Dipankar Gupta has written:
In terms of empirical detail a lot more is required than has been provided so far if the theory of the racial origin of caste is to be made convincing. The factual evidence given in favour of this point of view is exceedingly exiguous. In fact, the manner in which dark skin and fair skin have been read into Vedic texts is itself highly disputable. What is interpreted as 'fair skin' in the Vedas could easily mean, and most probably did mean, 'light', in which case it was not a matter of skin complexion but of knowledge. The Aryans then distinguished themselves from others not by their complexion, but by their belief that they were in possession of superior knowledge and wisdom. They were the carriers of light, and that is how they dispelled the darkness and ignorance that reigned during pre-Aryan times. Further, there is only one passage in the Vedas that purportedly depicts the Dravidians as being 'nose-less and bull-lipped'. The Sanskritist, Hans Hock has convincingly demonstrated that this particular sentence in the Vedas has been translated and interpreted in a highly dubious fashion. The same word, viz., anas, can also mean a person of poor speech and not someone who is nose less.
If the Aryans indeed brought in a new language then it is only natural that they should emphasize proper speech and pronunciation (uccharan) to differentiate themselves from those others over whom they ruled, or with whom they had an uneasy relationship. The term bull-lipped can also have a variety of meanings. Remember, the bull is not really looked down upon in India as it is in many European metaphorsŠThe bull in Indian tradition is not seen as dumb and obdurate but strong and determined. So the evidence from the Vedas is far from conclusive. Moreover, what is most striking is that this description of the autochthonous original inhabitants is to be found in only a single passage in the Vedas, and yet so much has been made of it. It should also be noted in this connection that the term Varna in the Vedas need not necessarily mean skin colour. Varna can also refer to order. Therefore, if there were four Varnas, or two (as in the early Rg Veda), then it signified that the society was stratified along four orders, or two orders, as the case may be. Each order was supposed to have a colour pennant of its own as they represented different phases of the sun's journey around the earth.The rising sun, the grandest of all, was red, and this was the colour given to the ruling kshatriyas. Brahmans were signified by the colour white because that was supposed to be the colour of the sun at noon. Vaishyas were yellow because that is the colour the sun took in the East, and finally sudras were blue, for that was the hue of the setting sun. To extrapolate racial segregation from factual material of this order is indeed far-fetched. It will then have to be admitted that even the Vedas concede that kshatriyas are superior to brahmans as they represent the rising sun. Further, from where does the colour yellow get any material substantiation? Why have yellow or red not received any attention from those who argue in favour of the racial origins of caste? Why are many of us committed to a two race theory and not a four race one? Quite clearly the thesis that caste originated from race is flawed as it is based on flimsy evidence. The race argument takes a further beating when we study gene distribution and racial measurements along caste lines. Once again no clear pattern emerges between different castes on the basis of the distribution of heavy gamma chains and light kappa chains in their genetic makeup. It has also been pointed out in this context that the presence of African haplotype among some people of North India obviously means that the term 'African' is a misnomerŠ Much earlier in 1960, Majumdar and Rao conducted a statistical study of so-called race elements in Bengal and came to the interesting conclusion that there were overwhelming physical similarities between high and low castes within the same geographical region. But the story was different between different regions. Upper castes in one area differed a great degree from upper castes in a different geographical locale. The same held true between lower castes in different regions of the country. In an influential paper published in 1990 in Current Anthropology, an international team of scholars undertook anthropometric exercises and found no differences between different castes. They took three important measurements, viz., head length, head breadth, and bizygomatic breadth. After examining a wide range of material they came to the conclusion that all efforts 'at typological/"racial"classification should be abandoned'