Records are preserved because they provide a charter for what historians believe about the present. Different historians may believe different things… - Edmund Leach

At one time social anthropologists used to complain that their archeologist colleagues had no sense of the overall coherence of human societies. Now, under the influence of Dumézil, who was himself in­ fluenced by Durkheim and Granet, most of the prehistorians who have specialized in India and Pakistan and most of their Indo- Europeanist philological colleagues have become committed to a functionalism of a wholly naive sort. They seem to assume that cultural systems and language systems are bonded together and intrinsically stable over long periods of time. If societies are left alone, they stay put; otherwise, they roll across the landscape like impermeable billiard balls. If the archeological record shows that in fact changes have occurred, their occurrence is always explained as the consequence of a movement of population that carries with it the products (both material and immaterial) of a preexisting, alien, self-contained culture. As a rule, the alleged movement of people takes the form of a military conquest. The mythology of the Dorian invaders of ancient Greece who reduced their Ionian predecessors to serfdom matches point for point the mythology of the Aryan invasion of northern India.

The Allchins, in their archeological capacity, have consistently emphasized the continuity that links the residues of the Indus civilization with those of the later classical India in the Ganges basin and further south. Furthermore, they repeatedly emphasized that archeology provides no clear evidence of any mass movement of peoples from Central Asia into northern India. So why do they continue to pay deference to the “racist” notions of nineteenth-century philologists in this way? (Incidentally, there is no “general agreement that the Indo- Iranian languages . . . were originally spoken in the steppes of Eurasia”)

The details of the theory fit in with this racist framework. Just as each member of the total family of Indo-European languages is lineally descended from one or another of a number of extinct “protolanguages,” so also are the speakers of these languages; hence the people who speak any particular language constitute an independent racial stock.