Of Possession

[alice]

Are there any natural languages for which the following are true:

1. It marks possession on the noun, e.g. with suffixes (e.g. Finnish)
2. It recognises different sorts of "possession" (e.g. alienable vs inalienable)

and if so, how does number 2 differentiate the different types?

[Mecislau]

The problem with your question is that a language can make the distinction between alienable and inalienable, yet not exactly put the two on equal footing. It's not like every language would have to have two sets of possessive affixes.

For instance, Proto-Yeniseian is sometimes believed to have had the distinction, as well as possibly some of its descendents (although the corpuses are too small to say anything definitive). In these cases, the difference was simply that inalienably possessed nouns always had to a possessor. So, for instance, PY *kiʔs "foot, leg" could never appear in isolation, but had to appear as *b-kiʔs "my leg", *k-kiʔs "your leg", *da-kiʔs "his leg", etc.

[linguoboy]

Are there any natural languages for which the following are true:

Hundreds, I imagine. In Osage, certain nouns--notably kinship terms--simply do not occur without possessive prefixes. With most others, however, possession is indicated with a possessive adjective (derived from a generic noun by means of the same set of prefixes).

[Whimemsz]

Hundreds, I imagine.

Probably. I mean, it's the case with all Algonquian languages--possession is marked on the head noun with person prefixes and suffixes and a possessive suffix, and inalienably-possessed nouns (mostly kin terms and body parts) are distinguished by the stem never being able to appear without possessive affixes.

[alice]

I mean, it's the case with all Algonquian languages--possession is marked on the head noun with person prefixes and suffixes and a possessive suffix, and inalienably-possessed nouns (mostly kin terms and body parts) are distinguished by the stem never being able to appear without possessive affixes.

Could you give an example or two of this?

[Whimemsz]

umm

like, you can say nikan, gikan, okan "my/your/his bone" but not *kan "a bone"?

[Astraios]

Oglala Lakota has ma- and mi- for first person alienable and inalienable possession respectively, yet the two are pretty much interchangeable in Northern Lakota and in Dakota. For example, mináǧi my spirit, my ghost, while manáǧi my shadow. This distinction is probably not maintained by younger speakers, though.

With stative verbs, it is normal to use the third person inalienably possessed form of nouns, but active verbs have the noun marked with the corresponding person:

Ištá maȟóȟote.
Ø-ištá // ma-ȟótA
3.POS.INAL-eye // 1SG.STA-grey
My eyes are grey.

Miíšta uŋ waŋbláke.
mi-ištá // uŋ // waŋ<wa>yáŋkA
1SG.POSS.INAL-eye // INS // see<1SG.ACT>
I saw it with my own eyes.


Alienable possession, if marked with prefixes, is more often marked with mitȟá-, nitȟá-, tȟá-, uŋtȟá-; while inalienable possession (especially body parts) is marked with ma-/mi-, ni-, Ø-, uŋ-. Additionally, alienable possession usually only takes prefixes if it's "sort of alienable" - I haven't a clue how it would be called, but things can be made less alien by using prefixes:

mitȟášuŋka
1SG.POSS-dog
my horse

vs.

šúŋka mitȟáwa
dog // be.mine
my dog


Kinship terms have a less regular set of inalienable possessives, where the least marked form is often the first person singular:

iná
my mother

nihúŋ
your mother

húŋku
his mother

[TomHChappell]

Look at WALS.info features 58, 24, and 57.

(Features 59, 117, and 55 might also be good comparisons.)

You probably want to look at those 21 languages in http://wals.info/feature/combined?id1=24&id2=58 that have obligatory possessive inflection, head-marked or double-marked.
And the 26 languages in http://wals.info/feature/combined?id1=57&id2=58 that have obligatory possessive inflection and have pronominal possessive affixes (whether prefixes or suffixes or both).
I think you want to look at http://wals.info/feature/combined?id1=24&id2=57 at the 64 languages which have head-marking or double-marking and have prononimal possessive affixes (whether prefixes or suffixes or both).

Feature 58 is about the fact that some languages distinguish between nouns that must be possessed and nouns that needn't be possessed; and, also, that some languages distinguish between nouns that cannot be possessed and nouns that may be possessed. To some degree it also discusses the difference between alienable and inalienable possession.

Feature 24 is about whether the noun for the thing possessed is marked, or the one who possesses is marked, or both or neither. Head-marking of a possessive NP marks the possessed noun; double-marking marks both the possessum and the possessor.

Feature 57 is about whether, when the possessor is pronominal, a morpheme is affixed to the noun for the possessum to indicate that, and if so, where.

Feature 59 is about the fact that, among languages which can indicate possession in more than one way, some classify the possessible nouns according to how their possession can be indicated.

Feature 117 is about ways in which a sentence can predicate of a possessor that it/he/she/they possess a certain possessum.

Feature 55 is about numeral classifiers; in a way it's similar to Feature 59's possessive classification.