Northern English Video: Additional Material

In my last YouTube video, I demonstrated how difficult it is to reliably compare how conservative two accents of the same language are. I used traditional Cumbrian, the non-southern-English dialect with which I’m most familiar, and traditional received pronunciation. I compared them up to about 1900, as this avoids the complexities of dialect levelling that have happened since then.

The resources for reconstructing the phonological history of northern Anglo-English are different in nature to the ones we usually use for southeastern Anglo-English. From the southeast, we have endless phonetic descriptions from about 1550 onwards, allowing us to reconstruct at quite a high level of phonetic detail. From the north, we have scant dialect texts which are often more focused on displaying the unique words of Cumbrian, implying the pronunciation through sometimes inconsistent spelling choices.

From the Middle English period (before any meaningful explicit phonetic descriptions), we have to try to map the spelling and rhyming evidence onto a naturalistic set of sound changes of the kind that we know to happen in modern languages, and ensure our reconstruction makes sense in light of how each dialect has ended up in recent times.

There are lots of reasons for thinking that the basic short vowel phonemes - those in TRAP, DRESS, KIT, LOT and BULL - have not changed much from their Middle English values in either RP or Cumbrian. For one, they show a lot of consistency in spelling over the last thousand years or so. For another, the spelling-pronunciation relationship is more ‘normal’ for a European language; the short vowel represented by <i> is now a high(-ish) front unrounded vowel in both RP and Cumbrian, and this is the sound it commonly represents across Europe. It makes sense that early scribes would choose this letter for a sound like [ɪ], and continue using it as the sound stayed the same in quality. There is more evidence that narrows down the qualities of these vowels, but I will focus on the long vowels in this post, as I feel that they are more interesting.

The ME Long Vowels

In the video, I chose the year c.1100 as a rough starting point when I compared the development of the Middle English long vowels. I think this is a good candidate for when the southern and northern vowel systems meaningfully ‘diverged’ from each other. This certainly doesn’t mean that the southern and northern long vowel systems were identical at this point, especially in terms of distribution: as I pointed out in the video, some words had a long vowel in the south and a short one in the north (such as finden ‘to find’).

The ‘common south-north’ vowel system I used for c.1100 was structured like this:

Clip 1: The seven short vowels c.1100.

This isn’t exactly the same as the chart that most books or websites will show for Middle English: they usually show the BOAT vowel as [ɔː], at the same height as the BEAT vowel. But at this early stage in the period, we still see this vowel spelled with <a> in both the south and the north.

The two areas diverge in spelling over the ensuing couple of centuries; for the BOAT vowel, <o> spellings become more common in the south and <a> spellings persist in the north. What does this spelling change mean? Well, the letter <o> had long been used in English to write a mid-height vowel with a back tongue position and lip rounding - roughly [oː] - and that trend applies more broadly across European languages as well (because that’s what the letter signified in Classical Latin). So this change in spelling suggests that southerners were leaning away from an open, unrounded [ɑː] vowel, and towards one with a higher tongue position and lip rounding. We know that this BOAT vowel didn’t merge with the GOOSE vowel in mainstream southern English, as they aren’t merged today and they are generally kept separate in Middle English rhyme schemes - so the most likely candidate is a lower rounded vowel, perhaps [ɒː] to begin with, and then eventually [ɔː].

Clip 2: Southern Middle English BOAT in its gradual ascent, [bɑːt̪, bɒːt̪, bɔːt̪].

We are fairly sure that this course of development did not take place in the north - mainly because northerners continued spelling BOAT words with <a>, and partly because traditional northern English and Scots have front, unrounded realisations of this vowel such as [e] and [jä]. So in the north, we may be looking at a continuation of this [ɑː] quality in BOAT words, which gradually took on more of a front tongue position something like [äː] or [aː].

Clip 3: Northern Middle English BOAT in a few possibilities of its original low quality, [bɑːt̪, bäːt̪, baːt̪].

This is where things become more complex than I let on in my video. Across the country, the TAKE vowel was probably also a low vowel with no lip rounding. We suspect this because these words continued to be spelled with <a>, and because when phoneticians did eventually start describing their vowels explicitly, the earliest ones (c.1550) tended to describe a low unrounded vowel like [äː ~ aː]. In some dialects of northern Middle English, the TAKE vowel actually merged with the BOAT vowel, so that the words hate, frame, stone, boat, one all took the same vowel. Vowel systems like to be symmetrical, so I would lean towards [äː] as the quality of this vowel in merged dialects. This merger was probably more likely in the northernmost parts of the area where Middle English was spoken, i.e. Scotland.

