Redundancy in morphological exponence

External participants: please contact Christian Döhler for an invitation to this event!

The best known example of redundancy in morphological exponence are multiple or extended exponence. Multiple exponence is typically defined as the marking of identical information multiple times within a single word, i.e. total redundant repetition of information. However, many languages display multiple marking of information that we might consider only partially redundant; in these cases we have morphological structures which indicate overlapping information rather than identical information. Yet for the most part, certain examples of the overlapping type have been

omitted from the discussion on redundancy in morphological systems. I argue that by taking a broad definition of redundancy as degree of functional overlap, a definition in line with range of disciplines, a typology can be drawn in which redundancy is a scalar property incorporating all these examples. Through careful examination of a range of languages, it becomes clear that there are two distinct parameters or ways in which morphological exponence may show redundancy. And by defining redundancy with regards to degrees of overlap, natural types emerge based on the logic of set relations, identity, subset and intersection; all of which are attested in natural languages. This logic informs a novel model-theoretic description of the typological space embedded in the formal calculus of set theory. By taking a formally explicit approach, we can precisely define each point in the typology and formulate a quantifiable measurement of redundancy. The definitions of the typological space provide insights into the functionality of morphological structure while precise measurements open up the opportunity for large scale quantitative work grounded in rigorous qualitative typological analysis.