| Summary: | <p>This thesis focusses on the tendencies of morphological change, as a source of evidence for how linguistic knowledge is stored and structured in the mind, and how speakers access and use this knowledge. I empirically test various hypotheses that morphological change is directed towards the optimisation of synchronic grammar. In chapter 4, I test the common claim that diachrony is guided by a universal preference for one-to-one mapping of meaning and form in language, implemented via Universal Grammar, and in chapter 5, I consider the diachronic implications of a set of claims by Carstairs-McCarthy (1998), that the degree to which synchronic systems of inflectional morphology deviate from this ideal meaning-form relationship is limited by a set of constraints on the acquisition of lexical semantics. I test the predictions made by these teleological theories against systematic data from synchronic verbal paradigms in classical Greek, and their subsequent diachronic restructuring, and show that the theories are not supported by this evidence. On the other hand, using the same data, I demonstrate two important factors influencing morphological change: the type frequency of formal patterns of implication between paradigm cells (chapter 4), and the token frequency of inflected words (chapter 6). I argue that this is best understood if we view diachronic tendencies as by-products of language use, and I make a case for a synchronic theory of morphology in which words are the fundamental stored units of language, and analogy is a psychologically real mechanism of both synchronic productivity and change.</p>
|
|---|