SATO Yosuke
Assistant Professor

My current research project lies in the phenomenon of argument ellipsis – the rather liberal omission of grammatical elements such as subjects and objects when their identity is recoverable from linguistic and non-linguistic contexts – within languages in Asia, including Japanese, Korean, Mongolian, Chinese, Malay, Indonesian, Javanese, and Singapore English. This phenomenon is prevalent in Asian languages, and hence provides valuable contributions to the working of our uniquely human language faculty and to comparative syntax of natural language in general. For example, in Colloquial Singapore English, we can say things like “Eat already or not?”, where the subject (i.e., you) and the object (e.g., lunch) are omitted. This kind of omission is simply impossible in (standard) varieties of English, in which the correct sentence would be “Have you eaten lunch already?”. A similar pattern characterizes the other Asian languages as well. Thus, in Japanese, we can say “moo kaitano?” (literally meaning “wrote already?”) to ask the question “Did you write a research paper” when the context is clear enough.

Why is it that argument ellipsis is possible in Asian languages, but not in many other languages spoken around the globe? Recent research on argument ellipsis in East Asian languages has proposed that this phenomenon is made possible by the lack of grammatical agreement, by which I mean things like person/number/gender agreement of the kind familiar from Indo-European languages such as Spanish and French. This agreement is called φ-agreement in the technical syntactic literature. φ-agreement is also manifested in English, albeit in an impoverished fashion, as in John likes to play the piano. According to this theory known as the Anti-Agreement Theory, all the Asian languages mentioned above permit argument ellipsis since they do not possess φ-agreement. The other languages such as English do not permit argument ellipsis because they have such a system.

My current research on Chinese, Malay, and Javanese, however, shows that the presence vs. absence of the φ-agreement system alone cannot wholly account for minute interpretive properties between null subjects and objects – the subject-object asymmetry – and suggests instead that this asymmetry falls out from the well-known discourse constraint imposed on subject positions in the agreement-less languages. Specifically, the referent of the subject has a very strong preference to be interpreted as a definite expression. In short, we must know the identity of the referent of the subject, whether it is pronounced or not.

For the last few years, I have been investigating the new hypothesis based on my in-depth investigation of comparative argument ellipsis encompassing Asian and non-Asian languages of the world (e.g., Turkish, Persian, Hindi, Bangla) that the topic element in discourse-oriented languages, including Asian languages, have the same function as the φ-agreement in Indo-European language such as Italian, Persian and English. This uniformity hypothesis has the far-reaching implication that topic and agreement, under close scrutiny, are in fact two sides of the same coin for the purposes of natural language syntax.

My research on argument ellipsis in Asian language has thus far been published in several internationally prestigious journals, including Linguistic Inquiry, Journal of Linguistics and Studia Linguistica. I am currently developing the uniformity hypothesis regarding the topic and φ-agreement into a book-length monograph and would like to complete it during my study leave (semester I, AY 2014-2015).

If you are interested to know more about argument ellipsis or other topics I am working on, come to talk to me in AS5-05-05!