The concept of human races has often been misused, and the whole issue has therefore come under intense criticism. Though it is unlikely that there exist populations of humans that have been reproductively isolated for long enough to have diversified to the same extent as races of other biological organisms, many genetic traits do show geographical (and demographical) distributions demonstrating historical endogamical traits. When the lines separating many of these characters roughly coincide, it is useful to refer to the common borders as race divisions.
Social Stratification
The majority (about 80%) of Indian society is broken up into about 2000 castes which can be further broken down into endogamous units which are called subcastes, the total number of these units in India is estimated to have been 75000 at its peak, and still about 43000. Any genetic study needs to take these into account, as well as the `gotra's (roughly speaking, exogamous lineages) within these. A preliminary study showed that about 87% of the subcastes were strictly endogamous, and about 5% allowed `anuloma' (hypergamous: woman marrying socially above herself) marriages. In almost all cases, the society is patrilocal (and patriarchal), and caste follows the father.
In addition to the Hindus who belong to the caste system, Moslems, and other established religious groups, the rest of the mainland Indian population, about 7%, is tribal in nature; 427 separate tribes are officially recognized. These are usually divided into the tribes of the Himalayan, Middle, Western and Southern India; Bengal has tribes belonging to the first two groups. Knowledge of the exact affinities of the tribal with Australian or African Negrito populations would be interesting, but investigations have not been able to uncover direct genetic affinities except for known or suspected cases of recent contact.
The castes can roughly be broken down into the upper castes, the middle castes, and the lower castes based on their social status. The division, especially between the upper and middle castes, varies according to region. Genetic evidence points to the different social groups and geographical regions having different set of Y chromosomes, pointing to the social identity staying unchanged along male lines; similar data from mtDNA studies show only slightly higher admixture in the female lines. However, the statistical signficance of all these results is weak except to support a tribal versus caste hindu divide.
Historical Migrations
A study of the skulls from ancient South Asia showed presence of three groups of people. Most of the hunter-gatherer skulls from South Asia clustered with upper paleolithic skulls from Europe. The skulls from Harappa were sharply distinct from these.
The foraging and farming populations of South and Central India are intermediate between the other two.
In India, from anthropometric studies, one used to find traces of seven races of humans who intermixed to create the Indian race. Modern studies within regional contexts are still rare; so one mostly has to look back to the global genetic studies. Northwest India shares with west Asia and eastern Europe (and pockets in Africa and South East Asia) the maximum heterozygosity known among world populations, with means between 0.35 and 0.37; and the rest of India (and Europe) is only slightly lower: 0.33 to 0.35. This shows the vast amount of admixture that has gone on in these regions: to be contrasted against Australia which has a homozygosity of less than 0.25. It is currently accepted that at least four strata are visible in the populations in different parts of India.
An australoid-veddoid substratum
A migration from the east of Austrasiatic and sino-tibetan language speaking groups.
Neolithic migrations from western Iran, probably proto-Dravidian.
The aryan expansion from north of Caspian sea via Turkmenia and Northern Iran.
Thus, for example, some researchers have concluded that the most likely synsthesis of different lines of evidence is that the Austric language speakers came to India c. 50-65 Ka BP from the northeast. the Dravidian speakers c. 8-4000 BC from the mideast with knowledge of wheat cultivation and cattle, sheep, and goat domestication (all middle eastern developments around 8000 BC),
the Indo-europeans in several waves since 4000-1500 BC with horses (domesticated c. 4000 BC around Ukraine; appears to move from northwest India in about 1900 BC to southeast India in 100 BC) and/or iron (used around 3000 BC in Anatolia; also appears to move from North West India in 900 BC to South East India in 400 BC; iron and horses were almost certainly distinct cultural traits which were not associated with one another), they had distinctive burial styles and may have performed cremation, the painted grey ware pottery associated with these people fits the iron users more than the horse riders; and in this mix, the Sino-Tibetans joined in in several waves since 8-6 Ka years BP bringing in rice cultivation (if it is not of separate origin in the Indian region, it may have started in south-east Asia around 8 Ka BP).
DNA evidence
mtDNA
Early mitochondrial DNA (which is maternally inherited) studies indicated that a vast number of Indian mtDNA lineages cluster with the East Asians, probably reflecting the proto-Dravidian. They also seem to be closely related to African (e.g. Ethiopian) populations, probably indicative of the Australoid-Veddoid substratum.
Modern studies indicate that the major mtDNA lineages in India belong to the typically asian M haplogroup, whose Indian variety probably originated around 48000 +/- 1500 years before present (i.e. about 46000 BC). This haplogroup shows no statistically significant linkage with caste as a whole. The lineages in this haplogroup do not segregate according to linguistic family, but some specifically Indian lineages (e.g. M3) correlate with the upper castes.
