[Spellyans] Excel SWF-KK-Traditional Corpus-WIP

Harry 'Double-H' Hawkey bendyfrog at live.com
Thu Feb 11 11:05:33 GMT 2016

Andrew J. Trim wrote:

    This is an interesting list.

Very pleased to hear it!

    It would be better if it were to include the attested Traditional
    Cornish forms, with frequencies. I expect that you have that planned

I sure do.  Slowly, but methodically...a lot of the traditional corpus simply does not have SWF forms though...I suppose I could try and make them up myself, though it would probably be better to post the list here and get someone else to do it...

    It would be better still if it were to include the Traditional Graph
    forms and their pronunciations for the plurals!

Yes, I already thought of that.  We're going to end up with a lot of columns I think.  Maybe I should make two sheets, one for main and the other for traditional?



    Words like karate, kayak, Korea, Koran do not become more
    traditional by spelling them with c, instead of k, so I would leave
    them as k.

Hmm. Okay...

    Double k occurs only in bykken, lakka, tykki. I would spell these
    with ck, instead of kk, to match all the others.

The SWF (main) rule is, ck in 'loan-words' (???), kk in everything else, correct? I don't remember seeing any rule for this in the traditional forms rules.  Personally I'd prefer universal 'kk' in the main form, universal 'ck' in traditional, but there you go...

    The SWF rule seems to be that we have to use kt but I prefer ct,
    instead of kt, in words like activita, whecter, octet.

Erm... aren't we pretty much designing a new orthography at this point? :-)



    For some reason, k did not get transcribed to c in the following:







    keskal to keskarydhes

    keskloppa to keskussulyans

    kesskol to kesskrifores





As far as I understand, k before e,i,y = no change.  K'lonka must be due to the apostrophe.  If 'klapkodh' was incorrect before, it's fixed now (in my version anyway...).



    The word distruksyon has been converted to distruxyon.(etc.)

Fixing soon...



    For some reason, ks did not get transcribed to x in the following:







Yes, the IPA stress symbol (hope I got this right) is between the k and the s in those words. Fixed in next version probably...



    It has been assumed that the Traditional Graph equivalent of kw is
    qw. However, the Traditional Cornish in your Awesome list shows that
    the overwhelming majority of Traditional words had qu, not kw and
    not qw. Any attempt to make Cornish spelling more traditional should
    reconsider qu.

Oh, didn't know qu was an option.  Maybe I should change them all?

    The word frikhwyth has been converted to friqwhyth.

    I think that this is wrong as it is not really an instance of kw but
    rather an instance of kwh (spelt here khw.)  I think that it should
    be spelt frikwheth.

Stupid worm-brained algorithm/programmer!!!



    The word dermatologydhes appears to have the dh transposed to hd in
    the plural (dermatologyhdesow). I believe that the on-line
    dictionary has this wrong.

You would not believe the amount of stuff it has wrong.

    The word bughwas has been converted to bugwhas. However, this is
    from bugh gwas, so wh is unlikely. If the Late Cornish pronunciation
    is correct, you may need an additional Late variant buhwhas.


Yes, I think the late Cornish pronunciation caused hw->wh in this case.  


    For some reason, hw did not get transcribed to wh in the following:



Because the recommended pronunciation is 'hw' and not 'ʍ' in these words, so my program didn't catch them. Thanks.


    Unstressed word-final -i to -y conversion is not happening whenever
    the word is marked L, M, [1], [2].

More algorithm improvements needed then...at the moment it only checks if the final letter is 'i'...and the pronunciation is not 'i:'


    The word nantei has been converted to nantey. I believe that this is
    wrong, and that nantei is correct.

Okay, any idea why?  Is it because it's a diphthong?


    Well, this has turned out to be rather a long post. I hope it helps.

    Thanks for putting so much effort into checking the database! I shall use your observations to improve it for the next version..well done!

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://kernowek.net/pipermail/spellyans_kernowek.net/attachments/20160211/521e1922/attachment-0001.html>

More information about the Spellyans mailing list