1
L563F10.A02 Meredith Tamminga
Assignment 2: Triggering Events
The widespread monophthongization of /ay/ in the South seems like a plausible trigger for the subsequent events of the Southern Shift—and indeed, it has standardly been analyzed that way. To be more explicit, the movement of /ay/ out of the long upgliding subsystem and into the long ingliding subsystem (/ay/à/ah/) is proposed to be the first step in a pull chain, where the second step is the lowering of /ey/ towards the original position of /ay/. The short vowel /e/ then raises into the newly empty space and develops a glide, effectively switching positions with /ey/.[1] I argue, however, that /ay/ monophthongization should not be analyzed as a triggering event for the Southern Shift.
One source of evidence for causality in sound changes in progress is geographic nesting effects. For example, when we consider the Canadian Shift, we observe that the isogloss defining it is contained entirely within the isogloss outlining the low-back merger. We take this as evidence in favor of an analysis where the low-back merger triggers the Canadian Shift. If we compare the spatial distributions of the reversal of /ey/ and /e/ and the monophthongization of /ay/, however, we might notice that there’s something off about the pattern.
Map 1: /ay/ monophthongization before voiced obstruents & finally, compared to the second stage of the Southern Shift (/e/-/ey/ reversal) [reproduced from the ANAE online]
At first glance, all seems to be in order. In the ANAE, the isogloss for /e/-/ey/ reversal sits tidily inside the isogloss for /ay/ monophthongization before voiced obstruents. We see this in Map 1 above. But upon closer inspection, it should be noticed that the isoglosses for /ay/ monophthongization before all obstruents, voiced and voiceless, pick out only northern Texas and parts of Appalachia. This is illustrated in Map 2.
Map 2. /ay/ monophthongization before voiced obstruents & finally, compared to /ay/ monophthongization before all obstruents (including voiceless) [reproduced from the online version of the ANAE]
I submit that this observation poses a problem to the usual analysis where /ay/ monophthongization triggers the Southern Shift. The pre-voiceless-obstruent environment is hardly marginal, including plenty of normal and common English words such as stripe, wipe, type, white, write, bite, bright, light, fight, like, strike, pike, lice, rice, nice, etc. In the face of a lack of monophthongization in this environment, the claim that /ay/ has “left” the long upgliding vowel subsystem seems spurious. A speaker living in such an area must encounter plenty of tokens of diphthongal /ay/ every day. The gap in the long upgliding subsystem that is allegedly left by /ay/ monophthongization is in fact still filled with the many voiceless-obstruent-final words that /ay/ monophthongization left behind. And if this is really the case, the logic that /ey/ lowers to fill that gap falls short.
[1] A third stage involves an analogous reversal of /i/ and /iy/; I will ignore that here.