CULTURE AND I.ANCUACn 



have over the sounds both of their own and of foreign languages 

 affords evidence of this Hniitation. Education and booklearning do 

 their utmost in civiHzed countries to suppress the vagaries of dialect 

 and the impediments of individual speech. The general effect of 

 education, as we see it today, is not, as it might be, to discover 

 and exploit individual genetic properties, but to obliterate them 

 in uniformity. For this reason the unlettered yokel or infant provide 

 us with the best evidence, and it is very clear evidence, of the natural 

 and uncorrupted genetic differences between the speech capacities 

 of individuals and populations. 



Yet another type of evidence is available for the genetic control 

 of language. A phonetic map of any area of the world reveals 

 gradients or clines parallel with those we saw in the blood groups 

 and like them sometimes agreeing with, but more usually over- 

 riding, the ordinary linguistic or rather verbal barriers. For example 

 the fricative dental (TH and DH) has arisen only in the Eurasian 

 region. It is most strongly developed in Iceland, Spain, Greece, India 

 and Burma, that is, peripherally. Such an orderly distribution seems 

 to indicate relationship going back beyond neolithic times. And 

 it extends from Iceland to Ireland and England, from Spain to the 

 Basque country, and from Greece to Albania, from Southern India 

 to Burma, in each case uniting what philologists (and those anthro- 

 pologists who take their lead from language) regard as divergent 

 stocks. Moreover, there is a steady diminution as we pass from 

 Ireland to Wales, Wales to England, England to West Jutland and 

 so on to Zeeland and South Sweden where the TH sounds dis- 

 appeared 500 years ago. We thus have a cline which may be expressed 

 in both space and time. 



It is, indeed, possible to draw maps showing the isogenic lines both 

 for blood-gene frequencies and for phonetic capacities or preferences. 

 Both of these characters are as nearly neutral in regard to natural 

 selection as one can imagine, and we might, therefore, expect them 

 to reflect equally well the long-standing genetic properties of the 

 population. When we do so we fmd a remarkable agreement in 

 Europe between 64-5 per cent line for the O chromosomes (in 

 the A-B-O series) and the boundary between those who use TH 

 in speech today and those who do not. We also fmd an agree- 

 ment between both of them and our knowledge of the historical 



361 



