There are two mtDNA and Y-Chromosome models used to explain the origin of Dravidian speakers in India. One model postulates a Pleistocene heritage for the Dravidian speakers based on the diversity of hapalogroup M, in India (Barnabas et al ,2005; Kivisild et al 2003; Thangaraj et al, 2006). Other researchers have claimed a Middle East origin for the Dravidian speaking people (Quintana-Murci et al, 2004; Renfrew,1996), eventhough Kivisild et al (1999a) has confirmed that there is a low-frequentcy of Eurasian mtDNA among Indians.

Although these origin models are popular in the literature there is another model for Dravidian origins that proposes that Dravidians originated in Africa and migrated to India during the Neolithic. This model is supported by archaeological, linguistic and mtDNA and Y-chromosome data.

African and Indian Skeletal Data
Using craniometric data researchers have made it clear that the Dravidian speakers of South India and the Indus valley were primarily related to the ancient Caspian or Mediterranean population (Gates,1961; Guha, 1936-37; Lahavory, 1963; Sastri, 1955; Schaffer,1954; Wheeler,1959; Winters,1985) . Lahovary (1963) and Sastri (1955) maintains that this population was unified over an extensive zone from Africa, across Eurasia into South India. Some researchers maintain that the Caspian civilization originated in East Africa ( Cole,1954; Lahovary, 1963).

Shared Archaeological Data
According to Sergent (1992), the Dravidian populations are not autochthonous to India either, but of African origin. The archaeological evidence also appears to support an African origin for the Dravidian speaking people (Lal,1963; Winters,1985).

Researchers have conclusively proven that the Dravidians are related to the C-group of Nubia (Lal, 1963), given the fact that both groups used 1) a common BRW (Sastri,1955; Lal, 1963); 2) a common burial complex incorporating megaliths and circular rock enclosures (Lahovary, 1963; Lal,1963; Sastri,1955) and 3) a common type of rock cut sepulcher (Lal, 1963; Lahovary,1963). The BRW industry diffused from Nubia, across West Asia into Rajastan, and thence to East Central and South India (Rao 1972:34).

Singh (1982) made it clear that he believes that the BRW radiated from Nubia through Mesopotamia and Iran southward into India. BRW is found at the lowest levels of Harappa and Lothal dating to 2400BC. T.B. Nayar in The problem of Dravidian Origins (1977) proved that the BRW of Harappa has affinities to predynastic Egyptian and West Asian pottery dating to the same time period.
After 1700 BC, with the end of the Harappan civilization spread BRW southward into the Chalcolithic culture of Malwa and Central India down to Northern Deccan and eastward into the Gangetic Basin. The BRW of the Malwa culture occupied the Tapi Valley Pravara Godavari and the Bhima Valleys. In addition we find that the pottery used by the at Gilund(2006), Rajasthan on the banks of the Bana River, was also BRW.

Archaeologists agree that Black and red ware (BRW) was unearthed on many South India sites are related to Dravidian speaking people. The BRW style has been found on the lower levels of Madurai and Tirukkampuliyur.



Haplogroup sharing between African and Dravidian Populations

Some researchers attempt to portray the Dravidians as Caucasoid people and try to link these people to western Eurasian people. Other researchers in India attempt to postulate an Indian origin for Dravidians because they mainly belong to the M haplogroup (HG) (Kivisild,1999a, 1999b).

Thangaraj et al (2006) recognize a Paleolithic origin of the M HG. The majority of Dravidian speaking people belong to the M haplogroup. Most geneticists agree that the M haplogroup is derived from L3. Kivisild et al (1999a) made it clear that all Indian mtDNA lineages "coalesce finally to the African L3a".

Metspalu (2005) argues that the earliest offshoots for L3, were HGs M and N developed in Arabia. Metspalu (2005) believes the MRCA for the M HG entered Asia 60-65 kya.

Metspalu (2005) maintains that "all the basal trunks of M, N and R have diversified in situ" (p.24). He makes it clear that in his opinion the M HGs are different from the subhaplogroup M of East Asia (Metspalu,2005). The most frequent HG in India is M2(Metspalu et al, 2004).

