[tied] Re: Latin pinso etc.

--- In cybalist@yahoogroups.com, Piotr Gasiorowski
<piotr.gasiorowski@...> wrote:

> 15-01-04 23:10, Richard Wordingham wrote:
>
> > I think it is undule harsh to describe the derivation of

_$arpe_ /

> > _$erpe_ from <serpens> as irregular. In particular, there is no
> > appeal to analogy, dissimilation or assimilation.
>
> If fact, it's just as straightforward and phonologically regular

as,

> say, <pãrinte> from <parens>/<parentem> or <pieptene> from
> <pecten>/<pectinem>. The generalisation of the Latin nom.sg.

rather than

> the oblique stem is rare but neither impossible nor unprecedented.

Whoops! I've a typo in my web page's implementation of
http://groups.yahoo.com/group/cybalist/message/18147 Rule 11. I've
encoded it as "i/1/r!_" when it should have been "i/1/#r!_". With
this error, it's falsely predicting *<pãrînte> (displayed as
<p&r1nte>). There's also a type in the rhotacism rule - it should
be "l/r/V_XV", not "l/r/V_V". (X represents the stressed/non-
stressed marker - suprasegmentals have to be written in-line.)

Richard.