[Spellyans] Final -dh, final -th, final -v, final -f

Nicholas Williams njawilliams at gmail.com
Sat Aug 11 14:30:45 IST 2012

I genuinely believe that there was a systematic opposition in Cornish after the PS between 
unstressed vowel + voiceless consonant and stressed vowel + voiced consonant.
This is why I refuse to write genev, orthiv, myghternedh, menedh, etc. It is one of the reasons that prevent me from
embracing the SWF in its present form.

It must be admitted, however, that there are exceptions to the stressed -eg ~ unstressed ek rule. Dan has rightly pointed out
gorryb and marreg. To be fair I discuss gorryb in CT 8.14.

Quite probably some will be unconvinced by my suggestion that the voiceless r in gorryb
dissimilated the voiceless final to voiced. The same people will refuse to believe, no doubt,
that marreg is for marheg with a similar dissimilation.

It must be admitted, however, that the overwhelming majority of words in unstressed -ek, -ak, -yk, etc.
have a voiceless final and those with a stressed vowel tend in the later texts are written with <g>, etc.

I have already drawn attention to the anomalous forms:

dyag, nownsag, methag, bothorag, golag, yddrag/eddrag and kronag.
I have no explanation for dyag, nownsag, methag. Methek, methak 'doctor' is attested at least 6 times to methag once.
Moreover Pryce gives nawnzack. 

 Eddrag may well have been contaminated by edrega.

The other three bothorag, golag and kranag all have a sonant r, l, and n immediately before the final unstressed syllable. 
It must also be admitted that for bothorag, golag and kranag voiceless finals are also  attested, i.e. bothorak, cronek and golok/golak.
I wonder whether the sonant in any of these cases might for whatever reason have been rh, nh and lh rather than than simple r, n and l.
We have excellent evidence that voiceless rh, nh and lh all existed in Cornish.
In which case, if our sources are reliable, we may be looking at a voiceless > voiced dissimilation in all three.

It must be admitted that gorryb, marreg, etc. are edge cases, and that the fundament distribution voiceless+k ~ voiced+g holds 
in the overwhelming majority of cases. 

There is not the faintest whiff of a hint that forms like gorryb, marreg, golag are sandhi phenomena. 
Apart from these apparently anomalous forms, all -ek, -ak words retain -ek, -ak before vowels as well as consonants.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://kernowek.net/pipermail/spellyans_kernowek.net/attachments/20120811/e815ee14/attachment-0002.html>

More information about the Spellyans mailing list