240 likes | 356 Views
New approaches to language and prehistory from typology, (genetics), and quantitative linguistics. S øren Wichmann MPI-EVA & Leiden University. Lecture II: More on the increasing importance of quantification.
E N D
New approaches to language andprehistory from typology, (genetics),and quantitative linguistics Søren Wichmann MPI-EVA & Leiden University
Lecture II: More on the increasing importance of quantification
Case study A: Using The World Atlas of Language Structures to make inferences about the peopling of the Americas(Not included in uploaded version)
Case study B: Distinguishing among farmer and hunter-gatherer language families
The language/farming dispersal hypothesis Human prehistory gives us a record of two very important, yet at first sight unrelated, examples of expansion. These are (a) the expansions of agricultural systems from hearth areas such as Southwest Asia, China, and Mesoamerica, and (b) the expansions of the world’s major language families. Some of the latter are of course associated with predominantly hunter-gatherer populations, but the majority occur in agricultural latitudes and their component languages are spoken by people who were already agriculturalists at the dawn of history. Many of these widespread agriculturalist language families, such as Austronesian, Indo-European, Niger-Congo, Uto-Aztecan, and Afroasiatic, had reached their precolonial geographical limits (give or take a few hundred kilometers) long before the local existence of any written records--their spreads belong among prehistoric farmers/pastoralists and small-scale social formations, rather than among the great conquest empires and charismatic world religions of history. Could the early dispersals of agriculture and the early spreads of certain major language families be linked effects of the same underlying set of causes? Do these causes relate to the demographic growth and rapid expansion profiles of early farmers? (Bellwood 2001: 182).
The policeman steps in. . . Campbell (2003): But not all extended language families are associated with agriculture. What about Tungusic, Uralic, Eskimo-Aleut, Pama-Nyungan, Salishan, Uto-Aztecan, Athabaskan, Algonquian, Siouan, Yuman, Chon, Jê? And not all small language families are spoken by hunter-gatherers. . . .
Bellwood gets nervous Bellwood’s answer to Campbell (2003: 468): “The immensity and complexity of the human past will always allow other hypotheses to exist, as it will also allow the existence of situations within which the hypothesis manifestly does not work. Critics of the hypothesis will always be able to rub their hands with glee as yet another non-matching situation is hauled out of the annals of archaeology or anthropology and paraded before an awed audience of non-believers.”
Can we save the theory? Maybe it is not the size or spread as such which is important, but rather its density, measured as the number of languages per amount of differentiation within the family.
Rationale Perhaps the rate of language change will be greater in a small community than in a large one. In a little band of hunter-gatherers it should be easier for individual innovations to perpetuate throughout the whole community than in a larger clusters of village inhabited by sedentary farmers (Nettle 1999). On the other hand, over time the population expansion will be greater among farmers (Golson 1982), causing a slow spread of the population over an increasingly large area. This, in turn, will lead to dialect differences and eventually the emergence of new languages.
Testing the hypothesis Let‘s mensure the density and see if the correlation works better. As a measure of number of languages, N, we simply use Ethnologue. As a measure of differenciation we use glottochronological age, mc, since this translates directly into lexical differentiation. D = N/mc (For practical reasons we calibrate D to a scale from 1 to 100)
The theory is now improved Using language family sizes (N‘): 70 ≺ N’≼ 100 agriculture present 1 ≺ N’ ≼ 70 no prediction possible N’ ≼ 1 agriculture absent A prediction can be made for 7 out of 36 families (19%). If we allow for one exception (Nilo-Saharan) it works in 35% of all cases. Using densities (D‘): 47 ≼ D’ ≼ 100 agriculture present 7 < D‘ < 47 no prediction possible 0 ≺ D’ ≼ 7 agriculture absent A prediction can be made (without exceptions) for 16 out of 36 families (44%).
It can even be extended somewhat The correlation narrows down to density in relation not necessarily to agriculture per se, but rather sedentism. Na-Dene and Algic data supports this. These are just below the cut-off value D’ = 47 that predicts the presence of agriculture. Driver (1969: 88): “fishing was more productive per acre than hunting or wild plant gathering. It was second only to agriculture in this respect. The relatively sedentary way of life on the Northwest Coast was made possible by the abundance of food available within a small territory.”
Not only have we saved Bellwood, we have also supported the initial hypotheses: (1) The rate of language change will be greater in a small community than in a large one (Nettle 1999). (2) Over time the population expansion will be greater among farmers (Golson 1982).
- Fin -Tomorrow: phylogenetic algorithms and software and they can be used in historical linguistics