When two communities live apart and communicate little with each other, over time some innovations tend to differentiate the respective linguistic varieties until two systematically distinguishable dialects are formed. It is also known that speech varies from generation to generation, and certain isoglosses have a different frequency in different age cohorts. This is the informal description usually given for new dialect emergence.

But, are there any studies or models that attempt to quantify the weight of factors? For example, could the presence in antiquity of mountain ranges or rivers be quantified with respect to linguistic diversity? How can we measure the degree of interaction in different linguistic communities and how can we study the numerical correlation between that degree of interaction and the degree of linguistic divergence?

  • In fact, there should be plenty of opportunities to study such questions, Modern Hebrew, Dutch in Flevoland, or the regions of Poland and the Czech republic that were resettled after ethnic cleansing from German speaking people should be regions for such kind of research. Commented Jun 6, 2022 at 10:53
  • @jk-ReinstateMonica At least in the Czech Republic, the border resettled regions are quite mixed, although mostly copying the neghbouring native areas. But the dialects are disappearing quite quickly with only some major interdialects remaining and with a strong influence of Common Czech as a sort of interdialect of Bohemia and sort of a common vernacular layer in many other places. There will be very little new dialect forming there and any systematic study of various effects on them would be probably impossible. It would be hard enough to find some evidence of any formation at all. Commented Jun 6, 2022 at 11:04

1 Answer 1


There are many studies that attempt to quantify the relative weight of such factors: there is a substantial sub-field of linguistics known as quantitative sociolinguistics.

For the most part, such studies are fairly descriptive, in the sense that they will look at some linguistic question, given a certain population, then try to figure out what non-linguistic properties correlate with what behaviors. As with all social research, you have to have a reasoned basis for thinking that a particular fact is worth throwing into the mix. For example, if you are studying American English in the Pacific Northwest, you would care most about race, income, and personal history w.r.t. location (mostly urban vs. rural but possibly controlling for proximity to Canada) – i.e. "Born and raises in California, moved here last year" vs. "Great Grandfather moved here in 1865". You have to know what is or has been "socially relevant".

Age is generally emerging as a highly significant factor in most of the world, because Grandfather tends not to hang out in global internet hangouts, and Grandchild does. There aren't any general and universal models that relate mountain ranges to dialect diversity, instead there is a simple observation that people who don't interact have a greater chance of diverging in their language than people who interact frequently. Mountains have been an impediment to interaction, same with rivers, snow, politics, but you have to independently measure the actual effect of mountains in a particular context.

  • 1
    Caleb Everett has been working on the effects of geography on languages, with some interesting results. The issue of dialect formation is a different matter, however; it's much more general and has to do both with physical and social separation, and with ingroup/outgroup identification and bonding.
    – jlawler
    Commented Jun 5, 2022 at 19:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.