Talk:Great Vowel Shift/Archive 1

Request for Expansion on Causes of Vowel Shift

I would like to see this article mention a few of the main theories on the cause of the vowel shift, and also mention how researchers were able to pinpoint the time at which it took place, since it was well before the invention of recording devices. -- Aaron W.

This response may be way obsolete, but it's worth a try.

Theories of why sounds change in this or that way come under the umbrella term functionalism. None of the theories work (including, most pointedly, the one stated below; which see). The ones that work best are very vague, such as the observation, repeated below, that stress-timed languages, e.g. English and the other Germanic languages, seem to have very unstable vowel systems (not just the long ones), whereas as syllable-timed languages, e.g. Japanese, Malayalam, and Spanish, tend to have very stable ones. Now of course that only points to the really interesting question, which is why -- and for which there seems to be no good answer.

As for the dating, the evidence is complicated and sketchy, having to do with such things as the appearance of spelling confusions, and the treatment of loan-words (whose arrival in written records can be dated) such as the failure of /ī/ to diphthongize in borrowings like redeem (first attested 1425 but then not again for 110 years), esteem (first attested 1528), breeze (first attested in the late 16th cent.) (examples from Jespersen). Had these words been borrowed earlier, they would have undergone the vowel shift, as did say devise (first attested in English around 1300), derive (first attested 1483), and so on. It's a pity that we don't have orthoepic manuals from Chaucer's day, but the social developments that inspired their creation didn't arise until the late 16th cent. Indeed, in all probability regional variation in the expression of the Great Vowel Shift was a major stimulus. Alsihler 19:39, 16 December 2006 (UTC)

Great Vowel Shift and Continental languages

Could the Great Vowel Shift explain why native English speakers often find foreign languages difficult, in the sense that it knocked English vowels out of sync with the Continent?

Yes, that's what I thought reading this article : according to the examples, the original pronunciation was very close to the French or German one. SeeSchloss 21:26, 4 Aug 2003 (UTC)

I think we're confusing spelling and language. The phonemic values of the vowel symbols and digraphs (like ea, ee, and ou) in English are indeed mostly different from the usual continental values of the symbols, even when it doesn't appear so: the phonetics of, say, the vowel of wrote are quite remote from, say, the phonetics of German rot "red": the English vowel is actually [əw] (America), [εw] (Brit.) But I would say that learning new values for the vowel symbols is at best a bit of a nuisance at the very outset, like learning that <j w> in German stand for /y/, /v/. The far bigger problem in a monolingual English speaker learning just about any foreign language is the complex nature of our vowels. The details differ from region to region, but our long vowels, mostly the reflexes of Middle English long vowels, are full of raising offglides, many of the short vowels have more or less pronounced inglides (i.e., toward [ə]), especially in environments of prolongation, as in the words bed fell jazz; /æ/ before /g/ and /ŋ/ have a pronounced raising & fronting glide (the nucleus of bang bag is very different from the nucleus of bad ban -- in some dialects it's gone on to become [e:], with no glide). /æ/ and /ɔ/ (which don't occur in many languages anyhow) are in many varieties of English decidedly long, and have offglides toward schwa, with or without raising of the onset, that is [æə] to [eə] or even [ɪə] in a word like cat. (And of course once you get into r-less dialects, things get even weirder, with say bad, bared, beard being totally homophonous). By the way, the offglides typical of present-day English long vowels considerably postdate the Great Vowel Shift itself (always excepting /ay aw/ < /ī ū/, of course). It might be added, too, that offglide features are found in other languages. In Karo Batak, from Sumatra, the vowel system is basically a sort of Latin Five, but both mid-vowels have an appreciable (if variable) upglide in final position especially. Alsihler 21:15, 31 January 2007 (UTC)

This is a minor point, but FYI, <j> in German actually stands for /j/, not /y/. (/j/ is the sound of English <y>; /y/ is the sound of German <ü>.) —Ruakh_TALK 03:02, 1 February 2007 (UTC)

I think I can answer this with an unequivocal no. I am not a native English speaker, and I found the problems of dealing with differences in phonetics negligible compared to other problems involved in learning a different language. In fact, I think the problems of dealing with differences in the relation of spelling and pronunciation is worse for foreigners trying to learn English than for native English speakers trying to learn another language. I suspect that the main problem for native English speakers when learning continental languages is grammar. English being grammatically a very simple language, the English are quite unprepared for dealing with such complications as (formal) gender, declensions, conjugations, etc. Grotendeels Onschadelijk 02:16, 14 April 2007 (UTC)

