As I have already pointed out Hart just does not want to go beyond Tamil region to compare the similar the cultural aspects. Of course, he goes beyond Tamil region but only to cover South India when it fits to his argument. Consider his footnote on menstrual blood taboo.
Though the segregation of women during their menstrual periods must have been
universal in South India during the time of the anthologies (as it is found in all South Indian cultures today and in medieval times),
Menstrual blood taboo is part and parcel of all West Asian derived pre-Christian religions like Judaism, Yazidi, Zoroastrianism and Caste system. I still believe ritualized menstrual blood taboo was first codified in West Asia though people living in present day India might have practicing it before entry of Indo-Turks (Y-Hg-R1a1 people also known as Indo-Europeans in popular literature) and Aryans (from Iranians/Caucasian lands). As Y-Hg-J2b appears to be older than any of those migrations.
Yazidis in Georgia have haplogroups K*,F*,P*,R2 and J2*. Though some have speculated Indo-Turk influence in Yazidi religion I rather believe it is built upon old West Asian beliefs. Lack of R1a1 in this highly endogamous community with caste like structure may show the oldest haplogroup distribution in that region before the advance of R1a1. But we need to check whether that J2* is entirely J2a or mix of J2a and J2b or J2b alone.
Now consider the exact word, pole, used by Hart to put forward his arguments.
This is what I found at Hobson-Jobson's dictionary;
HULLIA, s. Canarese Holeya; the same as Polea (pulayan) (q.v.), equivalent to Pariah (q.v.). ["Holeyas field-labourers and agrestic serfs of S. Canara; Pulayan being the Malayālam and Paraiyan the Tamil form of the same word. Brahmans derive it from hole, 'pollution'; others from hola, 'land' or 'soil,' as being thought to be autochthones" (Sturrock, Man. of S. Canara, i. 173). The last derivation is accepted in the Madras Gloss. For an illustration of these people, see Richter, Man. of Coorg, 112.]
I have already argued that we should restrict ourselves to the exact meaning of the word. In one of previous posts I mentioned about Mala in Andhra Pradesh equivalent to Holeya in Karnataka. In my opinion, the idea behind their designation must be the same. Hence I proposed it could have been derived from Sanskrit malIna meaning 'polluting'. Recently I was reading "At the botton of Indian society" (author: Stephen Fuchs). There Fuchs says;
An old Telugu dictionary derives Mala from maila, dirty. This sounds more appopripriate, for the Hindus consider the Malas as ritually impure
Now let us examine the relationship between pole meaning pollution and 'pole' meaning impurity due to menstrual blood and child birth. In Tulu, the word for former is pole and the word for the latter is pile. In Tulu too the word for the outcast is Poleya or Holeya showing the connection with the first word. But I am not in a strong ground here. According to Ravi Mundkur at Tulu Research, the word pole meaning pollution does not exist. I am following this database. Let us say pile is derived from pole. But that certainly shows the outcast Poleya has much to do with pollution than the menstrual blood (the uncontrolled thing that characterizes them). If there any strong correlation between menstrual blood and outcastyism is existed then it must have appeared in Tulu outcasts' identity. It should have been pileya and not poleya.
Polemics and Red herrings:
There is this idea going around that Brahmins' position was low in the early Tamil society. I wonder if Hart is responsible for that. He discusses about 'drum' the source of power and quotes;
...at that pindam [ball of sacrificial food] hard to get raised by the high one [uyarnton] to supplicate the god [in the drum, according to the old
commentary] wont to be hard in his might with mantras [mantiram] roaring
[i.e. shouted] out.
Then he discusses about the identity of 'high one' mentioned in that poem. He concludes that it must be a drummer whose status is higher than other drummers. In the run of the argument he discards Brahmins being high ones with an unqualified statement thus;
Brahmins are generally called Antanan and Parppan, not “high ones.”.
So? Was a Paraiyan drummer ever called 'high one' until Hart zeroed in on him? Who qualified most for that at that time? Who do you associate most with 'pinda' and 'mantra'(hymn)? Anyway, I too talk about likely scenario therefore not really a strong point.
