[Spellyans] SWF review results.

Craig Weatherhill craig at agantavas.org
Fri Apr 4 15:15:23 IST 2014


I strongly suggest that you send that to Jenefer and anyone else you can think of.

With regard to trad. graphs, the Treyarnon agreement has been ignored from the outset.  MAGA has published nothing in SWF/T, and ignored trad. graph and those who prefer to use them en bloc.  In the meantime, I hear terms like "Penwith separatists" (particularly for those who prefer to use Late variants, even if they're part of the main form), and "not SWF" being applied to trad. graphs.

Craig




On 2014 Ebr 4, at 11:41, Michael Everson wrote:

> On 2 Apr 2014, at 10:44, Michael Everson <everson at evertype.com> wrote:
> 
> Here are some comments I made on the Lytherennans Kernowek Facebook page regarding the “things recommended not to change”. It includes some changes. 
> 
> A Issues examined but where no change is recommended:
> 
> Issue: (2) <i> graph used inappropriately in prefixes and suffixes.
> Proposal for resolution:: Recommended: no change – retain the status quo.
> Scale of recognition: 1.8% (1 respondent)
> 
> Of course they don’t give any examples. I don’t know what this is about. 
> 
> Issue: (4) Distribution of <i> and <y> is unclear and incoherent
> Proposal for resolution:: Recommended: no change – keep the distribution as is.
> Scale of recognition: 7.1% (4 respondents)
> 
> This is fantastic. The distribution is unclear and incoherent — meaning you really can’t predict when you are supposed to use ‹i› and when ‹y› — and 7% of respondents criticized it. But the resolution? Leave it unclear and incoherent. 
> 
> Issue: (6+56) Use of <oo>, giving rise to unfortunate or ‘risible’ spellings.
> Proposal for resolution:: This was acknowledged to be a subjective issue.
> Recommendation: Retain <oo>. It is an umbrella graph for two pronunciations [o:] and [u:]. In addition too many words
> would be affected and its retention is in the interests of minimal change.
> Scale of recognition: 3.6% (2 respondents)
> 
> It is good that this was rejected. Doubtless it was an attempt to restore the unnecessary Kemmynism ‹oe›. And risible spellings? If a child laughs at “boos” or “poos”, well, then, it’s likely the child will REMEMBER those words. :-)
> 
> Issue: (7+19) Introduce minimal use of <z>.
> Proposal for resolution:: Recommended: no change – the distribution of voiced [z] is not the same for different periods and revived pronunciation systems. This issue might be reconsidered one day.
> Scale of recognition: 8.9% (5 respondents)
> 
> Again, I’m not sure what specific problem this was supposed to solve because the documentation was never released to the public. I asked many times and was never given an actual reason for the refusal. 
> 
> Issue: (9) Vocalic alternation - lack of a systematic and understandable rule.
> Proposal for resolution:: Recommended: no change, to leave as at Treyarnon; but we do recognise the need for further research on this issue.
> Scale of recognition: 8.9% (5 respondents)
> 
> Yes, well, I guess we’ve done this research. I’m not sure if there is a systematic rule for the alternation. It’s just part of the language. 
> 
> Issue: (12) Respelling of SWF/L <e’wedh>, <endella> and <e’mann> with initial <a>.
> Proposal for resolution:: Recommended: no change in order to keep correspondence between SWF/M <y> and SWF/L <e>.
> Scale of recognition: 1.8% (1 respondent)
> 
> Good that these were rejected. I do think that the overuse of ‹e› especially in unstressed syllables helps to ghettoize RLC unnecessarily. 
> 
> Issue: (16) Traditional graphs.
> Proposal for resolution:: Recommended: no change, The Guiding Principles for the SWF Review which were agreed by the Partnership stated that "issues that were closed in 2009 should not be reopened unless the solution found then is now causing problems." It was therefore clearly understood by the Board that they were not to re-consider those issues that were fundamental underpinning issues of the 2009 Treyarnon agreement. The use of Main Form Graphs and the status of Traditional Graphs were undoubtedly such fundamental issues, “closed in 2009", which were therefore outside of the remit of the Panel. However, the Board did, of course, discuss them and if they had been within the Board’s remit, it would nevertheless have been their recommendation to restate the position agreed at Treyarnon.
