Information Packaging In Chinese Narrative And Descriptive Discourse

There exists a general tendency for given information to precede new information in a number of languages, including Chinese. To capture the nature of this tendency, the aims of this study were to determine what discourse constraints ordering of arguments is subject to, how presupposition (given) and focus (new) behave in a clause and whether and if so, in what way the tendency holds for discourse as a whole in Chinese. This study was based on narrative and descriptive discourses produced by native speakers of Chinese under controlled conditions. The advantage of the controlled data is that the analyst has access to what is going on in the participants' mind and is able to decide on the cognitive status of a discourse entity or clause in a discourse model. It is argued that a complete functional account of information packaging will require reference to informativity, focus/presupposition, and cognitive status (hearer-and discourse-status). The primary advantage of the present approach over previous accounts is its emphasis on different types of givenness (Prince 1981, 1992) and on informativity.It is argued that although the given before new principle holds up nicely for Chinese, different types of'givenness/new' are responsible for different linguistic phenomena (constructions, syntactic slots, discourse slots, genre, etc.). In other words, different linguistic phenomena are subject to different types of'givenness/new'.The data indicate that the percentage of hearer-old entities becomes lower and lower across clause-initial,-medial and -final arguments. That is, the percentage of hearer-new entities becomes higher and higher across clause-initial,-medial and -final arguments. In addition, there does exist a very strong correlation between word order (ordering of arguments) and cognitive status of arguments:exactly speaking, clause-initial arguments, be they As, Os or Ss, tend to be hearer-old while hearer-new entities tend to occur in clause-final arguments, be they Os or Ss. Moreover, our findings underlie a preferred hierarchy for what type of entity is used by a cooperative speaker in clause-initial arguments, including A, S and O:hearer-old and discourse-old> hearer-old and discourse-new> hearer-new and discourse-new. In fact, hearer-old and discourse-new freely occur in clause-initial arguments, but hearer-new and discourse-new rarely does. That is to say, clause-initial arguments are subject to their hearer-status, rather than discourse-status. In intransitive clauses, there exists an almost perfect correlation between hearer-status of Ss and clause-initial and-final Ss: There is an almost perfect mapping between clause-initial Ss and hearer-old entities and there is an almost perfect mapping between clause-final Ss and hearer-new entities.It is further revealed that discourse-old O precedes discourse-old R in an AOVR clause with OVR in focus and that discourse-old R tends to precede hearer-new O in an AVRO clause with VRO in focus. In addition, when AOV/AVO in which OV/VO are in focus and which constitute grammatical variants, discourse-new Os tend to occur in AVO and discourse-old Os realized by nouns (i.e., expressions of higher informativity) are likely to occur in AVO as well as AOV, whereas discourse-old Os realized by pronouns (i.e., expressions of lower informativity) occur in AOV. It is also revealed that when AOV/AVO in which OV/VO are in focus and which constitute grammatical variants, O of AOV tends to be discourse-old but 0 of AVO has no preference for old or new entities.It is also found that although both focus and presupposition may extend over an entire clause, presupposition typically precedes focus. On the other hand, this tendency does not apply to every type of complex sentence. In a complex sentence with a temporal subordinate clause, the subordinate clause tends to be hearer-old and the main clause tends to be hearer-new, and presupposition tends to precede focus. In contrast, in a complex sentence with a yinwei-clause, the yinwei-subordinate clause tends to be hearer-new while the main clause of effect tends to be hearer-old and focus tends to precede presupposition.It is argued that the'given-before-new' principle holds for discourse as a whole, in that hearer-old information (even discourse-new) is omitted and hearer-new information is kept in transactional discourse, and in that there is a strong tendency for topical themes, including discourse-initial topical themes, to be hearer-old. However much common ground the speaker and the hearer have, the speaker is selecting the desired start of discourse for the sake of the intended hearer. However, the hearer-old effect on topical theme is genre-specific:There is an extremely strong tendency for topical theme in narrative discourse to be hearer-old, and topical theme of descriptive discourse must be hearer-old. In fact, the occurance of hearer-new topical thems will result into the breakdown of communication. It is also suggested that discourse-slots are sensitive to informativity in that discourse unit-initial participants and circumstances tend to be of high informativity, whereas unit-internal ones of low informativity.Finally, this paper presents a summary of the findings in this study and a discussion of theoretical implications and further research. The verb can not serve as the demarcation between definite and indefinite nouns in a Chinese clause. That is, grammar and pragmatic factors must go hand in hand in the investigation of information packaging in discourse.