On 21 March 2012 16:30, Jörg Rhiemeier <[log in to unmask]> wrote:

> > That's basically it: Suffixaufnahme is a phenomenon of *agreement*, and
> > happens only in languages that have word marking, i.e. all participants
> in
> > a noun phrase must be marked for the function of that noun phrase.
> > Suffixaufnahme just takes that rule without exception, i.e. it marks all
> > participants for the function of the phrase, even if they are already
> > inflected.
> >
> Right.  This is also the way it works in Old Albic:
> Adarama am catham ndaron Atambaradon.
> AOR-give-3SG:P-1SG:A the-OBJ cat-OBJ man-DAT Atambara-ABL-M-DAT
> 'I have the cat to a man from Atambara.'
Yes. As you may remember, I used an Old Albic example to illustrate
Suffixaufnahme in my LCC4 presentation :) .

> > >
> > That's only one aspect of surdéclinaison, converting adverbial phrases
> into
> > noun modifiers. Surdéclinaison has many other aspects, including but not
> > limited to:
> > - nominalisation of noun modifiers (to make the equivalent of "at the
> > neighbour's").
> Works the same way in Old Albic.  A genitive of an animate noun
> can be marked with a local case to create the equivalent of English
> "at/from/to someone's":
> Agama Mørdindoson.
> AOR-go-1SG:A Mørdindo-GEN-ALL
> 'I went to Mørdindo's (home).'
Interesting. So Old Albic has both Suffixaufnahme and at least one example
of surdéclinaison. Interesting, I am not aware of a single natlang example
that features both. Or did Old Georgian also allow such constructions?

> > - nominalisation of relative clauses (to render "the one that..." or
> > "whoever/whatever/...").
> In Old Albic, the relative clause is preceded by an article which
> is inflected for the case and number of the noun it modifies.
> Things like "the one that ..." are expressed that way, too:
> An larasa cenithi daroama am darath san.
> the-DAT sing-3SG:A good-SUP-INS give-FUT-3SG:P-1SG:A the-OBJ gift this
> 'Who sings best, to him I will give this gift.'
That's basically how Moten does it too, but since in Moten the article is
an inflectional infix, the result is surdéclinaison:

Kopejufeano |lezuj |lajteos luden joplude|n ige.
ko-pe-uf<e>an-no               j-lezu-j
|la-i-t<e>o-s                    l<d>en-n
j-opl<d>e-j-n                         i-ge.
INF-give<ACC.SG>-INF-ACC PRS-have.
I will give this to whoever sings greatest.

Literally: "(I) shall transfer this from me for the benefit of the one that
sings with the most greatness" (_joplej_ implies a transfer from the
speaker, so you don't need to indicate a subject).

(I'm curious whether the full interlinear is making things simpler or
actually more complicated to understand ;) )

The details are different (benefactive prefix instead of dative case,
separate article vs. infix), but this underlying structure is quite similar.

> > - formation of adverbial subclauses by inflecting conjugated verbs using
> > noun cases (I will discuss that in my next post).
> As above, by means of an inflected article in Old Albic:
> Amad alarasa Mørdindo am laras adarasa Phendrato son am darath.
> the-ABL AOR-sing-3SG:P-3SG:A Mørdindi the-OBJ song AOR-give-3SG:P-3SG:A
> Phendrato he-DAT the gift
> 'After Mørdindo sang the song, Phendrato gave him the gift.'
Interesting. This is indeed not unlike one of the ways Moten handles
adverbial phrases (by inflecting noun clauses as above), although the more
common way is slightly different (but still involves surdéclinaison).

Funny how the underlying principles are so similar...

> Is this what you mean?
Yes, but with inflectional affixes rather than separate articles :) .

> > - all kinds of restricted patterns and one-offs, which can be used to
> > create new lexical items (for instance, in Moten the word for "week" is
> > _negesizdan_, which is actually an inflected form meaning "for seven
> days").
> There may be such forms in Old Albic, too; they are not yet well
> explored.
It's a great way to derive new vocabulary in ways different from plain
derivation or compounding. And necessary in Moten since it has only very
little derivation.

> > And I'm sure you could imagine many more things to do. Those are just the
> > patterns of surdéclinaison that appear in Moten (and similarly in
> Basque).
> >
> > That's the power of surdéclinaison: it's far broader than Suffixaufnahme,
> Certainly!
> > to which it bears only a passing similarity (I'm nearly sure they are
> > actually incompatible: Suffixaufnahme requires word marking,
> surdéclinaison
> > seems to require group marking, i.e. marking NPs for function only once,
> a
> > the edge of the phrase).
> Surdéclinaison indeed seems at least to correlate with group
> inflection,

Yes, although Old Albic seems to prove you can have at least som forms of
surdéclinaison even in a word marking language.

> while suffixaufnahme is a matter of word inflection
> (AFAIK Georgian lost suffixaufnahme when it moved from word
> inflection to group inflection, but I am not sure).
Yeah, you cannot have agreement within a noun phrase in a language that
doesn't have word marking, so group marking seems to preclude
Christophe Grandsire-Koevoets.