| dc.description.abstract | In September 1914 Prof. R.P. Botha conducted a seminar on Relational Grammar, the syntactic theory of Perlmutter and Postal, and the present thesis is an attempt at evaluating the validity of one of the basic tenets of that theory, namely the concept of "terms". The motivation of, and principles underlying, Relational Grammar and the concept of terms was outlined in chapters 1 - 3. Perlmutter and Postal claim that within Chomsky's theory of Transformational Grammar, little provision is made for an explicit, a priori characterization of the concept "possible clause structures in human languages". Transformational rules operate for a large part blindly on given constituent structures, and can therefore potentially produce ungrammatical strings. Such ungrammatical strings are then blocked by what Perlmutter and Postal describe as ad hoc measures. What is required is a theory of clauso structure which will predict what structures will in fact be grammatical, and thus prevent ungrammatical strings in a non-ad hoc way. The theory of Relational Grammar is claimed to answer this need. At the centre of this theory is the concept of terms, with reference to which clause structure laws can be formulated. Botha discussed some clause structure laws in a fair amount of detail, and the present thesis was stimulated by the thought that if it could be established that the claims made for terms were also true for Afrikaans, a foundation could possibly be laid for an application of such laws to Afrikaans. What are terms? Perlmutter and Postal have claimed that certain noun phrases (NPs) stand in essentially different relations to the verb from other NPs. These NPs, which function as the subject, object and indirect object of the sentence are called terms. The difference between terms and non-terms is said to be universal. Four syntactic processes are mentioned in which terms, but not non-terms, can be involved, namely, Reflexivization, Quantifier-float, Equi NP-deletion and Agreement, of which only the first three are applicable to Afrikaans. The bulk of chapters 4 - 6 is devoted to a description of these processes for Afrikaans as a prerequisite to testing the claims made for terms, with respect to Afrikaans. The basic conclusions reached support these claims. Some minor problems could not be satisfactorily resolved. Part of the reason for this is traceable to the present unavailability of publications on Relational Grammar by Perlmutter and Postal themselves, with a consequent need on the researcher’s part to make assumptions which may prove to be incorrect. Nevertheless, the overall conclusion reached regarding terms was positive, and in addition valuable insights were gained into the operation of certain syntactic processes in the Afrikaans language. | |
