Re: Possessive Suffixes

[I hope this doesn't reach anyone twice. My first attempt at
sending this mail to the list failed, so I'm trying
again...]


Hello!

First of all, I should probably introduce myself... My name
is Julia Simon, I've been lurking on the list for a while,
and now I'm finally writing something. :-)

I'm a linguist by training -- mostly computational
linguistics and comparative Indo-European linguistics, but
I'm interested in pretty much anything that looks even
vaguely linguistic. ;-) I work as a technical writer,
though (in telecommunications, if anyone's interested).

My native language is German, but I've spent so many years
in a mostly English-speaking environment that I'm well on
my way to bilingualism. And I've been living in Finland for
<mumblemumble> years, so I'm trying to do my best with
Finnish and (the local variant of) Swedish as well. But
like any good linguist, I strive to know fascinating bits
and pieces of as many other languages as possible. ;-)

Unfortunately I haven't managed to create an entire
language of my own yet... I have snippets lying around (a
phoneme system here, a morphology there, some ideas on
syntax, diachronic changes, ... well, you get the idea).
Well, maybe some day. Until then, I'll just keep having fun
playing around with phoneme systems, inflectional and
syntax patterns, and so on.

Anyway...

On 5/18/05, Rob Haden <magwich78@...> wrote:
> On Wed, 18 May 2005 19:27:24 +0100, Ray Brown <ray.brown@...> wrote:
>
> >> That's what I was wondering about.  How do languages develop possessive
> >> pronominal suffixes from independent pronouns?
> >
> >Presumably: independent pronoun --> clitic --> affix.
>
> Yes, but sometimes the picture isn't that clear.  Take Finnish, for
> example.
Funny you should mention this -- I had wanted to bring up
Finnish as well (what with it being the one language with
possessive suffixes that I actually speak, and all that).

> The modern possessive suffixes there are:
>
> 1sg -ni
> 2sg -si
> 3sg -nsA
> 1pl -mme
> 2pl -nne
> 3pl -nsA
<nitpick> The third person suffixes have a second form -Vn
that is used with oblique case forms ending in a vowel, e.g.

  hänen talonsa "his/her house" (talo+nsA, "house+3SG/PL"
                     -- ends in a vowel but is not oblique)
  hänen talojensa "of his/her houses" (talo+i+n+nsA,
                                      "house+PL+GEN+3SG/PL"
                        -- oblique but ends in a consonant)

but

  hänen talolleen "for his/her house" (talo+lle+Vn,
                                   "house+ALLATIVE+3SG/PL")
  hänen talossaan "in his/her house" (talo+ssA+Vn,
                                   "house+INESSIVE+3SG/PL")
  hänen taloistaan "out of his/her houses" (talo+i+stA+Vn,
                                 "house+PL+ELATIVE+3SG/PL")

and so on.
</nitpick>

Sometimes I wonder if maybe the -Vn variant is the original
one, perhaps with an allomorph -nV or -in or whatever that
could be attached to forms ending in a consonsant, and the
-nsA variant developed because it's too easy to confuse
possessive -Vn with the illative suffix -Vn (as in taloon
"into the/a house"). No idea where the -sA comes from,
though. -- But I digress.

> The independent Finnish pronouns are:
>
> 1sg minä, stem minu-
> 2sg sinä, stem sinu-
> 3sg hän, stem häne-
> 1pl me, stem meidä- (dia. medä- or mejä-)
> 2pl te, stem teidä- (dia. tedä- or tejä-)
> 3pl he, stem heidä- (dia. hedä- or hejä-)
>
> Things aren't quite so clear here, but some of the obscurity is from later
> sound changes in Finnish, particularly /ti/ > /si/ and /sV-/ > /hV-/.  Even
> so, however, there seems to be an element in /n/ that is present in the
> formation of all the possessive suffixes except 2sg. While it could be
> some sort of derivational affix, I think it's more likely to be the result
> of the accusative and genitive cases of nouns used with the possessive
> suffixes.  So, for example:
>
> Nom. jalka-mi 'my foot'
> Acc. jalkam-mi ' '
> Gen. jalkan-mi > jalkani 'of my foot'
>
> In most cases, the acc. and/or gen. forms, originally allophonic variants,
> came to be generalized.
Yes, present-day -ni coming from an earlier *-mi does make
sense... and then we get a correspondence -mi : minä/minu-,
which goes nicely with the obvious -si (< *-ti) : sinä/sinu-
(< *tin- or whatever) correspondence.