The second most common haplotype, the U2i, separated from an West Eurasian lineage around 53000 +/- 4000 before present (i.e. about 51000 BC). This one is strongly correlated with caste; the upper castes having these in the highest proportion. A small fraction of the Indian population (about 5-10%) belongs to lineages (W,H,K in upper castes; J,T in other castes) also common in Europe, and which have more recent divergence dates. They probably have caste linkages, but the data set is, as yet, too small to be definite. Also, the divergence times have not been estimated, so it is difficult to pinpoint which migration this refers to.
Previous research, which had not detected the caste linkage of the European haplotypes had concluded that, assuming they were largely of Western Eurasian origin (e.g. in accord with the Dravidian-protoElamite or the Indo-european hypothesis), the divergence time is about 9300 +/- 3000 BP (i.e. about 7300 BC), which is deduced as an average over various number of unknown founders (i.e. gradual migration model, rather than concentrated invasion model). Some minor geographical gradients from the Punjab to the Andhra in distribution of European haplotypes also needs further study.
Y chromosome
The Y chromosome (which is paternally inherited) data is broadly similar. India groups clearly with the South/Southeast Asia cluster. The major European haplotype is pretty rare, but some European haplotypes are found amongst the upper castes, and in Punjab and Pakistan. A North African/Middle Eastern variety is found at low levels all over India, but many of the Indian haplotypes seem to be of Indian origin, possibly due to genetic drift in small endogamous units. The interesting aspect is the much stronger caste linkage in the genetic distance between the Indian and European populations found in these studies compared to those dealing with the maternally inherited mtDNA, though control over statistical and systematic errors is still lacking; as also estimates of divergence times.
Autosomal DNA
A similar caste linkage is also found in the autosomal studies: overall upper caste Indians are significantly closer to Europeans than other Indians. However, divergence times estimates are still lacking.
Physical Anthropological evidence
Physical anthropological studies very clearly separates the Indian populations (except some Andamanese tribals) from most of the other Asian populations, with people from the persian gulf, Arabia, Burma, SW China, Vietnamese and Malayan forming the border along the first principle coordinate. This component seems to be highly correlated with stature, and hence with temperature. The second principle coordinate, which mainly measures facial and nasal shape, serves to separate the various groups of Indians: (i) Maharashtra upper caste, (ii) Gujarat and Konkan, (iii) West tribal and lower caste, (iv) Central and Eastern tribal, (v) Bihar and Bengal, (vi) Vedda and South Deccan tribal, (vii) Orissa upper caste, (viii) UP upper caste, (ix) Pahari bhotia, (x) South India and Ceylon and (xi) Kashmiri, Punjabi and Pahari. The third component clarifies the separation of (i)-(iv), (v)-(vii) and (ix) out of the rest.
In a limited genetic tree, Indians form a number of distinct clusters: (i) Central Indian and Brahmins, together with South Indians cluster with Westen Asians (cluster formed by Iranians, Uzbeks, Caucasians, Lebanese and Turkish, Jordanian, Assyrian, Armenians), (ii) Sri Lankans and South Dravidians break off earlier, and (iii) North and Central Dravidians along with the East Indians break off before the cluster formed by the previous two groups and the Arabians including Bedouins. The first two principle components do not separate the Indian population: they only separate the caucasoid group discussed so far, from the South East Asians with Gurkhas, and from the North East and East Asians with the Bhutanese.
When only the Indian populations are analyzed, the branching order changes somewhat. Now, the third of the Indian clusters mentioned is an inner group and consists of a cluster of Munda and North Dravidians with Central Dravidians and a separate one of Marathan and Maharashtrian Brahmins with Bhils and then Rajbanshis, with a cluster consisting of Bengali Brhamins and Parsis splitting off its base. A cluster consisting of the second and most of the first of the previous groups is sister to it. (The first group, of course, is now resolved: Punjabi, Central Indian form a group with Punjabi Brahmins and the Rajputs; Vania and Jats form a sister cluster with Bombay Brahmins. Koli and Kerala Brahmins along with Pakistanis form a cluster off the base of all this.) Kanet and UP Brahmins seem to form a cluster branching off earlier, and Gurkha and Tharu a cluster even earlier. The Kerala Kadar seems to come from a completely different branch.
The detailed structure of the clusters probably needs more data to be established. As far as Bengal is concerned, the tribals cluster with North and Central Dravidians; and Bengalis as a whole cluster with these and Maharashtrians. However, the importance of caste shows up; except in Punjab and Maharashtra, the Brahmins do not cluster with the other castes. On the other hand, the Brahmins of different regions do not cluster together either; in a two principle component analysis, brahmins from Bombay, Kerala, UP, Punjab, Maharashtra and West Bengal show a steady progression along the diagonal. This could be due to genetic drift in these highly endogamous units.
http://members.tripod.com/~tanmoy/bengal/races.html
Bookmarks