Some thoughts on some of Alexander Gil's vowels

Just a couple thoughts on the possible phonetic values of some English vowels in the transcriptions of the 17th century spelling reformer Alexander Gil

The PRICE Vowel (modern /aɪ/) 

There is some disagreement in the field about how this vowel developed in Gil's period. Probably the most generally held view is that the process of diphthongization went /iː/ → [ɨi] → [ɘi]  → [əi] before 1500 and stayed there till the 18th century. Another possible view, to which Patricia Wolfe devoted a most interesting dissertation and recently put forward by Roger Lass, is that it took a peripheral path /iː/ → [ei] → [e̞i]  → [ɛi] → [əi] and the last stage wasn't reached until the mid-1600s. There are various other possibilities, like an intermediate path /iː/ → [ɪi] → [e̙i]  → [ɛi] → [əi], or an intermediately central path /iː/ → [ɪi] → [ɘ̘i]  → [əi] (moving, as it were, "diagonally" across the vowel space). Another is that the trajectory was central but that the timing is just different, and the [əi] stage was only reached in the early 17th century. Yet another variation on many of these takes a dip into the domain of [ɜi]. I can think of still more possibilities, and could plot out hypotherical vowel trajectories all the livelong day. 

The issues involved are complicated, but basically the main point is that no source before 1640 really describes a centralized vowel of any kind in this position whereas afterward they all seem suddenly able to. At the same time, the development of the WAIT diphthong (which in some types of speech at least did develop to [ɛi], where it remained for a long time) makes a peripheral path more complicated. The tension is between those who are inclined to credit direct testimony as much as possible, and those who have theories about how sound change works that make their testimony hard to believe on this point. For Alexander Gil's dialect specifically, the [ɛi] state is clearly spoken for. He recognizes this as its own phonetic entity and appears to take it mostly as a variant realization of his WAIT vowel (though, because his transcription choices are influence by spelling, he represents it as such in some of those places where traditional orthography contains <ei>, thus making it correspond partly to Middle English /ɛi/). This is a common theme with Gil: because he is trying to create a new and transparent spelling-system for everyday use which will be well-received, he often opts, among variant forms known to him, for that pronunciation whose transcription comports most with traditional orthography. (Though this is not always so, as with his pronunciations of "heard", "forth" and "yet"). Sometimes (as with the pronunciation of "other", "you", "mother", "fault", for example) he remarks that his spelling-conforming pronunciation is that used only by learned men. 

Gil also notes that his PRICE diphthong was "almost [ɛi]" and the use of his [ɛi] symbol for the word "eye" in particular (and the fact that he seems inclined to use e to represent reduced vowels in particular before /r/) suggests that he might have been happy to spell it as ei were it not for other concerns, like his desire to use a single symbol for this diphthong. He also seems to object to Hart's spelling of this diphthong with the letters ei in a way that could be taken to suggest that [ɛi] for the PRICE vowel was associated with a form of speech he didn't care for. 

Anyway, it seems to me that something high but retracted in the range of [ei]~[ɘi]~[ë̞i] is the least inconsistent with all the evidence available, and quite appropriate to the accoustic impression of an "e-like" quality. 

The MOUTH vowel (modern /aʊ/)

Gil represents this as a combination of his o and his u, as do Hart and (in a slightly odd way) Robinson. The traditional reconstruction of this entity is [ǝu], with posited trajectories like /uː/ → [ʉu] → [ɵu] → [əu] or /uː/ → [ʊu] → [ɵu] → [əu]. Again, Wolfe and Lass suggest the peripheral path more apparently in keeping with what pre-1640 sources actually seem to describe: /uː/ → [ou] → [ɔu] → [əu]. It seems hard to deny that the same phonetic element was not present in both the MOUTH and PRICE diphthongs before the 17th century in the kinds of speech being transcribed. Even Wallis in the 1650s (who describes what Dobson takes, IMO wrongly, to be /ʌu/) distinguishes two different vowels as the first elements of the MOUTH and PRICE vowels. In the latter, this is unarguably [ǝ], but in the MOUTH diphthong Wallis seems to describe something else, something similar to (what was probably even then) French [œ] in serviteur whose main distinguishing feature from [ǝ] was smaller lip aperture. In other words, there's some rounding involved. This sound Wallis recognizes as the same as the vowel of nut etc. This description would suit [ɞ] quite nicely and I find it rather hard to square with [ʌ]. The main objection to reconstructing a rounded back first element to the MOUTH diphthong early on is the question of how it was distinguished from the reflex of ME /ou/ (the "know" vowel, if you will). Hart, Gil, Cheke and Smith all transcribe the latter as if it were their GOAT vowel followed by [u], in other words /ɔːw/. I am inclined to think that is (more or less) what it was, given its behavior (including the tendency of the KNOW vowel before ME /l/ to merge in some types of speech with the MOUTH vowel). Admittedly this presents a slight typological problem. While /ɔːw/ as a phonemic segment can occasionally be found in languages (Ionic Ancient Greek, modern Somali), it's usually not in the absence of other long diphthongs. A closer parallel might be East Cree. But the evidence of Sir Thomas Smith's transcriptions would seem to suggest that the DEW vowel had (or could optionally have) a similarly long nucleus. Anyway, this all suggests the MOUTH vowel having a rounded vowel followed by [u] for this period. It may well have been [ɔu], but [ɞu] for Gil as well is also conceivable, perhaps even more so something near-back [ɔ̟u]. I take the latter as the best guess, or at least the most unlikely to be severely wrong.

The MEAT vowel

Gil's MEAT vowel (the reflex of ME /ɛ̄/) seems pretty safely /ɛː/ and for basically the same reasons given by Dobson. 

The GOAT vowel 

I find no reason to dispute the idea that Gil's GOAT vowel (the reflex of ME /ɔ̄/) is /ɔː/ (perhaps more specifically [ɔ̝ː]). He doesn't give the same kind of evidence he does for his MEAT vowel, and the orthoepists in general are a lot less clear about the height of this vowel than its front and unrounded counterpart, but the survival of [ɔː] into the 17th century in some forms of speech alongside [oː] is firmly in evidence. (John Florio (1611) reports GOAT [ɔː] quite unambiguously, finding in it a good counterpart to Italian /ɔ/ as against Italian /o/ which Florio equates with his reflex of ME /ŭ/). But pushing Gil's GOAT vowel higher doesn't solve the issues with the KNOW vowel, really.

The DUE vowel

I am completely undecided at this point between the general position of [ɪu̯] and Dobson's reconstruction of [yː] for this type of speech. I've gone back and forth on this, and I don't think the question of Early Modern [yː] as a minority variant is ultimately solvable given the nature of our evidence. I do think that the arguments made against [yː] have typically been very stupid. But the evidence for it isn't as clear as Dobson thought. I'll need to fill this out at length at some point.