English grammar is anything but simple. English is however, a great language to speak poorly. Jalwikip 07:40, 20 June 2007 (UTC)

From Vowel shift talk page

Comment moved from Vowel shift which redirects here

(An article on the Great Vowel Shift exists already. One on vowel shifts generally should appear here. Some specific examples could also be listed (e.g., what are the similarities and differences between the vowel shift that the Greek language underwent, and that experienced by English?).)

the last link is busted

The link to lancaster is broken:

http://www.ling.lancs.ac.uk/chimp/101/vowels/shift.htm

it comes up saying I don't have permission. I recommend the link be deleted.

Hwarwick

Clarification on table of vowel shifts

Knowing nothing of phonetics, I couldn't make out much of the table of shifts. After a bit of searching I decided the symbols were SAMPA so I've said this. Am I right? Also, people who didn't learn Latin would find it useful to have approximate vowel sounds for before as well as after. Some, but only some, are given in the paragraph following the table. For example, I still can't make out what /o:/ is. Thincat 16:13, 16 Dec 2004 (UTC)

Break and beak are also listed as one vowel sound in modern English. That's just not the case.

I don't think you read the passage carefully (unless it's been touched up since the complaint was made): to amplify the statement, earlier [ɛ:] (often spelled in late ME -ea-) raises to fall together with the outcome of [e:], namely [i:], as in beak. But that in some words, such as break, it remains a mid vowel. --All true, but in a discussion like this it strikes me as unwise to muddy the waters by bringing up "exceptions" (there are only three forms like break, the others being steak and great -- and Gay rimes great and cheat in The Beggars Opera). The problem is that in a "compromise dialect" like that of standard (south-midlands-based) English there are "exceptions" by the dozens to all these sound laws. It makes a bad impression on the non-linguist, but there's nothing the least bit remarkable about such raggedness in compromise dialects.

Oh, and [o:] (a long higher-mid back rounded vocoid) would be like the pronunciation of stone with an Irish accent (i.e. monophthongal and very peripheral), or the vowel of German Floh "flea" or French beau "handsome". Alsihler 18:30, 16 December 2006 (UTC)

I second the request for elucidation of the table, and for 'before' examples as well as 'after' ones in the par following—the phonetic symbols may as well be in Aramaic for any intelligibility they have for me, or the average reader. In fact, I don't even understand Alsihler's response to the points raised by Thincat—and would really like to.—Zoe Ocean 06:08, 4 November 2008 (UTC) —Preceding unsigned comment added by Zoe Buchanan (talk • contribs)

Statement in vowel shift in German and Dutch

The statement on vowel shift in German and Dutch, concerning the pronunciation of the word for "ice" ("Eis" and "ijs" respectively") is true only for German. "Eis" is pronounced as /ai/, but "ijs" in ABN (Algemeen Beschaafd Nederlands, or General Civilized Dutch) is pronounced [ei], the modern pronunciation of the Dutch-Specific "Long IJ".

Really? I mean, I suppose I ought to take your work for it, since I don't speak Dutch; but the times I've heard Dutch spoken, the sound of "ij" has always seemed to me to have a low nucleus, much closer to [æi] than [ei]. AJD 01:39, 15 July 2005 (UTC)

It is. Convention has it that it is pronounced [ɛi], but it's closer to [æi] I think. See also the Dutch phonology page, which is otherwise a bit of a mess, but the IPA charts are ok. Jalwikip 07:48, 20 June 2007 (UTC)

In Standard Dutch, ij and ei are homophones. In regard to spelling: Dutch spelling was not standardised until the nineteenth century. Before that spelling was typically phonetic and would depend on the writers dialect. There are some exemptions, mainly in place names and surnames, but also a few words: in the word 'bijzonder' the ij is pronounced [i].

In low-saxon, the ij is pronounced i. That is why kids in eastern Netherlands knew when to write ij or ei, whereas kids in western Netherlands often made mistakes. X10 05:56, 28 June 2007 (UTC)

On another note: why is Icelandic named as having participated in a similar shift? As far as I know, they kept their í and ú pretty sraightforward. The only thing resembling the Great Vowel shift is ó being pronounced somewhat like [oU].

