Christian Köttl wrote:
> The possessive suffixes were in the singular
> -ja "my"
> -ki "thy",
> -ka   "thy",fem.plk
> -Su "his"
> -Sa "her"
(snip etc.)

How, if at all, do the suffixed forms relate to the free forms of the

For example, in all the Austronesian/Malayo-Polynesian languages I can think
of, there's a pretty clear relationship between the two:

I-sg. *aku -- poss.sfx. -ku almost everywhere
II-sg. *kamu -- poss. -mu almost everywhere (*kamu may have been a pl. form
originally, not sure)
(also *kaw ~*kahu -- sfx. -ko in a few; definitely the sing., originally)
(III-sg. see below) --

Suffixes for the plurals are much less common, but where they occur:
I-pl.incl. *kita -- sfx. -ta
(I-pl. excl. *kami -- no sfx. attested AFAIK; it would probably have been
?*-mi, thus falling together with--
II-pl. *miu (Moluccan *kimi) -- sfx. -mi
III-pl. *sida -- sfx. -da (often > -ra)

There was also a possessive/attributive particle *ni, which shows up in N -
N possessives and idioms like *taqi ni aNin (shit (of) wind =) cloud, and
also shows up in the usual III-sg. suffix--

III-sg. *ia, poss. *ni ia > **nya > -nya in some langs. (/-ńa/ in those few
that have a palatal nasal, like Ml/Indo), but more often -na, and this too
is near-universal.

This *ni particle is probably present in some variant forms, also
Isg **-Nku, Ipl-inc. **-nta, IIIpl **-nda (perhaps trivially in the forms
with initial nasal)

The general lack of suffixes for the plural seems to be due to the
development of politeness/status distinctions and/or occasional loss of the
incl./excl. distinction, and/or frequent generalization of sing. forms to
the plural, esp. in III person.

In Ml/Indo. the I and II-sg. suffixes are informal; the more usual
construction is collocation, using alternate pronoun forms:

aku, -ku in familiar speech, e.g. rumah/ku 'my house'
saya 'I' non-intimate: rumah saya
and similarly in II-sg, where words like bapak 'father' or titles are used
in lieu of familiar engkau or kamu, so rumah bapak '(your) house' (note
rumah bapak saya 'my father's house')

III-sg. -nya can also be plural, also respectful II-- as, darimana asal/nya
(from.where origin/poss.) 'where do you come from?'; alternate III-pl rumah
mereka (+ 'they')

Here's the set of suffixes in Buginese:
-mu (sing. only AFAIK)
-na both sg. and pl. (pl. also N+na maneng (all) for clarity)-- also used in
N - N, e.g. bola/na Ali 'Ali's house'

-ta (both I-incl./excl, also polite II-sg/pl (alt. + maneng in pl.)
(-keng I-excl., full pronoun i/keng ult. < *kami, but both forms are

Here's the set in Kisar, a Moluccan language:
-u Isg (<*-ku or *-Nku)
-mu II sg/pl
-n(V) III sg/pl

-d(V) (<**nta) Ipl in/ex. --(V)is 0 or a V determined by rule

But very closely related Leti has:
-mu sg. only
-nV sg/pl

-mi II pl.

Leti retains the Ipl in/ex distinction, but has lost the suffixes, cf.
Kis. ik makrom/do (we master/poss) 'our master (relig. Our Lord)'
Let. it matrum/na (generalizing the III sfx.)

Also, the old IIIpl *-da survives in Leti as a plural marker (-ra), but not
in Kisar.

I don't have my Fijian dictionary at hand, but I know it has reflexes of the
singulars, *-ku, *-mu, *-na. It would be interesting to look at the plurals,
since Fij. has dual-trial-multiple.  Will check.

A multi-language study of the relationship between full/poss.-suffix pronoun
forms has probably been done, somewhere, but I don't know of any. Could be
interesting, though I'd suspect there's usually a pretty clear relationship,
just as in AN.