|Title:||Cause und Change: Thematische Relationen und Ereignisstrukturen in Konzeptualisierung und Grammatikalisierung (Cause and Change: Thematic relations and event structure in conceptualization and grammaticalization)||Add Dissertation|
|Author:||Holden Haertl||Update Dissertation|
|Institution:||Universität Leipzig, English Department|
|Abstract:||The dissertation aims to define the mapping mechanisms that are at work between the conceptual and the grammatical level. This is done from a procedural vantage point thus modeling a theory of language production. Two types of information are considered crucial for the mapping operations: Thematic relations on the extra-linguistic side and their indirect encoding in event structures on the grammatical/lexical side.
Conceptually, thematic relations are established by a so-called conceptualizer that assigns and calculates two basic features of objects: CAUSE and CHANGE. These two features are described as primitive entities that are associated with universal knowledge structures applied in concept acquisition. By means of these two features an event representation can be projected/conceptualized with all the thematic dependencies encoded directly.
A so-called thematic processor systematically maps the corresponding conceptual structures onto lexico-semantic - i.e. grammatical - structures, which encode only event structural properties ('Aktionsarten') of expressions. Because thematic structure is not visible in grammar, unambiguous event structural constellations have to ensure in grammatical terms that the extra-linguistic thematic properties of event conceptualizations can be inferred correctly.
The theory is tested empirically/experimentally and theoretically against the background of a special linguistic phenomenon: the class of German psych verbs, which express causal relationships in the abstract domain of mental states. The seemingly identical thematic relations of the two groups of psych verbs (e.g. 'fascinate' vs. 'adore') are realized in syntactic positions which systematically diverge. It is argued that 'fascinate'- as well as 'adore'-verbs express (implicit) causal relations on the conceptual level, their event structures, however, differ: While 'adore'-verbs express states, 'fascinate'-verbs realize temporally homogeneous activities and are therefore to be treated as any other activity verb like 'dance' or 'hunt'. Thus (through linking rules) the cross-wise syntactic realization of the arguments comes about systematically. These hypotheses are tested empirically by means of an ERP study (event related brain potential) and several questionnaire studies.