2002 Studies in the Intra-sentential Co-reference Rules of the 3rd person pronoun “ta” in Chinese. MA Thesis, Peking University, Beijing, China. [download pdf ]
Note: for update on this topic, see my article "An LFG analysis of pronominal binding in Mandarin Chinese"published in 2016 in the Proceedings of the Linguistic Society of America
Abstract: This thesis examines the intra-sentential co-reference rules of the third person pronoun “ta” in Chinese. I propose that the rules can be primarily categorized into syntactic and non-syntactic ones. In terms of syntax, this thesis combines the notion of c-command with the X-bar structure of noun phrases to study the co-reference rules of “ta” in simple non-embedding sentences first, and then the analysis is expanded to complex sentences with an embedded object clause. Based upon a thorough analysis of corpus data, I propose a new notion of “direct c-command” as the structural condition that determines the co-reference rules of “ta” in complex sentences.
More specifically, in complex sentences with one embedded object clause, the pronoun “ta” in the subject position of the embedded clause cannot refer to the entity denoted by the noun phrase in the subject position of the main clause. In comparison, noun phrases that are modifiers of the head nouns of the main clause subject may be co-indexed with the pronoun “ta” as the embedded subject, but not with “ta” in the embedded object position. Let’s use D to represent the head noun of the noun phrase in the subject position in the matrix clause, and Md to represent the noun phrase as the modifier of D. The noun phrase in the embedded subject position is directly c-commanded by D, and let’s use D(n) to represent this node. The object noun phase in the embedded clause is not directly c-commanded by D, and thus we use -D(n) to represent this. Thus the above-mentioned rules can be formulated as:
D ≠ D(n)
Md ≠ -D(n)
These rules show that the head noun phrase and their modifier in the matrix subject position have complementary co-reference possibilities. This thesis further applies this complementary rule to simple sentences as a simplified version. Such kind of rules can be called negative conditions, i.e. specifying which elements cannot co-refer, while on the other hand there are rules that specify which elements must co-refer, e.g. in the case of resumptive pronouns. Such pronouns must co-refer with certain elements in its scope, e.g. the head noun of a topicalized element.
In terms of co-reference that is not determined by syntax, this thesis discusses logophoric uses of pronouns in Chinese and their rules.
The analysis in this thesis also looks at the subcategorization of verbs in Chinese. I point out that the different types of co-reference rules are closely tied to different types of verbs. I study the 1308 verbs listed in A Dictionary of Verb Usage in Chinese and pick out those that can take an object clause and the control verbs (or “pivotal verbs” in traditional Chinese terminology), and categorize these verbs into different types. One primary distinction among these verbs is whether they have an internal specification of the directionality of the action denoted by the verb. For example, the verb pīpíng (“to criticize”) intrinsically directs towards someone other than that which is denoted by the subject noun phrase. I call such verbs “vector verbs”. In contrast, verbs such as “zhèngmíng” (“to prove”) are “non-vector verbs”.
2009 Issues in the Semantics of Mandarin Questions, Ph.D. Dissertation, Cornell University, Ithaca, NY. [download pdf ]
Note: much of the research in this dissertation has been updated and in many cases replaced by new analyses and conclusions in my book "Semantics of Chinese Questions: An Interface Approach" (Routledge 2019). Read more here.
This dissertation strives to explain certain long-standing issues in Mandarin
questions within a new framework, i.e. the Alternative Semantics theory, and also to
bring in hitherto unnoticed new data.
Part I of the dissertation examines argument wh-questions. Starting from
Tsai’s (1999) Lexical Courtesy Hypothesis, according to which wh-movement in
general should be avoided if possible, I present an analysis of Mandarin wh-in-situ
within the framework of Alternative Semantics (Rooth 1985, Shimoyama 2001) which
does not resort to LF movement or unselective binding. Furthermore I propose that the
scope marking of questions in this theory is achieved by focus intonation.
Experimental phonetic data are provided to support this important new claim.
apply this new theory to polarity, A-not-A and alternative questions in Mandarin,
showing that they are formed by syntactic specification of a set of alternatives on
different levels respectively. The Alternative Semantics analysis is further extended to
wh-existential and wh-universal constructions. I show that existential closure can be
applied either locally or non-locally as a consequence of the compositional semantics
in the wh-existential constructions. In the universal construction “mei…dou”
(“every…all”), the long-standing problem of double-distributivity is accounted for by
universal concord in the sense of Kratzer (2006)
Part II examines “how” and “why” questions using event semantics. Data
from Mandarin show that there is an event singularity presupposition in manner “how”
and causal “why” questions, and this presupposition leads to a singleton set when the
true answers are considered. This explains such cross-linguistic puzzles as the
distribution of the exhaustivity marker “all” in wh-questions and the lack of
quantificational variability effect in embedded manner and causal questions. I also
propose an analysis of verbal “how” questions in Mandarin (e.g. Yuehan zenme-le
Mali? literally “John how-ed Mary?”). The verbal “how” is treated as a ditransitive
verbal variable in the lexicon, and it can account for the three special constraints on
the use of such verbal “how” questions, i.e. the malefactivity reading, incompatibility
with negation, and lack of a ditransitive use.
I also propose a new typology of wh-questions based on the parameters of the
interpretational variability of wh-pronouns and scope marking strategies.