Sixty percent of of the Indian mtDNA lineages are M HGs (Thangaraj et al 2006). Kivisild et al (1999b) maintains that there are five M HGs in India: M1, M2,M3, M4, and M5. Thanaraj et al (2006) has revised the classification of HGs M3, M18 and M31 and defined the novel HG M41. Sun et al (2005) identified another 5 M HGs (M34-M40) in addition to the Indian mtDNA macrohaplogroup N.

The diversity of M HGs in India has led many researchers to suggest that the M clades have an in-situ origin (Metspalu et al, 2004; Thangaraj et al, 2006).These researchers speculate that although L3 originated in Africa, the M1 HG in Ethiopia and Egypt ,may be the result of a back migration to Africa from India (Metspalu et al, 2004; Thangaraj et al, 2006).

These researchers base this theory for a back migration to Africa from India, on 1) HG M1 is not found in India; and 2) the M HG's are only found in East Africa (Metspalu et al, 2004; Thangaraj et al, 2006). Both of these theories have little support when we look at the mtDNA data for Africa and India.

Barnabas et al (2005) noted that N,M and F lineages found in India could have originated in Africa (pp.13-14). He speculated these people migrated to India from Africa during the Upper Paleolithic.

Most researchers make it appear that the M1 haplogroup is only found in Ethiopia (Barnabas et al 20005; Metspalu et al, 2004; Thangaraj et al, 2006). These researchers maintain that the M1 HG is restricted to the Afro-Asiatic linguistic phylum (Metspalu et al, 2004). This is false M HGs are found in other parts of Africa where people speak non-Afro-Asiatic languages.

The M lineages are not found only in East Africa. Rosa et al (2004) found a low frequency of the M1 HG among West Africans who speak the Niger Congo languages, such as the Balanta-Djola. Gonzalez et al (2006) found N, M and M1 HGs among Niger-Congo speakers living in Cameroon, Senegambia and Guinea Bissau.

It is also not true that HG M1 is absent in India. Kivisild et al (1999b) found five M HGs in India: M1, M2, M3, M4 and M5. It is interesting to note that the M4 HG has the same 16311 coding region as the African M1 HG (Sun et al, 2005)

Kivisild et al(1999b) provides the first detailed discussion of the M subclusters in India and suggested an autochthonous development of these lineages in India. The researchers suggest that there were multiple M lineages when this haplogroup migrated to Asia (Kivisild et al, 1999). These researchers claimed that the expansion date for the five M subclusters expanded into India between 17,000-32,000 bp.

Kivisild et al (1999b) noted that 26 of the subjects in his study belonged to the M1 haplogroup. It clear from this study that sub-cluster M1 was found mainly in Kerala and Karnataka (Kivisild et al, 1999b). An interesting finding in the study was that most of the Indians with the M1 HG were members of upper caste.



Haplotype Sharing between Populations
The H1 haplotype is found among many Dravidians. Sengupta et al (2006) noted that the subclades H1 and H2 was found among 26% of the Dravidian speakers in their study, especially in Tamil Nadu. Ramana et al (2001) claims that the discovery of H1 and H2 haplotypes among the Siddis is a "signature" of their African ancestry. The frequency of the H1 subclade among Dravidian speakers is also and indicator of an African-Dravidian connection.

Another common haplotype in India is M2. This haplotype is also found in Africa. Luis el al (2004) observed that "There exists a west to east as well as a south –to-north clinal distribution with respect to E3a-M2. Bamileke and Benin display the highest frequencies (100% and 95.0% respectively)….(p.4). According to Luis et al (2004) the E2a-M2 is also present among East Africans in Rwanda, Kenya and Tanzania.

Semino et al (2002) noted that sedentary populations in Egypt like the Gurma fall into the M2 and M35 subclades. Semino et al (2002) found that 81% of Senegalese in his study fall into the M2 subclade.

In addition to haplotypes H1 and M2 in South India we also find the African 9-bp deletion. Watkins et al (1999) found the 9bp motif among four Indian tribal populations: Irula, Yanadi, Siddi and Maria Gond.

(A Clyde Winters Special)