The universality of language change

Human language would presumably function much better as a medium of communication if it did not change over time, because dialects would not diverge and mutually unintelligible languages would not multiply (see Chapter 10). But that is not what we find; on the contrary, any language or dialect recorded for even a few centuries can be shown to have changed over the course of its recorded history, and in recent decades William Labov and his students and colleagues in sociolinguistics have uncovered evidence of change in progress in practically every speech community in which they have looked for it. Apparently language change is universal. That is all the more surprising because in modern literate societies there are institutional forces (such as schools and newspaper columnists) that militate against linguistic change – and invariably fail. One naturally wonders why, but that question is not fundamental enough; we need to ask “how” before we can ask “why.” Let us begin by looking at the mechanics of change: how specific language changes originate, and what happens to them over time. (See already Paul 1960: 18–20, 24–5, 32–4 – originally written in 1880 and last revised in 1920!)

Potential sources of language change

A linguistic change has occurred when an innovation has spread and become accepted in a speech community. If we want to understand the entire course of the change, we first need to ask where the innovation came from.