> Scale of recognition: 3.6% (2 respondents)
> 
> But the “solution” of 2009 *DOES* cause problems, clearly. The solution is easy: Allow every person to choose which graphs he or she prefers. 
> 
> Issue: (24) Reduction of unstressed vowels to schwa.
> Proposal for resolution:: Recommended: There is no issue of spelling to be resolved here, but the SWF specification needs to be re-written as the wording is ambiguous and implies that all final unstressed vowels became schwa;
> Scale of recognition: 1.8% (1 respondent)
> 
> This is worrying, but of course we are given no details. 
> 
> Issue: (27) Diacritics
> Proposal for resolution:: Recommended: NO diacritics EXCEPT in pronunciation guides, dictionaries and teaching materials if the author so wishes. We are seeking a meeting with the Dictionary Board to ensure that such diacritics as are chosen are properly defined and used consistently.
> Scale of recognition: 1.8% (1 respondent)
> 
> Once again, we are betrayed by those who hold a grudge against us. This is remarkable. It’s worth my repeating here what I said in my submission to the Review:
> 
> =====
> The SWF does not permit the valid use of diacritical marks. This means that any judicious use of diacritical marks renders a word “not compliant” with the SWF. Prior to the publication of the SWF specification, the “agreements and rulings” document contained the following text: “Diacritical marks are not part of the SWF. (Publishers can add some if they think that they are necessary for pedagogical reasons.)” This was a slightly watered-down version of text which had been proposed for use as a footnote in the specification: “Diacritical marks are permitted to be used, optionally, to mark words with anomalous vowel length or quality.” 
> 
> Since the SWF has ambiguous spellings, it is reasonable to permit such spellings to be distinguished by those who feel that it is important that an orthography offer as much support to learners as possible. Accordingly, a ban on diacritical marks makes no sense. 
> 
> One advantage of the use of diacritical marks is that it leaves the underlying word-shape alone. The problem of inclusiveness between “Middle” and “Late” dialects of Revived Cornish is easily seen in the large class of words with an alternation bÿs~bës [biːz]~[beːz]. There are many stressed monosyllables with [iː] which do not have an alternation (like mis [miːz]) and there are also words with a short vowel (like bys [bɪz]), marking this class enables readers of “Middle” dialects to accurately identify words written in the “Late” dialect and vice-versa. (Another means of inclusively and accurately marking this class would have been to use an umbrella graph, such as beis.)
> 
> Where UC and UCR were ambiguous as to whether u meant /y/~/i/ or /u/, in the SWF the ambiguity has been shifted, so it is often not certain whether o means /o/ or /u/. Words like arlùth should not be written *arlodh or *arloth, and here the grave accent can indicate that the vowel is a reduced /u/ vowel, not a reduced /y/~/i/ vowel. For this class of words, the new spellings in o introduced into the SWF should be rescinded and the rules for the graph u should be that û indicates [uː], ù indicates [ʊ], and that u indicates /y/~/i/, length determined by the usual rules. This differs from UC and UCR, which wrote u for /u/ and ü for /y/~/i/, but in order to reduce the number of potential diacritics in a given text, it is better to mark /u/ because this is a much smaller class of words than those with /y/~/i/. (It is also easier because no allowance has to be made for distinguishing short ü [ʏ]~[ɪ] from long ü [yː]~[iː].)
> 
> Diacritical marks should permitted to be used, optionally, to mark words with anomalous vowel length or quality. If diacritical marks are not permitted, it automatically knocks any one who does use them out of compliance with the SWF spec. This has already been seen to have been used in a prejudicial way against publications in Kernowek Standard. Now KS may have derogated in other ways here and there from the SWF for various reasons, but for instance marking bÿs~bës words should not be considered to be a violation of the specification. On 2008-02-28, in an e-mail discussion between Albert Bock, Benjamin Bruch, Michael Everson, Trond Trosterud, and Nicholas Williams, the following text was agreed. It was subsequently not included in the SWF specification (despite it having been agreed) and no explanation was ever given for its removal. We request its reinstatement:
