[Spellyans] SWF review results.

Michael Everson everson at evertype.com
Wed Apr 2 10:44:04 IST 2014


Here are some comments I made on the Lytherennans Kernowek Facebook page regarding the “recommended changes":

Some SWF Review results have been made public:
DOUBLE CONSONANT SPELLINGS.
1. Introduce <ll> and <rr> in roots. Do not revert to single letter in unstressed non-final syllables where that changes the root.
2. Introduce <ll> for <dell>, <-ell> for the suffix which implies 'tool' or 'device', e.g., dewynnell, draylell, gwariell, hornell, karrigell, musurell, pibell, rostell, skitell, skubell, skwychell, tempredhell, torgentrell, yeynell.
SHORT <o> and <u>
1. Change to <o> in <arlodh>, in line with <toll>, <tomm>, <boka>, <bronn>, <koska>
<GH> and <H>
1. <gh> everywhere except word-initially.
PARTICLE <th> in SWF/L
1. Join particle with verb as in <thera>, without a gap.
PARTICLE (ow>- RLC VARIANT <o>
1. Retain <o> (present participle), but drop the apostrophe.
<NOWYTH>
1. To be respelt <nowydh> (M), or <nowedh>.
HYPHENS
1. Remove hyphens from numbers, particles and some other words, e.g. <ebost>.
2. Add hyphens in loose compounds of the form 'noun + verbal noun', e.g. <jynn-skrifa>, <skath-wolya>, etc.
3. <poslev/poslevow> to replace <poos-lev/poos-levow>.
4. <erbynn> to replace <er-bynn>.
(but personal use to remain optional).
VOWEL LENGTH IN MONOSYLLABIC LOAN WORDS ENDING -p OR -t.
1. To be written as <pp>, <tt> if they contain a short vowel.
2. Accept both permissible plural endings, e.g. <hattys/hattow>
DHYWORTH/DYWORTH
1. Accept <dhyworth> and <dyworth> as alternatives, and <dhort> and <dort> in RLC variants.
A/OA ALTERNATION
1. Remove RLC variants <toal>, <kloav>, <gwoav>, <gwloan>, spelling them only with -a-.
2. Add <boas>, <doas>, <moas> in RLC variants to the remaining <broas>.
<OO> for KK <OE>
1. <goolan> to become <golan>. Otherwise, no change.
<kk> AND <ck>
1. <ck> to be used in all borrowings; <kk> in all other cases, including those in doubt.
2. <okkupya> to be respelt <okupya>.
APOSTROPHES IN RLC VARIANTS
1. Retain before nouns like <'manyn>.
2. Do not use elsewhere where the beginning of a word has been shed, e.g. <skydnya> from <dyskynna>.
3. Do not use internally, e.g. <metern> from <myghtern> and <mos> from <mowes>.
4. Do not use after <kal> (KK <kalgh>).
5. May be used where -th, -dh is dropped, e.g. pro', for', warbar'. To be left to the discretion of individual writers.
<eu>
1. Add SWF/L <enkladhva>, SWF/M <ynkladhva> (presently SWF <ynkleudhva>).
2. Allow <teudhi and <tedha> as alternative spellings.
3. Add to SWF/L <bidhi> (currently <beudhi> only); and <briji> (currently <breuji> in SWF/L).
No change to <breusi>, <Meurth>, <feusik>.
GWR & GR (SWF/L VARIANT)
1. Write <gwr-> in RMC and RLC, inclusing <gwra>, <gwruthyl>.
GEMINATION & PROVECTION IN SUPERLATIVES/SUBJUNCTIVES
1. Retain <tth>, introduce corresponding <ggh>, <cch>, <ssh>.
WORTIWEDH, ETC.
1. <wortiwedh/wostiwedh> and <wostalleth/wortalleth> to be single words, without apostrophes.
1ST PERSON SINGULAR
1 RC variant <me>, <ve>, no longer required.


A spectacularly lousy set of amendments, lacking any sense of structure. Indeed, a number of them will have to be repealed, as they directly contradict the agreement made at Treyarnon.