??Didn't /ā/ become [aw] in Icelandic? Alsihler 16:59, 15 December 2006 (UTC)

The Ultimate Cause of the Great Vowel Shift

Note: My vowel charts don't show up correctly in this note. This passage still has the answer that you all seek.

The social factors presented in this article only tell why the sound change became so popular. Naturally a large city such as London would have had extensive influence over neighboring dialects through trade and education.

The real cause of the Great Vowel Shift is a linguistic principle in which a language may only have six different vowels of a particular set. This is due to the fact that vowels seek to differentiate themselves from other vowels for sake of clarity in pronunciation. Diphthongs are their own set and a language may distinguish its pure vowels either quantitatively (by lenght) or qualitatively (by tenseness and laxness.)

First some notes on vowel systems. There are three different sorts of languages with regards to vowel systems:

-Those that distinguish long and short vowels (Middle English, Latin, German.) These languages may also have tense and lax vowels, however, for sake of the rule of six, only long and short sets of vowels are distinguished from one another. One might say the long and short distinction supercedes the tense and lax one. More on that below.

-Those that distinguish tense and lax vowels (English, French, Italian.) In most instances, inquiring individuals are apt to comment that English does have a distinction between long and short vowels as well. Those that are designated long and the others short are due purely to conventions established in Middle English. English does in fact have long and short vowels, however, they are in complementary distribution and thus subphonemic (i.e. linguistically insignificant.)

-Those that do neither (Spanish.) There is the technicality in the Spanish language that they do have the lax vowel /a/, however, this perception is mostly prevalent in European languages. In reality /a/ is not realized as either tense or lax cross-linguistically.

In Early Middle English the vowel system was as follows (capitals are used throughout to signify lax vowel for simplicity, also @ = schwa):

i:, I u:, U

e:, E:, E @ o:, O:, O

æ A

Old English æ: became E: and a: became O: in all instances.

If we look at this sytem, we can see from the above descriptions that it is based on a long and short vowel distinction. Thus the two sets of vowels are:

[i:, e:, E:, u:, o:, O:] and [I, E, æ, A, U, O] (@ is a reduced vowel only occuring in word final syllables.

As we can see the rule of six is not broken at this point. There are six long vowels, six short vowels, and one reduced vowel. So what pushed it over?

In the thirteenth century a sound change took place whereby a non high (i.e. not I or U) short vowel became long in disyllabic words when in the first syllable if that syllable were open (i.e. when, most typically, followed by a single consonant.) Thus:

nA-m@ 'name' became nA:-m@

E-t@n 'to eat' became E:-t@n

chO:-sen 'we/you all/they chose became chO:-s@n

At this point, the E and O in this position merged with the original E: and O: phonemically (i.e. they were now understood as the same sound.) A:, however, was not realized phonologically because it occurred in complementary distribution with A. Thus, it occurred only in the first syllable of disyllabic if that syllable were open and A occurred elsewhere.

What caused the system to topple was this:

All Germanic languages are known for their strong word initial syllable stress. This is especially so in more northern Germanic languages, such as English. In prehistory, this strong word initial stress caused short vowels to disappear and long vowels and diphthongs to become short in word final (i.e. unstressed) syllables.

This trend continued in Middle English. Word final /n/ and /@/ disappear following the vowel changes described above. Because of this development A: and A are no longer in complementary distribution.

The A: in nA:m 'name' (<nA:-m@) occurs in the same phonetic environment as the A in nAm 'I took.' Due to this these two sounds became different phonemes and the Middle English vowel system now contained 7 long vowels and 6 short vowels.

Since the vowels could not properly distinguish themselves by the rule of six, i: and u: (both high vowels) "jumped ship" and became the diphthongs @j and @w. As the article states, the remaining vowels shifted upwards. The article erroneously mentions A: becoming fronted going to ej. Although it doesn't fit the classical model of vowel systems, the vowel A is in actuality a front vowel.

Classic View of Vowel Systems with Regards to the Position of the Tongue in one's Mouth

i, I u, U

e, E @ o, O

æ A a

Modern View of Vowel Systems with Relation to the Tongue's Position in the Mouth

i u

 I         U
e     @     o
 E          O
  æ
   A        a

The discrepency comes from more advanced technology better determining the position of one's tongue during the production of each vowel respectively.

Because each vowel can't shift into the position of the vowel above it one may construct a chronology of the vowel changes as follows.

