The author advocates a basically new approach to language change.
A change begins when a speaker says something new choosing between
several possibilities following nothing more complicated than the principle
variatio delectat. If other speakers like the new sound or expression, it may
become a language innovation. Since it is first used by a group of speakers,
it has distinct sociolectal value.
In order to communicate messages the old and the new sound or expression
function equally well. Traditionally linguists have generally tried to find some
weak point which is the cause of the language change. They have tried to
find causal explanations where there are none.
Language is a branch of human culture. Its changes can no more be
explained than changes in literature, music, visual arts or the length of
women’s skirts. The innovator may choose the new sound or expression
because of similarity with some existing feature of the language. The author
calls this assumed process a prompt, which is similar to a cause but is not
The notions of information value and redundancy, borrowed from information
theory, can elucidate some aspects of language change.