Translated abstract: |
Abstract | Language |
---|
Abstract
This is the first analysis of person reference in a narrative corpus of the under-documented Mozambican Bantu language Ngoni. After a short phonological and grammatical sketch, basic concepts like person reference, text, and text types are defined. Texts are primarily instances of communicative acts because language is rooted in communication. As building blocks of a model of person reference in narratives, four cognitive dimensions are discussed:
1) Information structure, which contains the activation status of referents and the conceptualization of linguistic devices in terms of position on a scale of decreasing weight, from modified noun phrases to personal affixes.
2) Episode structure, which encompasses the hierarchical anatomy and narrative units of texts. The former distinguishes macro- and micro levels, the latter provides the narrative script.
3) Thematic structure, which consists of thematic roles and prominence. These two components exist as inherent characterizations of referents by virtue of cultural expectations, but can be adjusted to the individual narrative, for example by assigning a central role to a supposedly less prominent referent. Another component is the agent-patient frame, predicting the encoding of referents in a chain of clauses.
4) Narrative profile structure, which refers to event line and back-grounding information. While the event line establishes the core of the narrative, the backgrounding frame predicts distribution of non-event line information. In addition, the dialogue frame predicts that the addressee immediately following direct speech is more likely to be the participant referred to than the speaker.
The effort hypothesis describes default referential mechanisms: The linguistic-cognitive effort required for activation is in inverse correlation with the predictability of the referent. This entails three specific hypotheses: a) introduction of a referent is nominal, b) maintenance is pronominal, and c) means for renewed reference are located between the two extremes of the effort scale. An analysis of a single narrative illustrates the methodology: Segmentation of a transcribed text into utterances is followed by division into narrative units and identification of background information types. All referential devices are recorded in a text chart which in turn serves to assign thematic roles. Deviation from default reference is then explored, employing the four cognitive dimensions of narratives previously developed. The same principled analysis is applied to seven further texts from the corpus, and the results are discussed according to the four heuristic dimensions. The analysis shows that the choice of referential devices correlates with the criteria elaborated in the cognitive model of person reference in narratives. | English |
|
Uncontrolled Keywords: |
Keywords | Language |
---|
Personenreferenz, Implikation, Informationsstruktur, Partizpanten im Dialog, Agens-Patiens-Schema, schwere Referenzmittel, Zeroreferenz, Quaestio, Zugänglichkeit, Kognition, Kommunikation, narrativ, Erzählung, oral, Oralität, Ngoni, Bantu, Textlinguistik, Textgrammatik, Pragmatik | German | person reference, particpant tracking, implication, particpants in dialogue, information structure, Agens-Patiens-schem, heavy encoding devices,
communication, narrative, narrative text, oral, orality, Ngoni, Bantu, textlinguistics, textgrammar, pragmatics | English | referência, implicação, estrutura da informação, esquema da agentividade, meios de referência pesados e leves
comunicação, narrativa, conto, oral, oralidade, Ngoni, Bantu, linguística textual, gramática de texto, análise pragmática | Portuguese |
|