Re: cases

Me govanen!

Jim Hopkins tetent:
>
> Hello Mario,
>
> Like many others on the List I am not a trained linguist.  As to the question
> as to how many cases a language should have, I feel that it is entirely a
> questions of taste.  Among the worlds languages many use no cases at all and
> others have many. My constructed languge "Druni" has nine cases.  I
> experimented with more but found that too ungainly and so retained only the
> ones I liked.
Very true.  It is always a question how many feel right in a particular
project.
Nur-ellen, my current chief conlang project, has two (which obviously as
few as you can have unless you have no case system at all), wherein one
(the agentive) only exists for animate nouns (and adjectives) while
inanimate nouns are always in the other (the objective).
Proto-Quendian, another conlang in my pipeline, the common ancestor of
Nur-ellen, Makaronesian and several others, will probably have nine
cases of which three will only exist for animate nouns and adjectives.
(Makaronesian also lies in the future, and I haven't settled yet on the
number of cases it will have.)  Modern Vandalic will have three cases
(nominative, genitive, accusative) not because it is my favourite number
of cases, but because it seems right to me for a Germanic language
heavily influenced by Arabic.  I have no idea yet about the number of
cases in (my version of) Pictish, though it ought to have quite a
number.

> The same holds true with verb forms and the shapre and sounds of individual
> words.  I did what I found pleasing and just kept refining the language until
> I was happy with the overall effect.
Very true.  And one should not get hung up with lesser-weight universals
one violates.  Nur-ellen is (and Proto-Quendian and Makaronesian will
be) an active language; it is stated that active languages never have
case markings on nouns, but Nur-ellen has and Proto-Quendian and
probably also Makaronesian will have.

OK, you might wonder what all those conlangs are meant to be.  OK, here
it goes:

Proto-Quendian: the proto-language of the *historical* Quendi, i.e. the
Bell Beaker people.
Nur-ellen: a Quendian language that survived in Britain into the
present.
Makaronesian: a Quendian language once spoken on the Canaries, Azores
and Madeira.

This Quendian family is related to Tolkien's Elvish languages; the
assumption here is that Tolkien based his languages on fragments of
historical Quendian languages he found.

Pictish: the pre-Celtic, non-IE, non-Quendian language of the Picts.
Ergative, agglutinative, VSO, initial consonant mutations,
triconsonantal roots.

Modern Vandalic: the language of Vandalland, the only Germanic language
spoken predominantly by Muslims and written in Arabic script.

Syld,
Jörg.