[Spellyans] Late Cornish adaptations

Owen Cook owen.e.cook at gmail.com
Mon Aug 4 12:52:28 IST 2008

2008/8/2 Michael Everson <everson at evertype.com> rug screfa:
>>> You'd propose <Jôwan> and <Kernôwek>, then?
>> Not I: you would need to.
> Me? I don't quite follow that.

I mean that I have been arguing for <ew> as an umbrella graph, but you
want both <ew> and <ow>. A consequence of the latter decision is that
there are two kinds of <ow>, one of which means /u:/ in prevocalic
position, and the other of which means /ow/.

> We could either mark the <bêwnans>~<bôwnans> [eU]~[oU] words with
> circumflex, or we could mark the <Jôwan> words with a circumflex. This
> makes me wonder if there are words in <ew> which don't have an
> alternation... in which case <bêwnans>~<bôwnans> is the better choice
> as we would be marking an alternation. We would then have
> <êwna>~<ôwna> 'emend' and <owna> 'fear'.

Well, I tried to make a list of words using Gendall's dictionary that
don't have alternation -- there were not too many of them. Marking the
alternation seems an awkward solution, because there are more of these

Afterwards, it occurred to me that a less invasive treatment might be
to use <ew> as an umbrella graph for /ow/, but only prevocalically.
That way prevocalic <ow> can only mean /u:/ for RLC, other <ow>s can
only mean /ow/, and prevocalic <ew> can only mean /ow/. In effect,
this would basically just mean that we should restrict ourselves to
the spellings clewes and kewar at the expense of clowes and cowar -- I
can't find any other words with prevocalic /ow/ coming from /ew/.

Oll an gwella,

More information about the Spellyans mailing list