INFERENTIAL CHANGE AND SYNTACTIC MODALITY IN ENGLISH

This study investigates changes in the Old English, Middle English (ME), and Early Modern English modal from the perspective of the abductive/deductive model outlined by Andersen (1973, 1974) based on the work of C. S. Peirce. Lightfoot's (1974) well-known proposal that the modals were categorially restructured in the sixteenth century is maintained in this work, but I challenge both his view of the independence of various ME changes and the proposal by Steele et al. 1981 for an earlier categorial reanalysis. With Roberts (1985) and Lightfoot, I hold that the ME modals continued to be underlying verbs, not auxiliaries, and I propose an abductive change in the subcategorization of the modals and the preterit-present paradigm in early ME which itself occurred gradually in speech communities and deductively resulted in the various gradual ME surface changes. These changes then constituted inductive pressure for a subsequent change with more rapid surface manifestations (i.e., the sixteenth century restructuring), which reflects the operation of a special type of inference called abductory induction by Peirce (1940). This often overlooked inferential type might better be called inductory abduction, since it still entails the formation of a hypothesis, which is the basis of abduction.

This study suggests that an inferential approach to language change holds certain implications for theories of grammar and change and in particular provides insight into the gradual/rapid issue. If purely abductive changes are prima facie 'possible', inductory abductive innovations are more 'probable' and may occur and actualize rather rapidly. I further argue that viewing language change within a cognitive framework obviates the need for preoccupation with elusive postulates such as Lightfoot's Transparency Principle.

In addition to the expansion of Andersen's model to elucidate developments in the modal, I examine the rise of various quasi-modal constructions in ME and the emergence of a group of 'primary' quasi-modals (e.g. need, dare) from former preterit-present verbs and weak verbs, and I find significant support in them for the specific two-stage analysis I have proposed for change in the modal.