2nd Shift of the Great Vowel Shift

[Anonimulo]

OK, so, you know English vowels? How ɔː rose to oʊ rose to uː rose to aʊ? And likewise aː rose to ɛː rose to eɪ rose to iː rose to aɪ? Etc, etc. Permitting exceptions. But then, permitting exceptions again, ɛː rose to eɪ rose to iː. I was just wondering if we know what words were victim to this second shift? Because not all of them were. Like break. Etc. Anyway, if there's some sort of list. Or can we just tell by plain spelling? Like the a-e/ey/ay spelling of ɛː and the ea seplling of eɪ?

[Soap]

Well, Scots didn't go through the shift, so for the most part they preserve the original pronunciation, with the exception of words that didn't exist in Scots at the time and thus are "loaned" from English English.

[Magb]

Looks to me like only break and steak have /eI/ out of all words spelled with <ea>. I don't know if this is related, but I've found five words with <ear> that have /E@r/: bear, pear, swear, tear and wear (Source: Rhymes:English:-ɛə(r)). All other words with <ear> have the BEER vowel. Whether all these spellings accurately reflect historical phonemes, I have no idea.

[Travis B.]

It should be noted that the merger of most meet and meat-words is actually not traditionally present in all of Anglic, even outside Scots. A range of traditional Anglic dialects, and not just Scots, in the British Isles never merged the two, instead shifting Early New English /eː/ (1) to things such as but by no means limited to /ɪə̯/ (2). However, Standard English is based off dialects that did merge the two.

(1) X-SAMPA: /e:/
(2) X-SAMPA: /I@_^/

[Anonimulo]

I also had "great", but. So far to me it seems spelling is the only clear indicator, maybe the only necessary one, too. But. I'm too inexperienced at this sort of thing to be sure.

But, doesn't Modern English spelling pretty much reflect Middle English pronunciation anyway? I mean, if the words weren't updated in Early Modern English.

[Travis B.]

Another note here is that one should not refer to Early New English /eː/ (1) and /oː/ (2) as /eɪ̯/ (3) and /oʊ̯/ (4), as they were not the latter. Rather, later /eɪ̯/ (3) and /oʊ̯/ (4) are the products of the (later) mergers of ENE /ɛː/ (5) with /ɛi̯/ (6) and ENE /oː/ (2) with /ɔu̯/, mergers unrelated to the Great Vowel Shift. These mergers are not universal in Anglic or even Anglic outside Scots, having occurred in Standard English but still not having occurred (at least completely) in some non-Scots Anglic dialects to date.

(1) X-SAMPA: /e:/
(2) X-SAMPA: /o:/
(3) X-SAMPA: /eI_^/
(4) X-SAMPA: /oU_^/
(5) X-SAMPA: /E:/
(6) X-SAMPA: /Ei_^/
(7) X-SAMPA: /Ou_^/

[AnTeallach]

It should be noted that the merger of most meet and meat-words is actually not traditionally present in all of Anglic, even outside Scots. A range of traditional Anglic dialects, and not just Scots, in the British Isles never merged the two, instead shifting Early New English /eː/ (1) to things such as but by no means limited to /ɪə̯/ (2). However, Standard English is based off dialects that did merge the two.

It's even more complicated than that. In traditional West Riding dialect the reflex of Old English /{:/ isn't merged with that of Old English /e/ lengthened in open syllables, so there are two diphthongs corresponding to the spelling <i>ea</i>; the former is [I@] or similar, and the second is [EI] or similar. Note that this [EI] is distinct from the reflex of Middle English /a:/, which is monophthongal: meat [mEIt] vs. mate [me:t] vs. meet [mi:t], and mean [mI@n].

[Magb]

I also had "great", but.

Oh yeah, and I guess it's worth mentioning the various words where <ea> is /E/, like lead (n.), stead, meadow, read (past tense), etc.

[Travis B.]

Yeah, it is really hard to make such blanket statements about the linguistics of Anglic dialects in nice short posts. Unfortunately, I myself have seen little really comprehensive coverage of traditional Anglic dialects of any real linguistic rigor, so it is hard for me to comment on them in a comprehensive kind of manner.

