This article is an experiment in treating simultaneously the evidence of language and the evidence of archeology to see if and how these may be brought to bear on common problems. It is not an attempt to solve problems but to compare some views of linguists and archeologists to see where areas of possible agreement and disagreement may be. It is hoped that these comparisons will evoke comment and criticism. Naturally linguists and archeologists alike are not in agreement on the interpretation of all of their respective materials. No effort has been made here to put forward any view as representing the consensus of all workers in any field. I am indebted to Professors Roman Jakobson and Joshua Whatmough of Harvard for much valuable advice in the preparation of this article and for criticizing various preliminary drafts of it. Professors F. N. Robinson and Vernam Hull of Harvard and Professor Kenneth Jackson of Edinburgh also contributed important help on Celtic. Professor C. F. C. Hawkes of Oxford and Dr. Marija Gimbutas gave great assistance on the archeological side, as did many other prehistorians who are mentioned in the appropriate places. But no one but the author is responsible for shortcomings.
It is of course clear that one cannot tell what language was spoken by an ancient illiterate people by the shape of their skulls, pots, or battle-axes (Hutchinson 1950). One has to have written words. It cannot even be insisted that one culture necessarily means one language. Many peoples speak two languages, and there are nations parts of which speak different languages. In the case of the Pueblo Indians, one culture is connected with several linguistic groups (Newman 1954). This is an extreme situation, but it serves as a warning that we are dealing with a highly speculative subject. Before exploring the prehistoric material, it would be useful to start with some historical analogies.
It cannot be insisted that an invasion necessarily imposes a new language. Persian, Greek, and Roman invaders brought new languages to Egypt, but Egyptians did not successively speak Persian, Greek, and Latin. It was only after Egyptian culture had been undermined by these events and by the introduction of Christianity that the Arab invaders of the early Middle Ages were able to turn Egypt into the Arabic-speaking country that it has been ever since.
Again, take the case of England and Wales. Here Celtic was the language of the Iron Age. The Roman military conquest introduced Latin, which became widely used, but Celtic was not by any means wiped out. Then in the fifth century A.D. there came the Germanic mass invasion that introduced English into the very part of the country where Latin was most used. English then replaced Latin, but, though Celtic is still spoken in Wales, Welsh speakers are now almost entirely bilingual. All . . .