Phonological Subcategorization, Infixation, and Relation-Specific Alignment

Brett Hyde, Jonathan Paramore


In the Phonological Subcategorization approach to infixation (Kiparsky 1986, McCarthy and Prince 1986, Inkelas 1990, Anderson 1992, Yu 2007), the position of an infix is determined by locating it at the appropriate edge of a particular phonological category, or “pivot”. Yu’s (2007) approach employs Generalized Alignment (McCarthy and Prince 1993) constraints to both specify the pivot category and locate the affix at the appropriate edge of the pivot. In this paper, we explore the consequences of replacing Generalized Alignment constraints with Relation-Specific Alignment (Hyde 2012, 2015, 2016) constraints in Yu’s Phonological Subcategorization framework. A Relation-Specific Alignment approach provides a solid foundation for the analysis of infixation. It retains the advantages that earlier Phonological Subcategorization approaches have over alternative approaches. For example, it avoids hyper-infixation, encompasses cases where infixes attach to prominent positions, and provides a general analysis of tmesis. Relation-Specific Alignment constraints also have several advantages over Generalized Alignment constraints, however. They simplify the specification of pivot categories, provide a uniform, general analysis of infixation and accent windows, and avoid Midpoint Pathology (Eisner 1997; Hyde 2012, 2015, 2016) effects. 


Infixation; Alignment; Phonological Subcategorization

Full Text:



Copyright (c) 2016 Brett Hyde, Jonathan Paramore

License URL: