Record:   Prev Next
說明 455 p
附註 Source: Dissertation Abstracts International, Volume: 41-04, Section: A, page: 1568
Thesis (Ph.D.)--Georgetown University, 1979
Within groups of word meanings identifiable as semantic domains or fields on the basis of some common core of meaning, a variety of smaller clusters are discernible, word meanings related to each other on the basis of any number of more particular components or features. Because the clusters overlap and compete with each other, hierarchical structures, features displays, and other static models are inadequate to structure the multiplicity of relations present in a semantic domain
In order to discover the tools needed for tackling the theoretical problem of structuring the multidimensional relations of the semantic domain, two principal types of analysis of lexical meaning are examined. Regarding componential analysis it is found that, while the usefulness and relevance of submorphemic elements of meaning has been established, the complex relations among the components within a meaning cannot be expressed by feature lists. This conclusion leads to an investigation of lexical semantics using the tools of the predicate calculus. In the discussion of predicates, arguments, and other parameters, the formulation or word meaning in propositional form is seen to be a flexible instrument susceptible of appropriate modifications to make it adequate for the task of getting at the components of meaning and expression them concisely and perceptively
Based on this conclusion, a group of 280 lexical meanings comprising the communication domain in a Greek-English dictionary of the New Testament being prepared by E. A. Nida, J. Louw, and R. Smith are analyzed by the extended predicate calculus procedure as the first step in providing the database for a solution to the structuring of a semantic domain. As an additional preliminary step these propositionally expressed meanings are translated into lists of labeled categories of components. Both the initial formation rules for the predicate calculus procedure and the subsequent translation rules for the componential display are discussed
The nature of the semantic domain as a composite of competing interactions between complex individual word structures demands that static structuring concepts be abandoned. A dynamic model for structuring the domain is proposed, consisting of projections (algebraic formulas) that operate on a database composed of the formalized word meanings, which have been further organized in tabular form. Capable of citing components and combinations of components found at various category levels, projections exhibiting increasing degrees of complexity are elaborated and demonstrated, producing both well known clusters in the domain (such as factives, and verbs of judging) and many new ones (public address words, intended hearer reaction words). In setting out these clusters the multidimensional structure of the semantic domain is demonstrated, as competing sets of subsets of meanings are seen to cross classify the data
The development of the dynamic field structure model has provided insights for the methodology of lexical semantics, defining the entities of analysis as predicates and arguments, determining the principles of their interaction in an enlarged propositional form, and incorporating much of componential analysis as one-place predicates. The projection formula model also suggests related research in areas such as syntactic or figurative correlates of word clusters, and generates the basic data for such studies
Host Item Dissertation Abstracts International 41-04A
主題 Language, Linguistics
Alt Author Georgetown University
Record:   Prev Next