Edition: Jonathan Cape, 1987
Review number: 399
It is very obvious that Romance languages such as Italian and Spanish are very similar, and the reason that this is the case is not hard to find - Roman domination leading to Latin dialects becoming the main languages spoken over much of south western Europe. It was only from the beginning of this century that scholars began to realise that many more languages were related to a lesser degree, covering most of Europe and, surprisingly, India - the Indo-European group of languages. This realisation immediately begs the question of what the reason for this might be, and this has been the subject of much speculation ever since.
By the thirties, the theory that became the consensus view was established. This was that there was a single race, the Indo-Europeans, which had, at some point in the prehistoric past, suddenly exploded from their homeland (thought to be in the Russian steppes) and established rule over a large area, changing the local language through a process known as "elite dominance". This theory rather unfortunately gained the attention of Adolf Hitler, and formed the justification (such as it was) for his view of the Aryans (from the name Aryas given to Indo-European speakers in the Sanskrit oral tradition in India) as a superior race.
By the seventies, the traditional view was strongly questioned, though the alternatives presented also seemed rather implausible. The problem is basically that the connections between the linguistic and archaeological evidence is tenuous at best, and often involves circular reasoning (the linguistic ideas are assumed when the archaeology is interpreted, and the results are then cited as evidence for the linguistic theory). Many of the arguments originally used to establish the theory are now considered simplistic, such as the assumption that a change in culture (in the archaeological sense of a distinctive style of surviving material goods) implies a change of language, and vice versa. In particular, no real evidence has been found of the destruction that would accompany a successful invasion of the type proposed.
Renfrew used this book to propose a new theory, one which seems a lot more convincing than those it sought to replace. Instead of elite dominance, which doesn't always change the language (think of India post independence, for example), he looks at other mechanisms by which the language of an area could change.
His theory is to do with the ways in which agriculture could well have spread in the early Neolithic period. Instead of conquest, this would have been more by infiltration as each successive generation created new fields a few miles beyond their parents'. As agriculture would have brought a vast increase in population density, the dominant language of a region would become the farmers', rather than that of the hunter gatherers they replaced. Pockets of non-Indo-European languages in Europe - the Basque still survives; others such as Etruscan were still spoken in historic times - mark places where the Mesolithic peoples learnt agriculture for themselves.
Renfrew puts forward many arguments to support his hypothesis, but the most telling is that it doesn't suffer from the problems of the standard theory. He says that it is untestable, but I suspect that useful evidence could perhaps today be obtained through DNA testing of prehistoric human remains.