> 
> “Diacritical marks are not a part of the mandated SWF orthography. However, publishers are permitted to be use them, optionally, to mark words with anomalous vowel length or quality.”
> 
> We also recommend that if diacritical marks are to be permitted, then the system for using them should be that recommended and implemented in Kernowek Standard. Amongst other things, An Beybel Sans is now a fact of Cornish public life, and users of all Cornish orthographies are likely to encounter it. If the bÿs~bës words are to be marked, they should be marked with the diaeresis, not written * bŷs~bês or something else, since the former spelling has been established in a large number of publications.
> =====
> 
> So instead of just permitting people to use them, the Review Board has made a statement (an unenforceable statement): “ NO diacritics EXCEPT in pronunciation guides, dictionaries and teaching materials if the author so wishes. We are seeking a meeting with the Dictionary Board to ensure that such diacritics as are chosen are properly defined and used consistently.” 
> 
> I wonder whether they will really screw up the Revival by choosing to use diacritical marks in a DIFFERENT way from the way they are used in KS. 
> 
> Issue: (28) No phonemic distinction is made between /iw/ and /Iw/; <iw> should be replaced by <yw>
> Proposal for resolution:: Recommended: leave it as it stands; this is important for those who make the distinction in Kemmyn pronunciation.
> Scale of recognition: 3.6% (2 respondents)
> 
> Except that nobody makes this distinction in Kemmyn pronunciation. 
> 
> Issue: (30) Inconsistent treatment of pre-occlusion: <jynn> & <gonn> etc.
> Proposal for resolution:: Recommended: no change – in items of which there are no attested RLC spellings, regular development of /N/ > /dn/ is assumed. <gonn> is not used in RLC, where [g]oram is used instead.
> Scale of recognition: 5.4% (3 respondents)
> 
> A total failure of understanding of the problem of lack of inclusivity for RLC users.
> 
> Issue: (31) M <ew> vs. L <ow>: pronunciation difference.
> Proposal for resolution:: This change happens within Middle Cornish (MC), not just Late Cornish. <ow> is universal, <ew> is earlier MC only.
> Recommended: retain <ew> as the M form and <ow> for both M and L.
> Scale of recognition: 3.6% (2 respondents)
> 
> We mark three classes of words: ‹ew›, ‹ow›, and ‹êw›~‹ôw›. The Review Board doesn’t even talk about classes of words. 
> 
> Issue: (32) Initial ye-/e- alternation in words like yehes~ehes not handled in a unified way. Learners unable to tell
> whether a RMC word in <ye-> has a RLC variant in <e->.
> Proposal for resolution:: Recommended: no change – retain the status quo.
> Scale of recognition: 1.8% (1 respondent)
> 
> Of course the status quo would be solved by permitting this class of words to be written ‹yê-›~‹ê-›
> 
> Issue: (44) <junya> - confusion over the vowel and pronunciation.
> Proposal for resolution:: Recommended: retain <junya> and teach pronunciation. Allow spelling <jùnya> in pronunciation guides etc. (see below).
> Scale of recognition: 1.8% (1 respondent)
> 
> Here we have the first glimmer of hope: Permit the use of ‹ù› as in KS. 
> 
> Issue: (46) Issue with the use of <eu>; for the small group of words always spelt with <o> in the manuscripts. Applies to seulabrys>/< seuladhedh>.
> Proposal for resolution:: Recommended: no change; remain with the present spellings for both SWF/M and SWF/L. More research is needed.
> Scale of recognition: 1.8% (1 respondent)
> 
> There is no warrant for ‹eu› in this class of words. 
> 
> Issue: (53) Final <i> / <ei> variants.
> Proposal for resolution:: Recommended: Remain as is. It would be beneficial to resolve this into one grapheme, but difficult to see how. Scale of recognition: 1.8% (1 respondent)
> 
> We write ‹-y› and permit ‹-ei› where necessary. This small class of high-frequency words is not problematic. 
> 
> Michael Everson * http://www.evertype.com/
> 
> 
> _______________________________________________
> Spellyans mailing list
> Spellyans at kernowek.net
> http://kernowek.net/mailman/listinfo/spellyans_kernowek.net





More information about the Spellyans mailing list