Great Vowel Shift

Step one: i: and u: became @j and @w respectfully Step two: e: and o: became i: and u: Step three: E: and O: became e: and o: Step four: A: became æ: Step five: e (< E:) became i: Step six: æ: (< A:) became e: Step seven: e: and o: became ej and ow (British @w)

All long vowels eventually merged with short in length due to their qualitative distinction. Hope this helps!

If this is the case and this is not original research (see Wikipedia:No original research), why don't you give some citations for this theory, noting that it is one theory among others, and work it into the article instead of just adding this to the talk page? After all, you seem to be knowledgable in the subject, and Wikipedia would be happy to have your contributions. —Lowellian (reply) 07:47, September 2, 2005 (UTC)

That doesn't seem to me to be true — French, for example, has sixteen vowels (not counting the halfway-in-between vowels that often appear in unaccented syllables), and while some of these are in fairly complementary distribution, most are not (and I think it would be hard to argue that the fairly complementary ones are simply allophones). How would you divide up /i/ (y), /e/ (et), /ɛ/ (est), /ɛⁿ/ (hein), /a/ (a), /ɑ/ (as), /ɑⁿ/ (en), /u/ (ou), /o/ (au), /ɔ/ (as in homme), /ɔⁿ/ (on), /y/ (eu), /ø/ (eux), /œ/ (as in heure), /œⁿ/ (un), and /ə/ (as in je)? — Ruakh_TALK 23:16, 28 November 2006 (UTC)

It doesn't work for me, either and I am just as happy that this is NOT worked into the main article. It's very mystifying. I'm unaware of any "linguistic principle" limiting the number of vowels to six (by "particular set" I'm guessing that you mean any group of syllabics distinguished by the same cavity features, e.g. all short oral vowels. It's not clear whether front rounded vowels would belong to the same "set" as "the Latin Five"). English, or rather my English, has seven short vowels: pit pet pat pot put putt bought, and dialects of English have an eighth in a distinction between the syllabics of four, hoarse vs fore and horse. French has eight short vowels of the ordinary kind, by my count, plus a reduced vowel and three front rounded ones, for a total of ten full short vowels (though to be sure the contrast between /a/ and /ɑ/ seems to be on the run; and at least in some environments, the contrast between /e/ and /ɛ/). A seven-vowel system seems to be the industry standard for sub-Saharan languages (two high vowels, four mid ones, and /a/).

Besides, the theory in question does not in fact have any obvious bearing on why all the vowels shifted. An overcrowded vowel-space would be simply dealt with by merging two or more vowels (as happened earlier, with the loss of three front rounded vowels and later when two products of the Shift, /ī/ and /ē/, merged as /ī/) and anyway, after all that commotion, there were still seven long vowels (which to be sure shrank to six before long, perhaps validating the Rule of Six, but I doubt it). What a lot of trouble to go to for nothing!

Another problem is that the Great Vowel Shift isn't an it, it's a they. The details differ quite a bit in the north. In Scottish, for example, /ū/ did not diphthongize. This did not bottle up /ō/, however; it didn't merge with /ū/, as it might have, instead it fronted, first presumably to [ö:] and then, because all phonologies with only one front rounded vowel have only a high one, to modern /ü/, usually written as bluid, duir for "blood, door". And I also have the impression that in the north, OE ā (from Proto-Germanic *ay) did not in fact raise and round, as in the south, but fell together with the outcome of lengthened ă (as in name). Hence pronunciations like hame, stane, gait for southern home, stone, goat, etc. This may be something else entirely: conventional wisdom takes such forms to be Scandinavian loans, reflecting the Scandinavian treatment of *ay. I'm not at all convinced that this is the case. There's too much of it, it seems to me, and it's too heavily concentrated in words with one particular vowel. But if the falling-together interpretation is correct, there goes the "trigger" envisioned for the shift. Alsihler 20:15, 16 December 2006 (UTC)

Swedish has eighteen distinct vowel phonemes, set on the long-short axis: a, o (/ou/), u (deep /u/), å (roughly a middle rounded /o/), e, i, y, ä, ö - long and short varieties of each and these are not in complementary distribution. Plus, tonal stress accent may alter the meaning of a word which is spelled and (apart from the accent) pronounced the same way (though in practice this affects only a limited number of words), e.g. anden - with a rise on the second syllable it means "the spirit", with a quick rise-drop in the first syllable and no rise later on it's "the duck"! So the "maximum six vowels" rule makes no sense in terms of general linguistics. /Strausszek (talk) 05:44, 4 September 2009 (UTC)