Towards a typology of change in person marking reference

Abstract

Previous studies of person marking referent shifts posited directionality: pl > sg and 3rd > 2nd > 1st person. The current study expands the documentation of person shifts, finding possible counter-examples to any directionality if all shifts represent the same phenomenon. However, comparison of shift trajectories allows them to be separated into two categories: paradigm-dependent and paradigm-independent shifts, analogous to the stages of known cycles like Jespersen’s Cycle for negation. Dependent shifts are context restrictions that occur after the introduction of more specific contrasting person markers. Independent shifts are context extensions. I analyze both shift types using a model of competition between referent-specific and referent-general markers causing pragmatic restriction of the more general form. Dependent shifts occur when the innovation of a referent-specific marker pragmatically restricts a general one to a subset of its original uses with no real change in reference. Independent shifts begin when some speakers use a referent-specific marker non-prototypically in more general contexts, triggering reanalysis of the marker’s reference by reducing specification to match the broader context. Crucially, using a general marker in more specific contexts would not trigger reanalysis, naturally resulting in directionality similar to ‘bleaching’.