[Spellyans] The two roles of â (a-circumflex) in Kernowek Standard

Michael Everson everson at evertype.com
Wed Sep 18 16:05:59 BST 2013

On 16 Sep 2013, at 11:30, Philip Newton <philip.newton at gmail.com> wrote:

> I must say that I find the dual role of "â" (a-circumflex) in KS rather confusing. Not only does it indicate anomalous (or unpredictable) length, as in _fâss_ or _shâp_, but it also indicates an optional pronunciation with
> [ɒ:] ([Q:]) as in _brâs_.

It has three roles. It also serves orthographically to distinguish "â" 'goes' from "a" (particle), and "dâ" 'good' from "da" (variant of "dha").

> I found this confusing because when I see the symbol, I keep having to wonder which particular meaning is intended.

Perhaps, but you are expected to learn. It isn't actually very difficult. And the number of words which use it is not very large. 

> I suppose this does account for words such as _brâs, clâv_ since those are long anyway, so the â cannot mark length.


> And I suppose it even makes a certain amount of sense in derived words such as _brâssa, clâvjy_ (where the "æ/ɔ" is short since it's a two-syllable word, but the spelling with "â" is kept from the monosyllable as with the "i" in words such as _tir, tiryow; gwir, gwiryoneth_).

In words where the vowel remains long in the polysyllable, the diacritic is retained: "shâp", "shâpya". 

> Even if it is a bit disconcerting to me that _brâssa_ and _fâss_ look so similar yet the â has different functions in both. (So you always have to know whether it's a derived word or not.)

Again, there are not many of these words, and most of them are of high enough frequency that you are expected to learn them. 

> However, in the KS edition of _Skeul an Tavas_, I see in the pronunciation section that _âls_ is to be pronounced "[ælz]~[ɒlz]".

Also "fâls" [fælz]~[fɒlz]

> And that is a situation where this breaks down.
> Followed by two consonants (that aren't "st"), an "a" in that position must necessarily be short.

This isn't difficult. We might have to state explicitly the rule to indicate that these two words in "ls" there is an alternation in pronunciation though not in vowel length. (Even if one were writing "fåls" and "åls" there would be anom

> Hence, an "â" in that position must signal unpredictable length… except that in this word, apparently, the vowel
> stays short but may have an alternative quality.

As in "brâssa". 

> At which point, KS "â" seems nearly as capricious to me as SWF "y" (is _bys_ long or short? does _glyb_ rhyme with _ryb_? etc.).

You are over-stating the case; moreover, "â" marks distinctions which are ignored in the SWF. If there are exceptions, one must judge them by their difficulty and their number. 

> Is it possible to resolve this part of KS in a way which enables one to predict the pronunciation of a word unambiguously given the spelling? Whether by splitting the two current roles of "â" or in some other way?

No one ever claimed that KS orthography was entirely free from occasional ambiguity. The question, again, is their nature and scope. 

> It might well be, given that much of the discussion was in 2008, five years ago. In which case this not-entirely-phonemic corner of the orthography must remain, and students encountering an "â" must continue to guess.

"Continue to guess"? At what point does one learn?

I am travelling and am not at home with my sources, but I did a quick look at words with "â" in An Beybel Sans. Out of the 18000 unique word-forms in the whole book, about 48 of them have "â". (That's 18,000 including distinctions between "Dâ" and "dâ", so the number of unique words is going to be somewhat less.)

Words with "â" can be divided into three categories; I give these with also a count of the number of instances each occurs in the book. Because of a feature in my concordance software, sometimes words are considered different because they have -ma or -na appended; the list below also gives a few possible errors in the Beybel where a diacritic has evidently been left off; I also do not guarantee that the list below is free of error. It is the scope of the "problem" that is the point.  

1) Orthographic "â" to distinguish these from other common words: (2)

â	(38)
Dâ, dâ, dâ-, dâ-ma, dâ-na, dhâ, dhâ-ma, tâ, tâ-	(70, 835, 2, 9, 2, 77, 1, 109, 1)

