[Spellyans] Normalization of words in -ak and -ek

[Michael Everson]

On 4 Jun 2009, at 07:42, Eddie Climo wrote:
> On 3 Efn 2009, at 23:25, Michael Everson wrote:
>> On 3 Jun 2009, at 23:05, Eddie Climo wrote:
>>
>>>> On balance I suggest the following normalization:
>>>>
>>>> KS -ak, -ogyon, -ogeth, -egy, -ogyl, -oges, -ogesow,
>>>
>>> A more balanced suggestion would be
>>> KS -ek
>>> as that is used by 3 out of the 4 varieties of RC. To accommodate  
>>> those who prefer LC forms, this could be expanded to:
>>> KS -ek (MC form
>>> KS -ak (LC form)
>>
>> We aren't making an MC/LC distinction for -ak and -ek, both of  
>> which are pronounced identically, [ək].
>> We're making an orthographic distinction, where -ak is used where  
>> the plural stem changes to -og-, and -ek where the plural stem  
>> remains -eg-.
>
> Yes, that point was made quite clearly, and is what I was responding  
> to. I'm suggesting that we should do so. Traditional Cornish  
> supports both endings, and KS should reflect this,

I do not follow you.

Currently, we have many -ak words with -og- or with -eg- in derivatives.
Currently, we have many -ek words with -og- or with -eg- in derivatives.

There is currently no pattern to this at all. I propose to restrict - 
ak to -og- and -ek to -eg- in a normalization which will certainly

If I understand your proposal correctly, you propose that we allow RMC  
and RLC users to choose -ak or -ek in the simplex on the basis of  
dialect preference, that is to say, RMC people will choose -ek and RLC  
will choose -ak.

But this leaves us no better off than where we are now. If you  
preferred -ek you would never know when a word went to -og- or when a  
word went to -eg-. And you'd be adding an additional layer of  
"preference" spellings to no advantage to readers or writers. This

>> Why? Because right now there's no help for learners.
>
> This sounds awfully like the condescending 'stupid learner' argument  
> used to justify the un-Cornish forms used by KK

Tosh. There is nothing "un-Cornish" about this proposal. Perhenn*k is  
attested exactly and only twice, with no derivatives, and once in -ek  
and once on -ak. I looked (respectfully) at all of the derivative  
forms of this word innovated by revivalists. Most were in -og-. So it  
appears sensible to make this normalization. The normalization does  
not "go against" the corpus -- we have seen perhennek and perhennak  
both occur, once only, and with no other forms at all to guide any  
aspect of the reconstruction of the plural, the feminine, or other  
derivative forms for this word. I have not, as George did, made any  
decisions based on Welsh or Breton or anything else apart from  
varieties of the Revived language. It is in the Revived language where  
most of these derivatives are found. The problem is that there is no  
system to the distribution of -ak and -ek vis à vis -eg- and -og-, and  
(1) there is no advantage to learners to the lack of system and (2)  
the choices made by Revivalists in filling out the paradigms seem to  
be pretty random, not based on any particular principle.

> And the economical thing about this MC/LC alternation is that  
> there'd be no call for yet another diacritic!
>>
>> Non sequitur?
>
> No. My closing comment follows on from what was said before.

This topic has nothing to do with diacritics. I have not proposed *-äk  
or -ëk or anything of the kind. Most -og- words have an -ak; some have  
-ek. Most -eg- words have -ek, some few have -ak. I propose a simple  
normalization. This will assist learners to know which stem a word has.

Your proposal, to allow RLC users to use -ak for -eg-/-og- words and  
to allow RMC users to use -ek for -eg-/-og- words, does not help  
anyone know which words are -eg- and which words are -og-. There is  
nothing either "economical" or indeed "wise" about that -- indeed it  
compounds the problem by adding yet another dialect option where none  
is needed (because -ek and -ak are both pronounced [ək]).

> As my signature suggests, I feel we already have too many diacritics  
> in KS, The one which is particularly egregious is the diaeresis used  
> for the  bys/beis alternation.

I'm not adding any diacritics. And I'm sorry you dislike bÿs/bës.  
KS1's beis would have been a true umbrella graph for these words, an a  
spelling which has some currency in the texts. It was rejected by the  
AHG, however, and they gave us many pairs with the alternation. Since  
this introduced ambiguity with many words without the alternation, we  
needed to mark these somehow. Ŷ was rejected for technical reasons  
(and yes, in my subsequent typesetting of Cornish I have had occasion  
to use display fonts which had ÿ but did not have ŷ). At least users  
of KS will be able to tell that bÿs is pronounced beez, not bizz,  
unlike users of KK, who conflate the two.

Your signature, therefore, doesn't apply. It *is* necessary to  
distinguish bÿs/bës words from bys/res words.

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