I still don't understand the necessity of this following paragraph;
If one uses purely economic criteria, then the high position and
influence of the Brahmins in later times must be mysterious, for it is not
clear that they had any tangible economic commodity or skill to offer.
Brahmins never really lived outside the village even when Hart believes they weren't at the top of the society. That argument reminded me something I read few years back. Those were words from an African though I don't remember the exact words and am writing them from my memory.
When the Whites came we hand the land and they had the Bible. They asked us to close our eyes and pray and said miracle would happen. When we opened our eyes miracle did happen. We had the Bible and they had the land.
Of course, Hart wants Brahmin to be non-high in the early Tamil society at any cost. He needs existence of caste in Tamil society based on certain other aspect (here controlled/uncontrolled power). Even when all his arguments are likelies/pluasiblies he makes a unequivocal statement in the end based on them.
As important as the Brahmins and the Brahmanical religion were, they were not the creators of the caste system in South India. They influenced the system profoundly, no doubt, but caste is found in most of its manifestations before the Brahmins became prominent.
Controlled and uncontrolled power:
Hart's arguments are centred around controlled and uncontrolled power. There were communities not really directly involved with possession and menstrual blood. These groups like leather workers and fishermen were confined to lower rungs all over India. But Hart tries to equate them with ghostistic groups like Pulaiyan, Panan etc...
The fish eating taboo could be found in Arab regions too. According to Stephen Fuchs Bedouin tribes considered fish a low food and higher classes abstained from it. They did not marry with the communities that ate fish or involved in fish catching. I believe Bedouins might have still kept that old West Asian taboo about fish that the Aryans from that region brought to India. I have come across myths of losing once high caste status involves fish eating or get caught with fish in South India. Fish and only fish no other meat.
Konkani and Bengali Brahmin communities do eat fish. Among them there are sections that do not. What we need is proper scrutiny of their haplogroup distribution. I have observed Bengali Brahmins do not have any J2a, a West Asian lineage. They are predominantly Indo-Turks(R1a1).
From Sengupta et al (2006) study.
R1a1-72%, R2=22% and H=6%
In my opinion, an objective study must keep clear of all kinds positive and negative qualifications. I found couple of them unwarranted in this paper.
For this purpose, they could offer to the king their whole tradition of sacrifice and magic, already developed for almost a millenium and extremely impressive.
In other words, the function of the untouchables no longer included serving kings—a function that must have enhanced their low status—but was limited to such mundane and undignified tasks as cutting hair, washing clothes, and catching fish.
Tulu society has never really moved away from its ghostistic traditions (probably true in the case of Kodava and north Malabar societies too). But even in that society I don't believe position of Brahmin and that of Poleya was different from the Tamil society in any degree in the past. The ghost possessor can be from any non-brahmin caste in the case of Kodava and Malayali society. But in Tulu society sometimes even Brahmin took up that role.
When you look at Tulu society then in my opinion there are two things that we can observe. There were occupational groups like leather workers that were deemed untouchable or low ones (fishermen). Then there are communities that were considered low like Holeya. A tribe can become an untouchable group as it moves closer to caste society(Koragas in Tulu region). However, Holeyas have been mainstream (if the dating of anthologies are correct) for nearly 2000 years now. That is probably for the reasons related to economics (as Holeyas were slave agricultural labourers in the past in South India).
But we should also note that people involved in ghost possession or shamanic traditions were not in high position in many societies (Siberian communities). Ghost possession or experience in my opinion is generally restricted to malnourished, illiterate classes of the society (and probably women). There can be no comparison of them with literate godistic priests.
Typical to societies around the world, South India too might have developed a society with masters and slaves. When Varnashrama was applied to the society I do not think slaves could be part of zUdras as the dominant classes themselves were grouped under them. It is more likely that Holeyas because of their controlled status(by landlords) than their uncontrolled status(by Hart) were relegated to untouchable status.
Early Evidence for Caste in South India
1. At the Bottom of Indian Society, Stephen Fuchs
2. MtDNA and Y-chromosome Variation in Kurdish Groups, Nasidze et al. 2005