The main thing for people to remember here, though, is that when people speak about "the history of English", they really are only usually speaking about the history of the dialects of southeastern England, and then only those that present-day Standard English and varieties allied thereto* are based off of. Even when it comes to southeastern Anglic dialects, there are definitely things that were going on historically that are generally missed in these kinds of things, such as that some dialects actually merged /w/ and /v/ as something like [ʋ] (1), which has since been lost, to my knowledge.

(1) X-SAMPA: [v\].

* for instance, my own dialect; while superficially this differs significantly from, say, General American, diachronically it is clearly actually very closely related to it, likely having a time-distance from it of only a hundred or so years, a hundred and fifty at the very most, and thus is closely allied to Standard English

[Anonimulo]

Sorry, I guess I should have mentioned I was speaking of prestige dialects such as GenAm and RP. The dia-phonology of English, per se. 'Cuz I'm working on this respelling of English that isn't totally better than what we have now, but. I have a preference for symmetry and I was going to remap the vowel phonemes of English like this:

Checked Vowels: /æ ɛ ɪ ɔ ʊ ɐ/ <a e i o u y>
Long Vowels: /ɑː eɪ iː oʊ uː ər/ <a e i o u yr>
Diphthongs: /ɛː ɔː aɪ ɛʊ ɔɪ aʊ/ <ai au ei eu oi ou>
Unstressed Vowels: /ə ɪ ʊ ər/ <a/o e/i u yr>
R-Colored Vowels: /ar ɛər ɪər ɔər ʊər ər/ <ar er ir or ur yr>

So, I wanted to drop those affected by that second shift I mentioned, whether or not it was actually part of the Great Vowel Shift. Compared to actual English, /ɛː/ does not exist, /ɛʊ/ is really /juː/, /ɐ/ is normally transcribed /ʌ/, and /ər/ as /ɜr/. Oh, and /ɔ/ is /ɒ/.

But anyway, this isn't about that. I'm just looking for the answers on the second shift with regards to standard English, and I'm pretty convinced that orthography is the key giveaway?

[Travis B.]

I should warn you that what you are attempting to do is much harder than it sounds on paper, due to the many vagaries of Anglic diachronics and dialectology. I myself have tried to do essentially what you are trying to do, but at the same time trying to have an internal representation that is as crossdialectally applicable as possible, even at the expense of having spellings that make sense diachronically but which are hard at times to relate to present-day standard varieties of English and which contain many distinctions that have since been lost in all standard varieties of English. I gave up on it partially because I do not seriously believe in English orthographic reform as a practical possibility but also simply because I did not have the resources to it well without making arbitrary and likely incorrect choices along the way. If you are even going to try this, I would very strongly suggest obtaining scholarly-quality resources on Anglic diachronics and dialectology or you will run into problems.

[Alces]

I think for the most part, words with <ea> (or long <e>) in the spelling accurately reflect Middle English /ɛː/, whether its modern standard value is with /iː/ or /ei/; at least, I don't know any obvious exceptions. If in doubt, you can go to a website like http://www.etymonline.com and look at the Old English form; /ɛː/ comes from OE /ǣ/, /ēa/ or open-syllable lengthened <e>.

[Anonimulo]

I here what you're saying, Travis, but this is just a personal thing for me. A little psycho experiment just to see what can come of it. I realize that using <yr> for /ɜr/ leaves out all the speakers who distinguish /ɛr ɪr ʌr/, but I'm not trying to make an orthography that is friendly to all Anglophones. Just the diaphonemes. Even then, my orthography isn't going to distinguish voicing in fricatives and <v> represents /w/ in words derived from Middle English and all that and /v/ in words from Anglo-Norman and Old French and that stuff, while <f> represents /f/ in all contexts and /v/ from native English words. So, definitely imperfect. Just a preferential thing.

[brandrinn]

The big question for me has always been /o:/ > /u/. It seems to be extremely finicky.
do, to, who
go, so, no
What gives?

[Travis B.]

The big question for me has always been /o:/ > /u/. It seems to be extremely finicky.
do, to, who
go, so, no
What gives?

Not all these words had /oː/ (1) in Late Middle English. While the first set of words had it, the second set had either /ɔː/ (2) or, in the case of so, /waː/ (3). It just happens that the differences between these two sets is not clearly delineated in present-day English orthography, even if it was in much Late Middle English writing in the case of swa for present-day so.

(1) X-SAMPA: /o:/
(2) X-SAMPA: /O:/
(3) X-SAMPA: /wa:/