth and the close association between the range and age of Dravidian languages and the South Indian Neolithic that Dravidian emerged in the south of India with farmers (some of whom may have reverted to hunter-gathering), although I would put the date closer to 5,500 years ago in line with the archeology of the South Indian Neolithic. But, an origin for Dravidian in Northwest India seems unlikely.

The lack of expansion of the Harappans out of Northwest India to the rest of India has an obvious explanation. Near Eastern Neolithic founder crops grow in Northwest India but not in the climate of South India where African Sahel crops grow much better, and the South Indian Neolithic involved crops domesticated in the African Sahel, which arrived later in a separate event.

There is really no archaeological evidence to support anything more than a thin trading relationshp between the Harappans and the South Indian Neolithics. And, the case that there is a Dravidian substrate in early Rig Vedic Sanskrit is increasingly viewed as weak, despite the fact that if the Harappans or their ancestors were the first people subjected to Indo-Europeans from Central Asia and were also the source of Dravidian, there should have been such a substrate. Finally, the crops argue against that origin.

Indo-European expansion is also quite sufficient to explain the ANI genetic component in South Indian Dravidian population. The evidence that Dravidian farming practices and language have a source in a Niger-Kordifani speaking population of Sahel farmers in Africa, as argued by Sergent, increasingly looks more likely to me, and has support from crops, from linguistics and from cultural carryovers. Given the lack of an mtDNA trace of that event, I think it is likely that the seed population of the Dravidian culture was probably a group of colonizing men and that Y-DNA haplogroup T, a haplogroup whose distribution coincides neatly with the linguistically inferred location of proto-Dravidian, is the genetic trace of this seed population. See some supporting evidence cited here.