LINGUIST List 5.1151

Thu 20 Oct 1994

Disc: French clitics

Editor for this issue: <>

Directory

1. John E. Koontz, French Clitics

In regard to Pierre Larrivee's 14 Oct 94 posting on French clitics:

I think Larrivee is wrong to suggest that 'It seems that treating
French clitics as "lexically-attached" elements, ... derives from a
universalist perspective.' I don't have the impression that Miller &
Auger had this intent, or have argued on that basis. This
perspective has come from Guy, Koo[n]tz and Spackman, to judge from
Larrivee's summary.

Speaking for myself [Koontz], it seems to me that the behavior of
French inflection as depicted in this thread and in other material I
have seen in the past is not inconsistent with that of typical
Eastern North American head-marking languages. But all I am trying
to do with contributions like this one is to to make that point. I
will leave the implications for the current status French to the
principals in the debate.

Like Larrivee I lack the competence to discuss the data from these
exotic, i.e., non-European, languages, in the sense that I am not a
native speaker of any of them, but this has never stopped a linguist
before, and I don't see it as a major obstacle here either. If there
were some existing question about the nature of these exotic
languages I might be a bit more nervous, but, as it happens, everyone
who has ever been exposed to them agrees that they are head-marking,
or specifically, that the pronominal elements are inflectional.
There are also usually some separate emphatic/contrastive
pronominals, but these are another matter.

In fact, the only real potential for questioning the head-marking
analyses of these exotic languages arises from comparing them with
the standing analysis of something like French, which is either
exactly like them, but supports an alternative dependent-marking
analysis, or else subtilely different, in which case we might well
ask ourselves why.

To put it another way, a facile Americanist response to the standing
analysis of French might be that the standing analysis rests on an
obsolete orthographic and analytic tradition. If the words are
written and analyzed as they are spoken, i.e., as hypothetically
transcribed and analyzed by a linguist without preconceptions,
encountering the language as it might be spoken by the
orthographically unimpaired natives of Hither North America c.
1500-2000 AD, then French would be head-marking and no one would be
the wiser. But a more serious response might be to ask ourselves
what basis there might be for the dependent-marking analysis of French
other than mere orthographic impairment, on the theory that this
might lead us to some insight into the real languages of Hither
North America.

In that vein I will tackle Larrivee's problem of (multiple)
dislocations, or, as I would call them encountering them in a language
of HNA, NPs in apposition to or alternation with inflectional
elements in the French verb (i.e., the French verb-clitic complex). I
will consider merely whether such things occur in the languages of
HNA and other such benighted places. The answer is that they do.

The chief difficulty in strictly exemplifying (multiple) dislocations
in the Siouan languages, the ones with which I am most familiar, is
that Siouan languages as a rule do not have explicit (non-zero)
inflectional elements for third persons, barring certain third person
plural objects. Thus, while the pronominal elements in Omaha-Ponca
are definitely prefixal elements, binding intimately and irregularly
with the verb stem and constituting with the stem a single accentual
domain and breath unit, they are only explicit for first, second, and
inclusive persons (in both case relationships) and some animate third
person plural objects. The first three possibilities in particular
do not lend themselves to multiple dislocations, but it is certainly
possible to provide an independent pronominal of emphatic or
contrastive significance, as in:

 1. wi uma~ha~=<b>-dhi~
 I Omaha <I>-am

 I myself am an Omaha;
 or, I'm an Omaha, though, of course, you aren't.

Furthermore, it is possible to have multiple dislocations of third
persons with zero concord:

 2. kki dhe= akha nias^i~ga=akha ga-<0>=bi=ama:
 and this the person the <he>-said-as-follows=QUOTE

 And, this one, the person, said as follows:


 3. dhe=akha Sigdhe Ma~kka~= akha <0>-wakhega <0>-0-gagha <0>-z^a~=i
 this the Tracks Medicine the <he>-sick <he>-made-it <he>-lay

 This one, Medicine Tracks, he lay pretending to be sick.

Examples with multiple dislocation for a third person plural object,
i.e., one that has explicit inflection, are harder to come by, but
here is one.

 4. edi= kki, tta~wa~gdha~=ma dhe= ma
 there when clans the these the

 u- <a><wa>- gi- <b>- dha= tte.
 LOC <I><them> DAT <I> tell FUTURE

 In that case, the clans, these ones, I will tell it to them.

 Note that u-gi-dha LOC-DAT-tell is simply the lexical
 structure of the stem 'to tell someone'.

It would be more profitable to identify examples of this pattern in
languages (Algonquian or Iroquoian?) in which third person concord is
almost always explicit.

I am afraid I do not understand the problem with the series of examples:

 1'. Le docteur, il lui a parli
 2'. Le docteur, il a parli avec lui
 3'. Le docteur, il a parli avec ce sale imbicile toute la journie

Larrivee supposes that what the head-marking analysis would call the
verb in (1'), i.e., "il lui a parli" has no case marking, but this is
definitely not consistent with any head-marking analysis I have ever
encountered. "Il" is a subject and "lui" an indirect object (or
their inflectional analogs in a head-marking language would be),
whether or not they are also inflections. As I tried to explain in
my first posting, when a verb has two personal inflections marking
different case roles, case is at least implicit, and languages of
this sort generally indicate case in the inflection, through a
difference in form of the inflectional morphemes, through a
difference in position, or in some other way.

So the Omaha-Ponca form:

 5. u- wa- dha-kkie
 LOC them you speak-with
 you spoke with them

is transitive and has a third person plural object pronominal wa,
and a second person subject pronominal dha. Wa can only be an
object; dha can only be a subject. Dha in particular is opposed
to dhi, the second person object pronominal, which can only be
an object. Any of these pronominals could govern one or more
dislocated NPs, as in:

 6. Uma~ha~=dhatha~=s^e Ppadhi~=dha~kha uwadhakkie
 Omaha you-the Pawnees the you spoke with them
 You the Omaha spoke with the Pawnees.

(I constructed this example myself, but I'm fairly confident of it.
My main fear is that I may have guessed at the wrong positional articles
for the nuances of the social situation: the Omaha person standing,
the Pawnees sitting, as indicated by the choice of article.)

The main difference between Omaha-Ponca and other Siouan languages,
on the one hand, and French, on the other, is that the former do not
have analogs of "il a parli avec lui," in which dislocated "avec lui"
alternates with inflectional "lui" of "il lui a parli," presumably
with some subtile difference in meaning. To find languages that do
that one has to look beyond HNA, e.g., to Africa or northern Asia,
and I will leave that for someone else.

John Koontz
Research Assoc., CeSNaLPS, U of Colo.