SUBJECT AND TOPIC IN CHINESE

Rint Sybesma


In general linguistics, the notion subject has always oscillated
ambiguously between semantic subject and structural subject. The
former notion involves a selectional relation to the verb: the
semantic subject is assigned some semantic or thematic role
projected by the verb. In John is sleeping, for instance, John
is interpreted as the sleeper by virtue of semantic relation
between John and the verb. The notion of structural subject, on
the other hand, does not necessarily involve such a relation: in
the Government-Binding framework, for instance, it is associated
with a particular position, which exhibits a number of
properties. One of the properties associated to this position is
that, for languages with agreement, the NP in this position (the
structural subject) agrees (in number and person) with the finite
verb. A second property is that the NP in this position is
assigned nominative case. As a final property, it is assumed that
every sentence has a structural subject, that is, the structural
subject position must always be filled. As a general rule, the
structural subject position will be occupied by the semantic
subject. If there is no semantic subject there are a number of
alternative strategies for filling the structural subject
position of the sentence: an embedded subject is raised (subject
to subject raising as in John seems to be sleeping), the
underlying object moves into the subject position (as in
passives) or some expletive element is inserted (Its seems
that...). Viewed in this way, there seems to be strong evidence
for distinguishing logical and structural subjects.
     From the Chinese point of view, however, this is not at all
obvious. Specifically, we may wonder whether there is any
evidence that Chinese has a structural subject position, as
Chinese lacks precisely those properties which play a role in
determining the presence of a structural subject: first, Chinese
has no agreement and it lacks overt case-marking of any sort
(subjects, objects and adjuncts are not distinct
morphologically). Furthermore, it is not obvious that in Chinese
every sentence has a subject: apart from the well-known fact that
Chinese exhibits seemingly random subject-drop phenomena, it has
also been argued that Chinese has no expletives. In sum, the
``standard'' procedure to determine the presence of a structural
subject fails in Chinese.
     However, there are a number of Ä as yet, quite vague Ä
indications that Chinese has a structural subject position too
and this project aims at making these indications more explicit.
One complicating factor is the fact that Chinese is a
topic-prominent language: all kinds of different constituents may
occur in preverbal position, preceding or following the logical
subject (if present), and it has been proposed that any preverbal
constituent which is not the logical subject should be analysed
as a topic. In this project we will try to determine what
differences there are between topics and (non-semantic) subjects
which occupy the structural subject position.


R.P.E. Sybesma (1960) studied Chinese and General Linguistics at
Leiden University (graduated 1987). From 1982-1984 he stayed in
China, where he studied Chinese and Chinese Linguistics at the
University of Liaoning in Shenyang. He went back there in 1987
to teach General Linguistics for a year. He obtained his Ph.D.
in 1992 at Leiden University. The title of his thesis is:
Causatives and accomplishments: The case of Chinese ba. He has
published a number of papers on Chinese syntax and co-edited two
conference proceedings and one book on Chinese linguistics. He
is also active as a translator of Chinese literary works into
Dutch.