Since no one in the revival says [dʒɪᵈn] or [ɡɔᵈn] it is absurd to make this assumption. Whether <gon> was lost in late Cornish is irrelevant to the complaint, which is that the inconsistent linkage of nn/dn and mm/bm is a source of potential error and therefore a violation of the principle of inclusiveness. Whether "gonn" is used in RLC is irrelevant. What is relevant is that RLC users will read it as "godn", which is an error. And as a late loanword it is completely ridiculus to posit a "regular development" to *[dʒɪᵈn]; that didn't happen, and positing it is in this instance an example of conlanging.


"1. Introduce <ll> and <rr> in roots." What does this mean? What roots? Why? The current rule is that monosyllables in -rr or -ll have a short vowel, and monosyllables in -r or -l have a long vowel. An example in KS is instructive: <war> [wæːɹ] 'beware' has a long vowel, and <wàr> [wɑɹ] 'on' has a short vowel. Is the SWF going to write *<warr> for the latter now?

Other than that I take it that the alternation (attested in the texts) dallath/dalathow will now be dallath/dallathow. This is no improvement.

"Introduce <ll> for <dell>." This is fine. KS already did it. The alternative would have been <dèl>.

"Introduce <-ell> for the suffix which implies 'tool' or 'device’." This is an absurdity. If I am not mistaken only one of these "tool" words is attested, and then with a single -l. This is not a valid orthographic morpheme that diserves special spelling. In fact the spelling will probably attract erroneous final stress in these polysyllables. This is an unnecessary Kemmynism.

"Change to ‹o› in ‹arlodh›, in line with ‹toll›, ‹tomm›, ‹boka›, ‹bronn›, ‹koska›." This makes no sense. The vowel in ‹arlùth› [ˈɑɹlʊθ] is unstressed and therefore is not in the same class with the others. In KS we write ‹toll› [tɔl] 'hole', <tomm>~<tobm> [tɔm]~[tɔᵇm] 'warm', ‹bùcka› [ˈbʊkə] 'goblin', ‹bronn›~‹brodn› [bɹɔn]~[bɹɔᵈn] 'breast', ‹củska› [ˈkʊskə] 'sleep'. Củsca and bùcka have a different vowel from the rest of these. No improvement here.

"<gh> everywhere except word-initially." Guaranteeing another generation mispronouncing [ˈbiːən] as [ˈbɪkən]. The morphophonemic alternation found in SWF1 and in KS is better than this unwarranted Kemmynism.

"Join particle with verb as in <thera>, without a gap." This fails the inclusivity test. Writing ‹th era› with a gap was bad typographically. Writing ‹thera› without a gap is bad morphologically. Writing ‹th'era› as KS does correctly assists non RLC readers in recognizing the relation between ‹th'› and ‹yth› while avoiding the typographic hypercorrection that could occur in ‹ˈth era›. This is a splendid way of ghettoizing the late variety of Cornish. 

"Retain <o> (present participle), but drop the apostrophe." Again, this violates the principle of inclusivity. There's no reason RLC cannot either write ‹ow› like everyone else and just elide it in pronunciation, or just drop it in writing and provect where necessary. Writing ‹o'› was unnecessary. And writing ‹o› will surely confuse it with the verb. Once again, a splendid own goal by the RLC, which will marginalize text in that variant because it emphasizes something that makes no difference.

"To be respelt <nowydh> (M), or <nowedh>." Nearly every speaker of Revived Cornish has voiceless final consonants in final unstressed position in polysyllables. This orthographic decision simply underpins an unnecessary and unimplemented Kemmynism.

The hyphenation rules proposed are so incoherent that there is nothing to be said. The rules in KS are clear, and provide a better model for anyone (since personal use rules here).

