Re: Lenition

On Monday, June 24, 2002, at 05:01 , Christian Thalmann wrote:

> --- In conlang@y..., Christophe Grandsire <christophe.grandsire@F...>
> wrote:
>
>> Well, I don't want to begin the flamewar that happened a few month ago
>> about
>> whether voiced sounds are "softer" or "harder" than voiceless sounds, but
>> suffice to say that most people seem to find voiced sounds to be softer
>> than
>> unvoiced ones (I am of them)
>
> Same here.
..and here

>
>> I still really don't understand by what logic English speakers call
>> voiced
>> sounds "harder" than unvoiced sounds. I was extremely surprised the
>> first time
>> I heard such a claim, and I still cannot imagine what it can mean.
>
> I was quite flabbergasted myself, but I do have a theory by now.  =P
..and so also was I flabbergasted.  I can assure Christophe that not all
anglophones call voiced sounds "harder".

> I assume those people somehow confound hardness with loudness.  The
> human speech apparatus can produce voiced sounds with much greater
> volume than unvoiced ones.
Maybe.   But as American English seems to voice all intervocalic plosives,
one would assume that such voicing was because the sounds are easier
or "softer".

In fact the majority of Welsh "soft mutation" changes are pretty much the
same
as those that affected Vulgar Latin a couple of millennia back.   The only
additions, so to speak, are:
/m/ --> /v/
/K/ --> /l/
/r_h/ --> /r/   (both sounds trilled)

The change /m/ --> /v/ was originally accompanied by nasalization of the
preceding vowel.   Breton still preserves this as, I believe, do some Irish
dialects.  The nasalization has been lost in Welsh.   But as [m] can be
regarded as nasalized [b], then the shift of /m/ --> nasalized /v/ is
parallel
to the shift /b/ --> /v/ (well attested in proto-Romance, Greek, and many
other natlangs).

The changes /K/ --> /l/ and /r_h/ --> /r/ is on a par with the voiceless -
->
voiced feature found with the voiceless plosives; but it's further
compounded
by the fact that the 'aspirate mutation' caused original /sl/ and /sr/ to
become
/K/ and /r_h/ respectively.   The original aspirate mutation coupled with
a
tendency to devoice initial sounds has brought about the modern Welsh
system.

{ll} btw is *not* by a devoiced (or aspirated) approximate (like the /l/ in
English "play".   It is a voiceless lateral _fricative_ denoted in X-SAMPA
by [K].   The sound also occurs in the Nguni branch of the Bantulangs in
southern Africa and in some of the indigenous langs of the New World.
The voiced equivalent also occurs in the Nguni langs and is quite a
different sound from [l].

Ray.

Replies