Clip 4: Northern Middle English merged ‘hate’, ‘frame’, ‘stone’, ‘boat’, ‘one’.

This is clear from modern Scots varieties in which many of these words still take the same vowel as each other, albeit one with a much higher tongue position:

Clip 5: My attempt at modern Scots ‘hate’, ‘frame’, ‘stane’ (‘stone’), ‘ane’ (‘one’).

However, dialects within northern England had the option of keeping the BOAT and TAKE vowels separate, as evidenced by their development in Cumbrian by 1900:

Clip 6: 20th-century Cumbrian ‘hate’, ‘frame’, ‘styan’ (‘stone’), ‘yan’ (‘one’).

Börje Brilioth, a Swedish linguist who described a variety of Cumbrian in 1913, said that the presence of the modern [jä] diphthong was difficult to predict from the surrounding sounds. By Brilioth’s time, it generally occurred in words where:

1. [ä] in Early Middle English had lengthened because it was in an open syllable.

2. [ɑː] in Early Middle English remained long, and was not followed by /r/ at the end of the syllable.

However, plenty of words in the first category instead had [ɛ̝ː]. This might be because Cumbrian also underwent that Scots merger of the BOAT and TAKE vowels, but the merger never managed to completely affect all of the TAKE words. Alternatively, it might because there was no such merger in the first place in Cumbrian, but influence from neighbouring Scots caused many TAKE words to end up with the BOAT vowel. Or it might be that the merger did completely happen in Cumbrian, but later influence from southern standard English gradually reintroduced TAKE words to their ‘original’ lexical set.

In any case, in both Scots and Cumbrian, this front development of the BOAT vowel did not completely rid the vowel inventory of long back vowels. As I stated in my video, the MOUSE and GOOSE vowels remained back vowels through the earlier Middle English period, probably with the same qualities that they had in the south:

Clip 7: Early Middle English ‘mouse’ and ‘goose’, as they were in both the south and the north c.1100.

However, there was a third long back vowel that I didn’t mention in my video. Where Old English [o] had lengthened in open syllables, it became [ɔː], and was joined by some French loan words. In southern England, this new long [ɔː] soon merged with the BOAT vowel, which is why ‘bone’ and ‘close’ now have the same vowel in standard English. In Cumbrian, on the other hand, ‘bone’ has ended up as [bjän] by 1900, and ‘close’ as [klwɔs].

The other ‘weird’ (by metrocentric standards) development in Cumbrian is that the GOOSE vowel, which was [oː] in Early Middle English, flew across the vowel space and took a front tongue position while keeping its lip rounding, becoming [øː]. This change must have permeated through northern England and Scotland alike, because traditional northern English and Scots both have front realisations of this vowel: [ɪə] or [jʊ] in Cumbria, [ɪ] in parts of Scotland.

Clip 8: 20th-century Cumbrian ‘gius’ (‘goose’), and my attempt at modern Scots ‘gis’.

Jordan (1974) places this northern sound change in the 1200s, so that the northern long vowel inventory by 1300 looks like this:

Clip 9: The vowels of northern Middle English c.1300, where GOOSE has fronted. Here, I have putatively included TAKE as a separate vowel from BOAT, although they may have been merged across a lot of the north.

It’s clear from modern northern forms that this GOOSE vowel fronted at some point, and the evidence for this happening prior to 1400 is that it began to be rhymed with the <u> vowel that had been loaned from French, which was probably pronounced [yː]. In my video, I stuck to an account in which this fronted GOOSE remained [øː] until the great vowel shift, at which point it raised to [yː] (in accordance with all the other long vowels raising by about one ‘cardinal position’ on the IPA chart) - but Jordan (1974) suggests that these rhymes with French loans imply that it had actually already raised to [yː] by the time of the great vowel shift.

I’m not sure which of these is more likely, but I might now lean towards the latter interpretation (which is not the one I presented in my video), in which the BOAT and GOOSE vowels are already respectively [ɛː] and [yː] by 1400, and have diphthongised at least to [ɛə] and [ɪə] by 1700. This is something that my opinion has flipped on over time, and I hope to follow this up with another article on that specific subject soon, but this is getting quite long and I’d like to post it to coincide with the video, so I will end this here.

Thank you very much for reading :) Click below if you’d like to subscribe, but even if not, I hope you enjoyed the post!

Jordan, R. 1974. Handbook of Middle English Grammar: Phonology. Translated and revised by Eugene Joseph Crook.

Brilioth, B. 1913. A Grammar of the Dialect of Lorton (Cumberland).