The geography and development of language isolates.

Matthias Urban

doi:10.1098/rsos.202232

Abstract

This contribution theorizes the historical dynamics of so-called language isolates, languages which cannot be demonstrated to belong to any known language family. On the basis of a qualitative review of how isolates, language families or their branches lost territory to other languages through time, I develop a simple model for the genesis of isolates as a function of proximity to major geographical barriers, and pit it against an alternative view that sees them as one manifestation of linguistic diversity generally. Using a variety of statistical techniques, I test both accounts quantitatively against a worldwide dataset of language locations and distances to geographical barriers, and find support for the position that views language isolates as one manifestation of linguistic diversity generally. However, I caution that different processes which are not necessarily mutually exclusive may have shaped the present-day distribution of language isolates. These may form elements of a broader theory of language isolates in particular and language diversity in general.

Highlights

This contribution theorizes the historical dynamics of so-called language isolates, languages which cannot be demonstrated to belong to any known language family
On the basis of a qualitative review of how isolates, language families or their branches lost territory to other languages through time, I develop a simple model for the genesis of isolates as a function of proximity to major geographical barriers, and pit it against an alternative view that sees them as one manifestation of linguistic diversity generally
It has almost become a truism to say that linguistic diversity is unevenly distributed across the globe

Summary

Introduction

It has almost become a truism to say that linguistic diversity is unevenly distributed across the globe. Isolates can be seen as one particular manifestation of linguistic diversity generally There is another way to look at isolates that takes the Normal Diachrony Assumption as its starting point, but develops it further through a different line of thought: if isolates are conceived of as the sole survivors of former language families that were largely superseded by later language spreads, they invite to theorize the expansion and diversification of languages and language families, and reductive processes that are the mirror image of expansion and diversification. They may be crucial as background information for developing descriptively adequate synchronic and diachronic models of the factors that shape synchronic linguistic distributions

Two models for the distribution of language isolates and language diversity

50 Maximal extension Present-day distribution

Analysis and results

Language data

Geographical data

Wilcoxon–Mann–Whitney test

Findings

Bayesian logistic mixed effects regression