The Syntactic Processing of Particles in Japanese Spoken Language

THE SYNTACTIC PROCESSING OF PARTICLES IN JAPANESE SPOKEN LANGUAGE Melanie Siegel Department of Computational Linguistics University of the Saarland Postfach 151150 D-66041 Saarbriicken, Germany siegelOdfki.de Abstract Particles fullfill several distinct central roles in the Japanese language. They can mark arguments as well as adjuncts, can be functional or have semantic funtions. There is, however, no straightforward matching from particles to functions, as, e.g., ga can mark the subject, the object or an adjunct of a sentence. Particles can cooccur. Verbal arguments that could be identified by particles can be eliminated in the Japanese sentence. And finally, in spoken language particles are often omitted. A proper treatment of particles is thus necessary to make an analysis of Japanese sentences possible. Our treatment is based on an empirical investigation of 800 dialogues. We set up a type hierarchy of particles motivated by their subcategorizational and modificational behaviour. This type hierarchy is part of the Japanese syntax in VERBMOBIL. 1 Introduction The treatment of particles is essential for the processing of the Japanese language for two reasons. The first reason is that these are the words that occur most frequently. The second reason is that particles have various central functions in the Japanese syntax: case particles mark subcategorized verbal arguments, postpositions mark adjuncts and have semantic attributes, topic particles mark topicalized phrases and no marks an attributive nominal adjunct. Their treatment is difficult for three reasons: 1) despite their central position in Japanese syntax the omission of particles occurs quite often in spoken language. 2) One particle can fulfill more than one function. 3) Particles can cooccur, but not in an arbitrary way. In order to set up a grammar that accounts for a larger amount of spoken language, a comprehensive investigation of Japanese particles is thus necessary. Such a comprehensive investigation of Japanese particles was missing up to now. Two kinds of solutions have previously been proposed: (1) the particles are divided into case particles and postpositions. The latter build the heads of their phrases, while the former do not (cf. [6], [12]). (2) All kinds of particles build the head of their phrases and have the same lexical structure (cf. [1]). Both kinds of analyses lead to problems: if postpositions are heads, while case particles are nonheads, a sufficient treatment of those cases where two or three particles occur sequentially is not possible, as we will show. If on the other hand there is no distinction of particles, it is not possible to encode their different behaviour in subcategorization and modification. We carried out an empirical investigation of cooccurrences of particles in Japanese spoken language. As a result, we could set up restrictions for 25 particles. We show that the problem is essentially based at the lexical level. Instead of assuming different phrase structure rules we state a type hierarchy of Japanese particles. This makes a uniform treatment of phrase structure as well as a differentiation of subcategorization patterns possible. We therefore adopt the 'all-head' analysis, but extend it by a …

Content truncated to 500 words