The dynamics of this process of language shift are exceedingly subtle and complex.
One of the principal differences between an ecological and an earlier mechanistic approach to languages is that the former subscribes to the interdependency between linguistic and numerous other ecological factors. It sees the well-being of individual languages or communication networks as dependent on a range of language-external factors as well as the presence of other languages and thus not as something which can be meaningfully studied in its own right, though it is possible to focus on certain symptoms of language death and decline, as will be been done in Chapter 11. The arguments put forward here have their analogies in studies concerned with the disappearance of natural species. The change of a single link in a ecological network can precipitate very considerable overall changes, the disappearance of one species typically leading to that of a dozen of others.
The term ‘ecology’ is derived from the Greek word for house or home. We can employ this house metaphor by pointing at the removal of only small, but important, parts of a building which can lead to its collapse or vastly reduce its fitness for habitation. Few approaches of linguistics cater for such a concept of language. The one which I have found most suited is the ethnography of communication approach developed by Hymes (1974), which focuses on a wide range of components of speech situations and speech events occurring within a language ecology.
I shall examine the use of Hymes’ scheme for the study of the language decline shortly. Before doing this, however, let me contrast the ecological approach with another, atomistic approach influenced