<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;charset=ISO-8859-1">
<title></title>
</head>
<body text="#3333ff" bgcolor="#ffffff">
<font face="Arial Unicode MS">Dear Claire, dear Irina, dear Typologists,<br>
<br>
maybe that I didn't get Irina's question right, still let me add the
following: Among the many functions of O-Split paradigms (traditionally
called Differentiated Object Marking, DOM (Bossong)), there is one type
related to what can be termed Fluid-O: A referent in O-function is
marked for case (or agreement etc.), if the speaker wants to add a
pragmatic 'comment' (be it definiteness, be it topicalization etc.).
Taking up the standard definition of Fluid-S, we can say that Fluid-O
is governed by the speaker's pragmatic 'intention' rather than by
semantic (categorial) properties of the referent in O-function (which
would give us a typical Split-O, as in Slavic) [I have summarized the
Split-Typology of S, A, and O functions in my 2000 paper in General
Linguistics (the Accusative Ergative Continuum, GL 37,71-155)].
Contrary to referents in S/A function, the O domain is prototypically
marked for pragmatic motives (S/A are prototypically liable to semantic
splits). This prototypical motivation accounts for the fact that most
(if not all) 'lexical' Split-O procedures can be derived from pragmatic
patterns. Now, as I have said above, one of the pragmatic features of
Fluid-O is that of topicality. Usually (but far from always), the
marked variant of O in a Fluid-O system refers to some kind of Given
Topic, anchored either in a preceding textual pragmatic 'head' or in
the frame/script/kwoledge system of the speaker/hearer (e.g.
typicality). In this sense, many Fluid-O paradigms would not entail a
true 'case' marker or so (horribile dictu: 'accusative'), but a
pragmatic marker (> 'Given Topic, Typicality' etc.) restricted to
referents in O-function. Hence, a Turkish phrase like <br>
<br>
(1) c^ocuk et-i ye-di<br>
child meat-ACC eat-PAST:3sg<br>
'The child ate the meat'<br>
<br>
should better be glossed:<br>
<br>
(2) c^ocuk et-i ye-di<br>
child meat-gTOP:O eat-PAST:3sg<br>
[gTOP = Given Topic and/or Typicality]<br>
<br>
In this respect it is interesting to ask whether languages with a
Fluid-O pattern also know a complementary way of marking NewTopic in
O-function. One of the many options would be to use some kind of
indefinite marker, derived e.g. from the numeral 'one'. However, I am
not sure whether there are languages which use such an 'indefinite
articel' with referents in O-function exclusively: Most often, the
domain seems to be extended to the Subjective resulting in an
'ergative-patterned' cluster {S/O} [indefinite referents in A-function
seem to be extremely rare in discourse]. In Udi [South East Caucasian],
there is an interesting distribution of Fluid-O marking that also
involves the notion of NewTopic, compare the follwing two phrases
[Vartashen dialect]:<br>
<br>
(3) ga"d-in-en eq'-ne uk-sa<br>
child-SA-ERG meat-3sg:A eat-pres<br>
'The child eats meat.'<br>
<br>
(4) ga"d-in-en eq'-n-ux u-ne-k-sa<br>
child-SA-ERG meat-SA-gTOP:O eat-3sg:A-$-PRES<br>
'The child eats the meat [you know which I mean].'<br>
[SA = stem augment, DAT2 = gTOP:O-marker < Allative, $ = second part
of discontinuous lexical stem] <br>
<br>
The alternative pattern of (4) ( ga"dinen eq'nuxne uksa) is possible,
however judged odd by many speakers [(3') 'ga"dinen eq' uneksa' would
have the reading 'the child EATs meat']. Although the floating clitic
-ne (3sg:A) can be occasionally be added to referents in S-function
(and even in A-function), these two usages are extremely rare. Hence,
we can safely claim that in Udi, the adding of a floating agreement
clitic to a referent is strongly coupled with the NewTOP-function,
whereas the case marker DAT2 signals gTOP of a referent in O-function. <br>
<br>
Naturally, the Fluid-O type (as illustrated in the examples above)
calls for further parametrization. For instance, it is crucial whether
the gTOP-marker is also used for other functions, compare the Chuvash
pair:<br>
<br>
(5) ac^a-sen-e ta^mran yapala-sem
tu-ni-n-e ka^tart<br>
child-PL-DAT ceramics thing-PL make-INF-3sg:POSS-DAT
show:IMP:2sg<br>
'Show the children how to produce ceramics!'<br>
<br>
(6) xe^vel-e te s'i-me e^nt^e v^apa^r<br>
sun-DAT TOP eat-NEG:FUT:3sg now vampire<br>
'Now, the vampire will no longer eat the sun.' <br>
<br>
Here, the DATIVE case (-e) encodes both IO (Indirect Objective) and
gTOP:O. This Fluid-O type comes close to Split-O procedures as known
for instance from Spanish. Hence, we can assume that topicalization
strategies of referents in O-function often are coupled with (or
metaphorically derived from) other functional domains. It would be
interesting to draw in more details a picture of the grammaticalization
paths leading to the pragmatic Fluid-O pattern illustrated above
[unfortunately, Heine/Kuteva (World Lexicon of Grammaticalization] do
not (as far as I can see) mention such paths]. <br>
<br>
Best wishes [and hoping that what I have said at least modestly
contributes to Irina's question]<br>
Wolfgang<br>
<br>
</font>
<blockquote type="cite" cite="mid42F18942.4020002@rice.edu">Maria
Koptjevskaja Tamm wrote:
<br>
<blockquote type="cite">Dear colleagues, I have been asked to forward
the following message on behalf of Irina Nikolaeva.
<br>
Best wishes,
<br>
Maria Koptjevskaja Tamm
<br>
<br>
-----------------------------------------
<br>
Dear Typologists,
<br>
<br>
If anyone knows of a language that has a topic marker (a dependent
marker: particle or case affix) used exclusively on direct objects,
please let me know at
<br>
<br>
<a class="moz-txt-link-abbreviated" href="mailto:irina_a_nikolaeva@yahoo.com">irina_a_nikolaeva@yahoo.com</a>
<br>
<br>
Thank you in advance.
<br>
<br>
</blockquote>
<br>
</blockquote>
-- <br>
<div class="moz-signature">
<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1">
<title>schulze adr</title>
<small><font face="Arial Unicode MS">#############################<br>
<small><span style="font-weight: bold;"><big>Prof. Dr. Wolfgang Schulze
</big></span></small><br style="font-weight: bold;">
<span style="font-style: italic;">Institut für Allgemeine
und Typologische Sprachwissenschaft
</span>(IATS)<br style="font-style: italic;">
[General Linguistics and Language Typology]
<br>
Department für Kommunikation und Sprachen / F 13.14
<br>
Ludwig-Maximilians-Universität München
<br>
Geschwister-Scholl-Platz 1
<br>
D-80539 München
<br>
Tel.: ++49-(0)89-2180 2486 (secretary)
<br>
++49-(0)89-2180 5343 (office)
<br>
Fax: ++49-(0)89-2180 5345
<br>
E-mail: <a class="moz-txt-link-abbreviated"
href="mailto:W.Schulze@lrz.uni-muenchen.de">W.Schulze@lrz.uni-muenchen.de</a>
<br>
Web: <a class="moz-txt-link-freetext"
href="http://www.ats.lmu.de/index.php">http://www.ats.lmu.de/index.php</a></font></small>
</div>
</body>
</html>