Giancarlo (2001) discusses the ways in which the GVS has been represented by others in diagrammatic and tabular form as a way of simplifying, visualising, and generally making palatable to the uninitiated student the phonetic complexities of the shifts that it covers. His overall point is that it is a way of telling “the story” of English. However, he admits that it is the story of the development of standard English that is at stake rather than the story of the GVS itself. The dispute that the diagrams have generated—possibly precisely because vowel shifts have been represented as diagrams—is whether we have a chain reaction in the whole of the vowel system and whether the set of movements was sparked off by the shift of the long high vowels [i:] and [u:]. That is, at issue is whether this movement “dragged” the mid-high vowels [e:] and [o:] into the places vacated in the vowel trapezoid by the diphthongi- sation of the long high vowels, or whether the initial movement was the raising of the mid-high vowels [e:] and [o:] and the mid-low vowels [e:] and [э:] into the upper slots, “pushing” the occupants of those slots out. In any event, we have a mixture of “drag-chain” and “push-chain” effects that can be represented in the form of a diagram without really knowing whether the conventional wisdom reflects the facts. This dispute can still be found in literature on the GVS today, and, since it is not my purpose to argue about the fine points of historical phonology in this chapter, I will not discuss the merits of the two explanatory camps of drag-chainers and push-chainers (cf. fig. 6.1 above).
The other major issue concerns the process of the diphthongisation of the high vowels. Did [i:] move to a more central position, say, something like [li] and thence to [ai], or did it move to [ei] first? Did [u:] move to something like [uu] and thence to [эи], or did it move to [ou] first? In addition, there are varieties of English in northern England and Scotland in which certain parts of the “chain” shifts did not occur, a point that is not lost either on Stockwell and Minkova or on Lass. The classic example is the failure of [u:] to shift to [au] in Lowland Scottish English, the Northumberland dialect and Tyneside Geordie.
The explanation for this failure neatly exemplifies the kind of arguments that are used to bolster the drag-chain or push-chain theories. Push-chainers such as Lass maintain that because [o:] was fronted to [0:] in these northern dialects, it did not come under pressure to diphthongise. But it is relatively easy to imagine a drag-chainer claiming that if a dialect could be found in which [o:] was fronted and [u:] nevertheless diphthongised, the push-chainer’s argument disintegrates.
Further dispute revolves around the nature of the evidence used to support different interpretations of what, how and when shifts in the GVS took place. Wolfe (1972) and Lass (1989) both rely on the real-time evidence of orthoepists such as Hart, Gil and Bullokar in the sixteenth century and Cooper, Daines and Wallis in the seventeenth, whereas Stockwell and Minkova challenge the phonetic accuracy of the orthoepists’ observations and instead rely on spellings and rhymes. This leads them to posit much earlier datings for the first shifts in the GVS, and it also leads them to challenge the push-chain theory. To find orthographic evidence of this kind, Stockwell and Minkova need to examine texts that are written in localised varieties of English. But even Lass’s argument on the partial failure of [u:] to shift forces him to consider northern varieties of English.
As we saw above, several varieties of English, both within Great Britain and in the wider English-speaking world, also show different end points in the diphthongisation process, [ai], [rn] or [oi] for [ai] and [эи] and [du] for [au]. If we add to this the fact that since the GVS a number of shortening and lengthening processes have taken place together with the fact that certain dialects in the United States (Boberg 2000; Labov, Ash & Boberg 2006) and in New Zealand (Bauer et al. 2007; Trudgill & Hannah 2002) are currently undergoing shifts in the short back and front vowels, then it is time to start reconsidering the significance of the GVS and to give up the teaching heuristic as a simplistic justification of the development of the standard.
In the following section I indicate how conventional representations of the GVS have been challenged and how this leads to the need for a much more careful analysis of the phonetic processes involved in vowel shifting in general. The final link in my argumentative chain concerns the producers of the vowels themselves. After all, even if we believe with Weinreich, Labov and Herzog (1968) that languages do not change, but that people change languages, we still need to fix the dialectological and phonetic evidence to a theory of how speakers perceive innovations, of how they adopt them and of how those newly adopted innovations are diffused through space and time. This problem will be tackled in the final section.