I suspect that the 1sg inflectional suffix -n plays a part
in this as well. I don't know enough about Finnish
historical morphology to be able to tell whether or not
this -n used to be *-m at an earlier stage, but I do like to
think so, because a triple correspondence (*-m>-n :
*-mi>-ni : min-, just like 2sg -t : *-ti>-si : *tin->sin-
and 1pl -mme : -mme : me-) would be even nicer. :-)
(Besides, I do know enough about Finnish historical
phonology to know that an old word-final /m/ will appear as
/n/ in modern Finnish.)

As you see, some of the person suffixes of Finnish verbs
bear a remarkable resemblance to the possessive suffixes:

  talo+ni "my house" : lue+n "I read"
  talo+si "your(sg) house" : lue+t "you(sg) read"
                                    (possessive -si < *-ti)
  talo+mme "our house" : lue+mme "we read"

The impression is of course spoiled a little by those
suffixes that don't quite match, namely

  (hänen) talo+nsa "his/her/its house" : luke+e "he/she/it
                                                     reads"
  (heidän) talo+nsa "their house" : luke+vat "they read"
  talo+nne "your(pl) house" : lue+tte "you(pl) read"

The underlying forms of the verb suffixes are -V for 3sg,
-vAt for 3pl, and -tte for 2pl. So we end up with

    possessive    person/agent  independent pronoun (nom.)
1sg  -ni (< *-mi)  -n (< *-m?)   minä
2sg  -si (< *-ti)  -t            sinä (< *tinä or some such)
3sg  -nsA/-Vn      -V            hän
1pl  -mme          -mme          me
2pl  -nne          -tte          te
3pl  -nsA/-Vn      -vAt          he

<insert wacky theory about double-consonant correspondences
and how -nne is really -tte in disguise, or maybe vice
versa; and another, even wackier, theory about how
possessive -Vn and inflectional -V both lengthen the
preceding vowel and therefore *must* be related>

(The 3pl suffix doesn't quite "fit"; the form seems to be
related to the active present participle. In different
contexts, the form "lukevat" can be analyzed as luke+vat
(read+3PL) "they read" or as luke+va+t
(read+PRES.PCPL.ACT+PL) "reading" (or "the reading ones",
or whatever translation goes best with the context):

  pojat lukevat "the boys are reading"
  lukevat pojat "the reading boys"

So that suffix probably can't be linked to the possessive
suffix or to the independent pronoun no matter how hard I
try. *sob* ;)

But I do wonder if an earlier form of the language maybe
had a more, um, versatile person marker that could be used
with nouns (to mark possession: house+1sg "my house") and
verbs (to mark the agent: read+1sg "I read") on the one
hand, and with some sort of generic "talking about an actual
person now" stem to form something that corresponds to an
independent personal pronoun (actualperson+1sg "I") on the
other hand... something along the lines of

   talo+mi : luKe+mi : mi+n-
   talo+ti : luKe+ti : ti+n-

and so on...

[rest snipped because I don't have all that much to say
about Greek]

Regards,
                       Julia 8-)

--
   Julia Simon (Schnecki) -- Sprachen-Freak vom Dienst
_@"  schnecki AT iki DOT fi / helicula AT gmail DOT com  "@_
si hortum in bybliotheca habes, deerit nihil
                                        (M. Tullius Cicero)

Reply