Triggering Events

William Labov
Professor of Linguistics
University of Pennsylvania

 Martinet argued that linguistic changes often represent the long-term structural consequences of some remote external disturbance of the speech community. Since every linguistic stage is the result of some preceding stage, the question can be raised as to whether there are linguistic triggering events which are not merely links in the chain. It is proposed that the typical triggering event is the unpredictable resolution of an undetermined situation which entails a series of irreversible consequences. Three cases in the recent history of North American English will be considered.

The first case involves a conflict of short-a systems in western New York State in the early 19th century. One resolution led to the general raising that triggered the Northern Cities Shift, and another to the even more widespread nasal system in New England, the Midland and the West--tensing before all nasals and nowhere else.

A second case is the low back merger of cot and caught, etc. It is only one possible resolutions of a delicately balanced and unstable opposition that was itself the result of a concomitance of historical accidents. The collapse of /o/ and /oh/ was then the triggering event for the Canadian Shift of the short front vowels, and the alternative Pittsburgh Shift involving the rotation of wedge and the low back vowel.

A third case to be considered is the continental forward shift of /uw/. This is a prima facie counter example to Martinet’s proposal that the fronting of /uw/ (and /ow/) is a response to overcrowding in the back vowels, since /uw/ fronting is particularly strong in areas of the low back merger. However, it will be proposed that the initiating cause is not to be found in vowel configurations but in a very different type of event in the consonantal system. One further characteristic of a triggering event is that it is orthogonal to the chain it initiates, In this case it is the loss of the /y/ glide after coronal consonants, a process almost complete throughout North America. The phonemic information conveyed by the glide was transferred to the /iw ~ uw/ opposition after coronals, which in turn proved to be unstable, and in merging triggered the massive fronting of /uw/. Evidence for this scenario is provided by the Atlas of North American English, which shows 400 to 500 Hz differences between /uw/ after coronals and after non-coronals in all intermediate stages of the fronting process.