Hallo conlangers!

On 22/04/2017 15:37 Jyri Lehtinen wrote:

> 2017-04-22 10:06 GMT+02:00 R A Brown <[log in to unmask]>:
> On 21/04/2017 23:20, Jyri Lehtinen wrote:
>
>> Oh yes, the Daghestanian hoax cases are a lot of fun.
>>> They are the sort of thing you can develop when you go
>>> nuts with case compounding or merging case inflected
>>> postpositions to nouns etc.
>>>
>> Right, so 'hoax cases' because they are not case suffixes in
>> the manner in which one finds such suffixes in fusional
>> languages like Sanskrit, Greek (ancient & modern), Latin,
>> Icelandic etc.?  It's always seemed to me that case suffixes
>> of agglutinating languages could be regarded as enclitics.
>>
> The "hoax cases" was actually a reference to Comrie and Polinsky's classic
> paper "The great Daghestanian case hoax" (
> https://www.academia.edu/661228/The_great_Daghestanian_case_hoax) which
> discusses their structure. It's really a nice paper to have a look at.
>
> The point is that the Daghestanian local cases have such rich and
> transparently modular paradigms that treating them all as independent cases
> may not be the most intuitive way for describing them. You certainly want
> to discuss the various distinct parts that make up these case suffixes.
> Doing so raises the question if only parts of these suffixes should then
> count as the cases and if the remaining parts would be better grouped under
> some other grammatical category.
>
> None of this, of course, means that it's necassarily a bad thing to call
> all the Daghestanian local cases simply as cases. But it makes you think
> that saying things like "Tabassaran has the most cases of any language with
> its 48/50/52" doesn't necessarily mean what you think it does.

Yep. It is not really appropriate to speak of "fourty-something cases"
in these languages the same way Latin has six or Finnish has 15. The
point is that those "case suffixes" are bimorphemic, there are not
fourty-something case markers to learn. Perhaps the better analysis is
to consider these forms locative, allative and ablative cases of derived
nouns "inside of X", "top of X", etc., where "X" is the lexical noun.
These are simply not "cases" of the same kind as the more conventional
grammatical cases these languages also have.

I once read an article by Alexander Kibrik titled "Nominal inflection
galore" (found in a volume "Noun phrase structure in the languages of
Europe", edited by Frans Plank) where he discusses Daghestanian nominal
morphology - and puts those "compound cases" (which he calls "locational
forms" or something like that) in a separate category from the more
traditional cases such as absolutive, ergative or genitive.

If you discount the "compound cases" you will find that most
Daghestanian languages have something like 5 to 10 cases -
well-developed case systems, but not out of the ordinary.

> As to the question of considering the Daghestanian cases as clitics, I
> don't remember seeing any reson for doing that. The case suffixes all look
> to be quite well bound to the nouns and hence be bound suffixes.

Case suffixes usually aren't considered clitics in agglutinating
languages in general. Usually, they are suffixes the same way as
case/number suffixes in IE languages are, just with the difference that
number is expressed by a separate suffix. A language where it may be
appropriate to speak of clitics is Japanese, which according to some has
about a dozen case suffixes, while others consider the same items
postpositions. In Basque, one may speak of clitics because the case
marker is always attached to the last word in the NP, even if this is
not the noun itself. But in a language like Finnish or Turkish, they are
best considered suffixes rather than clitics.

But defining the boundaries between "affix", "clitic" and "particle" is
not always easy.

>> Clearly the system I outlined could be extended to include
>> other spatial meanings such as 'above' (to the top of, from
>> the top of, on top of), 'below', 'behind' and so forth.
>> Maybe extended to included temporal meaning such as: until,
>> from/since, at [time]   :)
>>
>> Ray
>>
> It sure can be expanded quite a bit. Actual productive temporal cases
> aren't too common on nouns, but there's no reason why you shouldn't explore
> them as well. On adverbs they are absolutely fair game.
>
> In Kišta I have a simple set of four core local cases:
>
> lative (movement to),
> illative (movement into, originally a compound of locative and lative),
> locative (location at),
> ablative (movement from),
>
> where the difference between lative and illative is that illative refers to
> movement into a concrete location while lative is more general and is used
> for things like movement at, near, or towards. I'm, however, also
> considering that at least dialectally some local postpositions get
> agglutinated onto nouns. These postpositions are themselves inflected for
> the lative, locative, and ablative cases and follow unmarked nouns, so as a
> result they will give a handful of directional affixes that can be inserted
> between the noun stem and the case ending. That'll be my way of introducing
> these hoax cases into a conlang.

In my planned Macaronesian languages (descendants of Old Albic which
will have Daghestanian-like "compound cases"), the sources of those
"compound cases" will be nouns expressing locational relations, like
'top', 'bottom', 'inside', 'outside', etc.; in Old Albic these are
constructed with the lexical noun in the partitive/locative case (much
like English 'top of the box'), but these can be compounded (like
'box-top'). These compounds are then grammaticalized as "compound cases"
in the Macaronesian languages. What shows that these are not the same
kind of cases as the more traditional cases is that articles and
adjectives agreeing with nouns marked that way only agree with the final
case marker, e.g. "the-ALL box-top-ALL red-ALL" 'onto the red box', not
"the-top-ALL box-top-ALL red-top-ALL".

--
... brought to you by the Weeping Elf
http://www.joerg-rhiemeier.de/Conlang/index.html
"Bêsel asa Éam, a Éam atha cvanthal a cvanth atha Éamal." - SiM 1:1