Re: THEORY: Reduction of final consonants

>As some of you may know Tolkien's Quenya doesn't allow any
>non-alveolar consonants in word final position. Underlying consonants
>with another point of articulation are changed into an alveolar with
>the same manner of articulation as the underlying consonant; at least
>this is the theory, while AFAIK only m > n and k > t are actually
>attested in Tolkien's writings.
>
>The constraint itself is not surprising, being attested in Finnish and
>ancient Greek, which in addition allows only r n s and disallows even
>t l finally. The problem lies in the way the constraint is enforced:
>in these natlangs disallowed final consonants are simply deleted,
>although in Finnish there was -k > ? > zero or assimilation to the
>initial consonant of the following word. So the questions are: what
>other natlangs have a similar constraint? Are there natlang instances
>or ANADEWS for an assimilation rule like in Quenya? And are there any
>examples of what migu happen to underlying p in Finnish or other
>langs, other than deletion? The reason I ask is of course that my
>conlang Kijeb has such a constraint.
>
>/BP
Finnish does have underlying /m/ becoming /n/, too:
avain ~ avaimen "key/.GEN"
and arguably /h/ > /s/ in some cases:
mies ~ miehen "man/.GEN"

It's not certain whether a final *p existed, but -pi at any rate became -u,
later only vowel length, while other -Ci suffixes became -C. Dialects retain
/vi/, /:vi/ and /:pi/ as allomorphs.

John Vertical