Gereon Müller [+]
Deponency is a widespread phenomenon in the world's languages (Baerman et al. (2007)). It raises intricate questions for all theories of inflectional morphology; essentially, I think one can make a point that the simplest approach to the phenomenon would be one that envisages the possibility that a wrong form does in fact emerge as grammatical in these cases. Accordingly, in the overview article in Müller (2013) I suggest that one should try to pursue an optimality-theoretic approach to deponency that identifies the core property of deponency with an optimal violation of a faithfulness constraint in a candidate, triggered by a lexically marked incompatibility of the stem with the faithful exponent. This works technically, but it raises the question of how well the treatment of lexically marked exceptions actually works in OT. In contrast, an approach in terms of harmonic serialism offers a somewhat new perspective on deponency (while maintaining the basicidea that an unfaithful candidate eventually emerges as optimal here): On this view, deponent stems are inherently equipped with features that may be incompatible with the features properly encoding the syntactic context, and this can give rise to a situation where the expected exponent is used up early, is subsequently replaced by another exponent (faithfully realizing the inherent feature of the deponent stem that does not correspond to the syntactic context), and consequently can never come back again since it is not present anymore in the morphological array.