This paper has three aims: (1) to generalize a compulational account of the discourse process called CENTERING, (2) to apply this account to discourse processing in Japanese so that it can be used in computational systems for machine translation or language understanding, and (3) to provide some insights on the effect of syntactic factors in Japanese on discourse interpretation. We argue that while discourse interpretation is an inferential process, syntactic cues constrain this process; we demonstrate this argument with respect to the interpretation of ZEROS, unexpressed arguments of the verb, in Japanese. The syntactic cues in Japanese discourse that we investigate are the morphological markers for grammatical TOPIC, the postposition wa, as well as those for grammatical functions such as SUBJECT, ga, OBJECT, o and OBJECT2, ni. In addition, we investigate the role of speaker's EMPATHY, which is the viewpoint from which an event is described. This is syntactically indicated through the use of verbal compounding, i.e. the auxiliary use of verbs such as kureta, kita. Our results are based on a survey of native speakers of their interpretation of short discourses, consisting of minimal pairs, varied by one of the above factors. We demonstrate that these syntactic cues do indeed affect the interpretation of ZEROS, but that having previously been the TOPIC and being realized as a ZERO also contributes to the salience of a discourse entity. We propose a discourse rule of ZERO TOPIC ASSIGNMENT, and show that CENTERING provides constraints on when a ZERO can be interpreted as the ZERO TOPIC.

/pdf/japanese-discourse-and-the-process-of-centering-4ttfoahaat.pdf

Japanese discourse and the process of centering

Two Finnish language comprehension experiments are presented which suggest that the referential properties of pronouns and demonstratives cannot be reduced straightforwardly to the salience level of the antecedent. The findings, from a sentence completion study and visual world eye-tracking study, reveal an asymmetry in which features of the antecedent Finnish pronouns and demonstratives are most sensitive to, both in terms of their final interpretations and during real-time processing. In particular, the syntactic role and linear position of the antecedent, two factors which have been claimed to influence referent salience, have different effects on the interpretation of pronouns and demonstratives. Contrary to what is commonly assumed, pronouns and demonstratives cannot be mapped onto a unified salience hierarchy, because they exhibit different degrees of sensitivity to syntactic role and word order. We offer an alternative approach to anaphor resolution, the form-specific multiple-constraints approach.

Interpreting pronouns and demonstratives in Finnish: Evidence for a form-specific approach to reference resolution

Focus, pragmatic presupposition, and activated propositions

Story grammars have been proposed as a means of expressing a theory of stories. Story grammarians claim that stories are a linguistic form in much the same sense that sentences are, and that we can attribute constituent structure to stories in much the same way we attribute it to sentences. Just as sentences and their constituent structure can be characterized by sentence grammars, so stories and their constituent structure can be characterized by a story grammar. However, the analogy between story grammars and sentence grammars is ill conceived. It is based on a category error that assumes stories to be textual entities like sentences. This is demonstrably not the case. Moreover, this confusion is the cause of most of the misunderstandings about story grammars and what they can accomplish. Once this mistake is acknowledged, the possible contribution of story grammars to a theory of stories is considerably diminished. In place of story grammars, I propose the rudiments of a theory, called the theory of story points. Although it lacks some of the aesthetic appeal of story grammars, the theory of story points seems a more promising route to a meaningful theory of stories. The theory is being used as a component of a computer story-understanding system under development at Berkeley. In addition, some very preliminary experiments conducted on the basis of this approach seem to lend it some psychological plausibility.

Story grammars versus story points

The prediction of purely semantic or inferential theories of discourse understanding (cf. (Hobbs, 1985)) is that there should be no difference in coherence between (1) and (2). According to these theories, the agent in (1c), realized by the pronoun he, can only cospecify the discourse entity realized by Jeff because the verb soaping can only be related to the washing event, while the agent in 2c can only cospecify the discourse entity realized by Dick because the verb buffing can only be related to the waxing event.

Centering in Naturally-Occurring Discourse: An Overview

The elements of the stock of knowledge shared by the speaker and the hearer change their salience, in the sense of being immediately accessible in the hearer's memory. The hierarchy of salience is argued to be a basic component of a mechanism serving for the identification of reference. Some of the regularities of this mechanism are discussed, the description of which is a necessary prerequisite of an automatic understanding of connected texts.

On the role of the hierarchy of activation in the process of natural language understanding

I. It is one of the fundamental aspects of the process of communication that one of the participants in this process, the Speaker, attempts to cause the other participant(s), the hearer(s), on the basis of the rules of the given language, to modify some elements of the Information stored in their memory. Since the System of the knowledge stored in human memory is very large and structured in a very complicated way, it is useful for the Speaker to specify first those elements that are easily accessible in the hearer's memory (salient, activated), and are to be modified (or introduced into new relations with some other elements); only then does the Speaker specify which modification or which new relations are to be achieved. The thousands of years of functioning of natural languages äs means of communication between people have certainly had their impact 011 the structure of the language: the structure of the sentence is shaped to enable the hearer to determine in a quite simple manner the elements the Speaker wants to \"speak about\" and to identify them in his memory. These elements were chosen by the Speaker from those elements of the stock of knowledge which he assumes to be shared by himself and the hearer. However, the Speaker cannot simply choose any of these elements, but only those that are believed to be activated at the given point of the discourse, that are in the foreground of attention (or in very close connection with the highly activated elements). The hearer then does not need to make much effort to identify them. The stock of shared knowledge (comprising not only knowledge in the literal sense, but a wide ränge of psychological phenomena including beliefs and other attitudes shared by the Speaker and the hearer) is one of the crucial notions in the theory of topic and

Jarka Vrbová

Papers

On the role of the hierarchy of activation in the process of natural language understanding

On the salience of the elements of the stock of shared knowledge