Re: Derived adpositions (< Linguistic term for ease of changing word-class)

On Wed, Aug 13, 2008 at 4:57 AM, Jörg Rhiemeier <joerg_rhiemeier@...>wrote:
>
>
> In Old Albic, many prepositions are case forms of nouns which
> express relations.  Some examples:
>
> _and_ 'inside'
> _andal_ (inside-LOC) 'in'
> _andan_ (inside-ALL) 'into'
> _andad_ (inside-ABL) 'out of'
>
> _ras_ 'top'
> _rasal_ (top-LOC) 'on'
> _rasan_ (top-ALL) 'onto'
> _rasad_ (top-ABL) 'from the top'
>
> _dal_ 'bottom'
> _dalal_ (bottom-LOC) 'under'
> _dalan_ (bottom-ALL) '(to) under'
> _dalad_ (bottom-ABL) 'from under'
>
> etc.

Oh! This is similar to what I'm doing with Cl. Arithide. Although I do have
a distinction between -dyniān, which is simple "to move out of/to exit"
(it's a verb, from the postp. _dus_ "outside") and -irior which is "out from
the inside of". There is some complication too, due to the existence of
different semantically-based lexical classes ("genders" is too confusing, I
think).

So parallel to yours, I have

innos "top surface", with -os for abstractions (declension I)
--> innas "place of the top surface", i.e. "top", with -as for places (II)
--> innum, with locative case

Epsolinnum "on top of the building" (epsol = building)

and

tandos "underneath-ness" (there has to be a word amidst English's millions
for this...)
--> tandas "region underneath (not just surface)", which bears an awkward
similarity to the Malay/Indonesian for "toilet"
--> tandum, with locative case

Saluntandum "on earth", literally "under the heavens" (salum = heaven(s))


>
>
> The objects of these relational nouns are coded as inalienable
> possessors, appearing in the partitive case (locative if inanimate),
> they repeat the case ending of the relational noun as an instance
> of suffixaufnahme:
>
> _andal amalal cotholol_ (inside-LOC the-LOC-LOC house-LOC-LOC)
> 'in the house'
> _andan amalan cotholon_ (inside-ALL the-LOC-ALL house-LOC-ALL)
> 'into the house'
> _andad amalad cotholod_ (inside-ABL the-LOC-ABL house-LOC-ABL)
> 'out of the house'
>
> In some daughter languages, the relational nouns form possessive
> compounds with their objects (an alternative way of expressing
> inalienable possession, in Old Albic used mainly with pronominal
> possessors (e.g. _macoth_ 'my house') and phrases such as _cothodvar_
> 'house-door'), resulting in what is essentially a Daghestanian-style
> case construction kit:
>
> _amal cothandal_ 'in the house'
> _aman cothandan_ 'into the house'
> _amad cothandad_ 'out of the house'
>
> _amal cothorasal_ 'on the house'
> _aman cothorasan_ 'onto the house'
> _amad cothorasad_ 'from on the house'
>
> _amal cothodalal_ 'under the house'
> _aman cothodalan_ 'to under the house'
> _amad cothodalad_ 'from under the house'
>
> ... brought to you by the Weeping Elf
>
Hence Cl. Ar. could be said to save a few letters here and there from not
needing to repeat case endings. ;p

dag = "house"

_dagirum_ 'in the house'
_dagirae_ 'into the house'
_dagedyniān/dagirior_ 'out of the house'

_daginnum_ 'on the house'
_daginnis_ 'onto the house'
_daginnior_ 'from on the house'

_dagetandum_ 'under the house'
_dagetandis_ 'to under the house'
_dagetandior_ 'from under the house' (which sounds similar to Hokkien
"everybody gains")

Eugene

Reply