In his masterpiece On the Origin of Species, Charles Darwin proposed that human groups and languages evolved in concert following a tree-like history of splits and isolation. But it turns out that the association between genes and languages is often transient. As Edward Sapir observed in 1921, "language moves down time in a current of its own making." Still, languages require human hosts, and geneticists routinely test for associations between genes, languages and geography.
The results are mixed. Gene trees can extend as far as one likes into the past. But language trees do not: unwritten languages change relatively quickly, and persist in a reconstructable form for at most a few thousand years. If populations remain isolated for centuries, as in the rugged mountains of the Caucasus or New Guinea, all three – genes, languages and geography – become strongly correlated. They co-evolve. Otherwise, population movements and language drift cause these associations to weaken.
Observing language use in human societies today is one thing, but reconstructing language use in the distant past would seem to be out of our reach without the use of a time machine. Linguists debate how long language families have persisted, but that is a different question from the social life of languages in the past. Can we learn anything about Sapir's currents of language in deep time?
Interestingly, the answer is yes. And the answer lies in the trees.
The key point is that at the community scale, genetic and linguistic trees become joined. Human communities are organized by kinship systems, which are themselves adaptations. Every living person speaks at least one language, invariably inherited from one or both parents. If people sometimes marry into villages where a different language is spoken, their children will learn the language of the new community. We can see this in paired gene-language trees, which record not only the language replacement, but the movement of the parent to a new community. By studying genetic and linguistic trees together, the recent movements of living people and their languages can be traced, but crucially, we can also see the travels of their distant ancestors as well.
If we perform this analysis for whole communities (and even better, for groups of villages), a fine-grained picture emerges of the interaction of these two symbolic systems, language and kinship, as they play out from one generation to the next against the backdrop of genetic inheritance. We can discover the basic principles of their kinship system, the histories of their languages, and the extent and nature of contacts between communities, extending deep into the past. But to see all this, we first need to understand how, where and why the trees are joined. The next two models explore these connections.
The results are mixed. Gene trees can extend as far as one likes into the past. But language trees do not: unwritten languages change relatively quickly, and persist in a reconstructable form for at most a few thousand years. If populations remain isolated for centuries, as in the rugged mountains of the Caucasus or New Guinea, all three – genes, languages and geography – become strongly correlated. They co-evolve. Otherwise, population movements and language drift cause these associations to weaken.
Observing language use in human societies today is one thing, but reconstructing language use in the distant past would seem to be out of our reach without the use of a time machine. Linguists debate how long language families have persisted, but that is a different question from the social life of languages in the past. Can we learn anything about Sapir's currents of language in deep time?
Interestingly, the answer is yes. And the answer lies in the trees.
The key point is that at the community scale, genetic and linguistic trees become joined. Human communities are organized by kinship systems, which are themselves adaptations. Every living person speaks at least one language, invariably inherited from one or both parents. If people sometimes marry into villages where a different language is spoken, their children will learn the language of the new community. We can see this in paired gene-language trees, which record not only the language replacement, but the movement of the parent to a new community. By studying genetic and linguistic trees together, the recent movements of living people and their languages can be traced, but crucially, we can also see the travels of their distant ancestors as well.
If we perform this analysis for whole communities (and even better, for groups of villages), a fine-grained picture emerges of the interaction of these two symbolic systems, language and kinship, as they play out from one generation to the next against the backdrop of genetic inheritance. We can discover the basic principles of their kinship system, the histories of their languages, and the extent and nature of contacts between communities, extending deep into the past. But to see all this, we first need to understand how, where and why the trees are joined. The next two models explore these connections.
References:
Sapir E. 1921. Language: An Introduction to the Study of Speech. Harcourt Brace & Company, p 160.
Sapir E. 1921. Language: An Introduction to the Study of Speech. Harcourt Brace & Company, p 160.