Topic markers on direct objects

Fri Aug 5 14:13:39 UTC 2005

I’m afraid I have to disagree with my friend Wolfgang’s characterization 
of Udi.  First, regarding Irina’s original question, Udi does not have a 
particle or affixal topic marker, let alone one used exclusively on direct 
objects.  Second, the clitics that Wolfgang refers to as “floating 
agreement clitics” do not mark new topics, but focus.  These clitics are 
the subject of an entire book (my Endoclitics and the Origins of Udi 
Morphosyntax, CUP, 2002), and chapter 3 explicitly establishes that the 
markers in question indicate focus, not new topic or any other discourse 
function.  It is true, as Wolfgang notes, that these clitics are 
infrequent with subjects of transitive or intransitive verbs; they occur 
frequently with direct objects and with NPs in various adjunct functions. 
This is entirely in keeping with their being markers of focus, not of 
topic.  Third, use of the dative case for a direct object in Udi is also 
not topic marking, but indicates instead definite direct objects, as shown 
in section 11.1 of the work referenced above.  It is true that 
definiteness is related to topicality, but a definite direct object need 
not be topic.  In sum, Udi does not have explicit marking of any kind for 
topics; in particular, “floating agreement clitics” and dative case have 
been shown to have other, very specific, discourse functions.

Alice C. Harris
Professor 
Department of Linguistics
SUNY Stony Brook
Stony Brook, NY  11794-4376
Phone: 631-632-7758, 631-632-7777
Fax: 631-632-9789
e-mail: alice.harris at stonybrook.edu

Wolfgang Schulze <W.Schulze at lrz.uni-muenchen.de> 
Sent by: Discussion List for ALT <LINGTYP at LISTSERV.LINGUISTLIST.ORG>
08/04/2005 05:18 AM
Please respond to
W.Schulze at lrz.uni-muenchen.de

To
LINGTYP at LISTSERV.LINGUISTLIST.ORG
cc

Subject
Re: Topic markers on direct objects

Dear Claire, dear Irina, dear Typologists,

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 

(1) c^ocuk   et-i                ye-di
     child       meat-ACC   eat-PAST:3sg
     'The child ate the meat'

should better be glossed:

(2)  c^ocuk   et-i                      ye-di
      child       meat-gTOP:O   eat-PAST:3sg
      [gTOP = Given Topic and/or Typicality]

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]:

(3)  ga"d-in-en         eq'-ne           uk-sa
      child-SA-ERG  meat-3sg:A  eat-pres
      'The child eats meat.'

(4) ga"d-in-en         eq'-n-ux                   u-ne-k-sa
      child-SA-ERG  meat-SA-gTOP:O  eat-3sg:A-$-PRES
     'The child eats the meat [you know which I mean].'
[SA = stem augment, DAT2 = gTOP:O-marker < Allative, $ = second part of 
discontinuous lexical stem] 

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. 

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:

(5) ac^a-sen-e     ta^mran    yapala-sem   tu-ni-n-e            ka^tart
     child-PL-DAT  ceramics  thing-PL        make-INF-3sg:POSS-DAT 
show:IMP:2sg
     'Show the children how to produce ceramics!'

(6) xe^vel-e    te       s'i-me                        e^nt^e  v^apa^r
      sun-DAT  TOP  eat-NEG:FUT:3sg   now      vampire
      'Now, the vampire will no longer eat the sun.' 

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]. 

Best wishes [and hoping that what I have said at least modestly 
contributes to Irina's question]
Wolfgang

Maria Koptjevskaja Tamm wrote: 
Dear colleagues, I have been asked to forward the following message on 
behalf of Irina Nikolaeva. 
Best wishes, 
Maria Koptjevskaja Tamm 

----------------------------------------- 
Dear Typologists, 

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 

irina_a_nikolaeva at yahoo.com 

Thank you in advance. 

-- 
#############################
Prof. Dr. Wolfgang Schulze 
Institut für Allgemeine und Typologische Sprachwissenschaft  (IATS)
[General Linguistics and Language Typology] 
Department für Kommunikation und Sprachen / F 13.14 
Ludwig-Maximilians-Universität München 
Geschwister-Scholl-Platz 1 
D-80539 München 
Tel.:     ++49-(0)89-2180 2486 (secretary) 
             ++49-(0)89-2180 5343 (office) 
Fax:     ++49-(0)89-2180 5345 
E-mail: W.Schulze at lrz.uni-muenchen.de 
Web: http://www.ats.lmu.de/index.php 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20050805/13e95a4e/attachment.htm>