Principles of linguistic change 1 - internal factors

We hold that words change their pronunciations by discrete, perceptible increments (i.e. phonetically abrupt), but severally at a time (i.e. lexically gradual) . . . (Wang & Cheng 1977, p. 150 – my emphasis)

There is no basis for contending that lexical diffusion is somehow more fundamental than regular, phonetically motivated sound change. On the contrary, if we were to decide the issue by counting cases, there appear to be far more substantially documented cases of Neogrammarian sound change than of lexical diffusion. The upshot of this discussion is not, however, a victory for the Neogrammarians. There are more than enough solid examples of sounds changing one word at a time to support the view that lexical diffusion is deeply rooted in the process of change. But where it is rooted, and where it flourishes, is more difficult to decide. We cannot resolve a paradox by saying that both parties are right. Though we have advanced evidence on both sides, the situation is no less chaotic than it was at the outset. If language changes in two different ways, the data must contain some hint that will tell us when it changes in one way, and when in the other.

Much of the discussion concerned synchronic variation; it was found that the choice of one linguistic variant or another is not determined by the need to preserve information, but is influenced by the tendency to maintain parallel structure and parallel articulation. The chapters in this part considered how this syntagmatic principle might affect the course of change, which is essentially paradigmatic, following the principle that functional effects are primarily the consequences of linguistic choices rather than the causes of them.

Given phonological and morphological variation, the functional hypothesis predicts a tendency for speakers to choose one variant or the other in a fashion that will preserve information. Most of the results cited show the opposite: in the stream of speech, one variant or the other is chosen without regard to the maximization of information. On the contrary, the major effects that determine such choices are mechanical: phonetic conditioning and simple repetition of the preceding structure.

Labov (1983)

This study concerns the choice between the passive (1) and the generalized active (2) in English.

Figure 20.3 shows that the mechanical effects of priming and parallel structure were by far the most important.

• The passive was favored most strongly if the preceding instance of the variable was also a passive (.69 vs. .31).
• There was also a strong tendency to maintain parallel structure: that is, coreferential subjects are maintained in the same syntactic position (.62 vs. .38).
• The third effect is one that might be considered functional, though it does not involve the loss or retention of information. It involves the tendency for given information to be placed first, and new information to follow. There is a tendency to favor the passive if the subject of the passive is given information (.56 vs. .44), that is, if it occurs somewhere in the preceding four clauses.2

When an element of the system is found to cooccur frequently with the signal being deleted, it is increasingly eligible to assume the distinctive feature representing the semantic feature in question.