"To be written as <pp>, <tt> if they contain a short vowel." THIS IS A VIOLATION OF THE AGREEMENT MADE IN TREYARNON. The rules are simple: In stressed monosyllables the vowel is short before a voiceless consonant and long before a voiced consonant (at least for the stops as here). Therefore the ONLY words which are problematic here are words like ‹shâp› [ʃæːp] where the vowel is unexpectedly long. What this rule has done is revert to the incorrect half-length rule of KK, where ‹hat› would be [hæːt] and ‹hatt› would be [hæt]. The correct thing to do here is to do what the other recommendations do (for dûk and jùnya) and to permit a diacritical mark if desired for the exceptional words like ‹shâp›. I say again, THIS RECOMMONDATION GOES BEYOND THE SCOPE OF THE REVIEW AND IS AN ATTEMPT TO REVERT TO KK PHONOLOGY.

"Accept both permissible plural endings, e.g. <hattys/hattow>" This is stupid. This is a matter of morphology, not orthography.

"Remove RLC variants <toal>, <kloav>, <gwoav>, <gwloan>, spelling them only with -a-." On the face of it this is a violation of the principle of inclusivity because RLC users will not pronounce ‹tal, clav, gwav, gwlan› correctly. However, so long as people mark these words (as KS does) by writing ‹tâl, clâv, gwâv, gwlân› there is no problem. So from that standpoint this is an improvement.

"Add <boas>, <doas>, <moas> in RLC variants to the remaining <broas>". This is a mistake. The alternation in ‹brâs› is [bɹæːz]~[bɹɒːz]. The other alternations are [boːz]~[bɒːz] which occurs ONLY in these three words. RLC users can learn their preferred pronunciation for these three words and be done with it. There is no reason to mark ‹bos, dos, mos› specially.

"1. <goolan> to become <golan>. Otherwise, no change." I don't know what this is about. We write ‹gùlla› pl ‹gùllys› 'seagull'.

"1. <ck> to be used in all borrowings; <kk> in all other cases, including those in doubt." Yeah, right. This just burdens the learner with working out which are the borrowings. Is it "gocky" or "gokky? Is it "tycky" or "tykky"? In my view "whecka" is a better spelling than "hwekka".

"2. <okkupya> to be respelt <okupya>." What a blessing. We write ‹ocûpya› [ɔˈkuːpjə] and are pretty happy with that.

As for the apostrophes, I can again say that the only rules for apostrophe use that are coherent and pleasing to the eye are those established and implemented in KS.

"Add SWF/L <enkladhva>, SWF/M <ynkladhva> (presently SWF <ynkleudhva>)." We're writing ‹ancladhva›.

I don't know what the rest of the ‹eu› words are about because of course the discussion documents were never made public.

"Write <gwr-> in RMC and RLC, inclusing <gwra>, <gwruthyl>." Good. An improvement. We already did that.

"Retain <tth>, introduce corresponding <ggh>, <cch>, <ssh>." Pretty much useless, as [θː] [xː]~[hː], [tʃː], and [ʃː] are not part of the revived language. Unnecessary Kemmynism. Unwelcome.

"<wortiwedh/wostiwedh> and <wostalleth/wortalleth> to be single words, without apostrophes." Will wonders never cease. We write ‹wortyweth/wostyweth/wàr an dyweth›.

There appear to be no concessions to KS. One can see quite clearly that in a number of instances they have re-introduced Kemmynisms precisely to distance the SWF from KS where there was previously similarity. The spitefulness is palpable.

To answer one question before anyone asks it: Since not one of the recommended changes here would improve the accuracy or authenticity of KS were it to be adopted, the orthography of KS as presented in Desky Kernowek, An Beybel Sans, and the rest of the more recent KS publications will not be changed. "Alys in Pow an Anethow" will be taken out of print and revised for a second edition because the orthography was made more precise during the preparation of An Beybel Sans, but that is not a reaction to this review, but rather just some house-keeping (since Alys was the first KS publication). (I don't know whether any of the other early publications will need revision; Alys certainly needs some.)

KS is an accurate, authentic, inclusive, and stable orthography. I commend it to everyone. To anyone who is using the SWF: Stick with SWF1. SWF2 as proposed is markedly worse and less coherent.

Michael Everson * http://www.evertype.com/





More information about the Spellyans mailing list