2) Vowels whose length would be expected to be short but which are marked as long: (33)

acâcya	(27)
agât	(1)
blâmya, blâmyas, blamys, vlâmya	(3, 1, 2, 8)
CÂF	(1)	35322
câken, câkys, gâken	(8, 8, 4)
câss, câss-ma, câss-na, gâss, hâss	(43, 4, 6, 10, 2)
consecrâtya, consecrâtys, gonsecrâtyas	(10, 7, 1)
crysoprâs	(1)	73358
debâtys	(1)
dylâtya, dylâtys, tylâtya	(5, 4, 2)
Ewfrâtes	(37)	234
fâmya	(1)
fâss, [Fassow], [fassow], fâssow	(399, 1, 9, 3)
flâtyow, flâtys	(1, 1)
Forsâk, forsâkya, forsâkyas, Forsâkyowgh, forsâkyowgh, Forsâkys, forsâkys	(1, 145, 59, 2, 1, 2, 29)
frâmys	(25)
frâs, [frasow]	(62, 1)
galargân, alargân-ma, galargânow	(5, 1, 1)
Grâss, grâss, grâss-ma, grâss-na, grâss-offrydnow, grâss-offryn, grâss	(8, 125, 1, 1, 2, 1, 56)
hâtya, hâtyas, hâtyoryon, Hâtyowgh, hâtys	(85, 2, 1, 2, 12)
lâss	(1)	5439
Nâ, nâ, comondment-ma-nâ, mans-nâ	(92, 17, 1)
pâss	(3)	17486
plânys	(1)	42712
plâss, plâss-ma, plâss-na, plassyow, pednplâss, blâss, flâss	(15, 2, 2, 2, 1, 2, 1)
plât, [plat], brestplât, plâtyow, plâtys	(17, [1], 31, 2, 1)
separâtya, Separâtyowgh, separâtys	(13, 1, 1)
shâmya, [Shamys], shâmys	(14, 1, 62)
shâp, Shâpya, shâpya, shâpyas, shâpys	(6, 1, 4, 2, 3)
spâss, spâss-ma	(52, 1)
stât, stât-na, stâtly, stâtys	(13, 1, 1, 2)

3) Vowels where "â" marks a dialectal vowel alternation, usually long but occasionally short in derivatives: (10)

âls, âls-ma, âlsyow	(15, 1, 20)
Brâs, brâs, brâs-ma, brâs-na, brâs-pàn, brâs-saw, Brâssa, brâssa, Brâster, brâster, brâster-ma, brâstereth, brâsyon, VRÂS, Vrâs, vrâs, vrâs-ma, vrâs, vrâssa, Vrâster, vrâster, vrâstereth, vrâsyon	(36, 962, 21, 3, 1, 1, 4, 87, 2, 33, 1, 4, 9, 1, 1, 216, 4, 15, 4, 15, 7, 9)
Cân, cân, Gân, gân, gân-ma, gân-na, [Ganow], Gânow, [ganow], [ganowow], hân	(33, 31, 1, 10, 11, 1, [1], [1], [196], [3], 3)
Clâv, clâv, clâv-, clâv-oll, glâv	(1, 84, 1, 1, 3)
Fâls, [fals], fâls, fâls-cabel, fâls-derivas, fâls-descadoryon, fâls-dùstuniow, Fâls-dùstuny, fâls-dùstuny, fâls-duwow, Fâls-esperans, fâls-fara, fâls-gober, fâls-gobrow, fâls-gwarek, fâls-gwreg, fâls-hunrosow, fâls-lavarow, fâls-profecy, fâls-profettys, fâls-prontyryon, fâls-venyn-na, fâls-vesyon, fâls-vesyons, fâls-was, fâls-wesyon, fâls-wonesyjy, fâls-yn, [falsa], [falslych], fâlslych, Fâlsury, fâlsury	(3, [4], 82, 1, 1, 1, 2, 1, 3, 4, 1, 1, 9, 4, 2, 1, 1, 1, 3, 1, 2, 1, 2, 5, 6, 7, 1, 1, 5, 5, 2, 1, 40)
Glân, glân, [glander], glânder, lân, lânder	(4, 136, 1, 2, 1, 3)
Gwâv, gwâv, gwâvas, gwâvy	(1, 16, 1, 1)
gwlân, wlân	(14, 5)
hâv, hâvy	(24, 1)
tâl, dâl, thâl, adâl	(15, 12, 8, 83)

Of these 10 words, 3 of them end in -ân (cân, glân, gwlân), 3 of them end in -âv (clâv, gwâv, hâv), 2 of them end in -âls (âls, fâls), 1 in -âs (brâs), and 1 in -âl (tâl)

I don't believe that this is too difficult for anyone to learn. 

Michael Everson

More information about the Spellyans mailing list