Re: the fascination of illV

>
> My opinion is that the paradigm canis, canem; cane:s, cane:s, should
> regularly have given (in ER): cane, cane; cane, cane. This is sub-

optimal,

> and the plural forms were given the ending -i, the plural of the o-

stems,

> resulting in sg. cane, pl. cani. I don't believe in

palatalization -es >

> -e(i) > -i, nor much in a variant plural form -i:s, which I

associate

> rather with early pre-Clasical Latin than with "latino tardo".
>

I still think an west/east "shibboleth", western -s, eastern "front
glide" would explain it most economically, including the eastern 2nd
sg. -i. We already saw in Caxton what people felt towards plural -s
;-).

BTW on the sociology of unpractical old plurals: Not only in English,
but also in Dutch, and Danish do they occur in two large semantic
groups:
1) parts of the body
2) farm animals

ad 2)
Dutch eieren, kalveren, koeien (sg. koe)
Danish gæs, høns (sg. høne), ænder (sg. and), køer (sg. ko)

Torsten


Torsten