Diachronic Sources of Allomorphy

Abstract In approaches to phonologically conditioned suppletive allomorphy (PCSA) based on allomorph selection via subcategorization frames, it is a coincidence from the point of view of the grammar that PCSA often appears to be phonologically optimizing. This is not true in Optimality Theoretic (OT) approaches, where phonological markedness constraints drive and explain the allomorph selection. Thus the subcategorization approach has been criticized for not capturing the role of markedness in allomorphy. A defense against this critique is that apparent optimization in some cases of PCSA is not coincidental but reflects historical changes. The forces behind the sound changes and other historical developments leading to PCSA can account for the apparent naturalness of synchronic patterns. This chapter gives proposals for the origin of PCSA in two specific cases, attempting to show that PCSA that appears to be phonologically optimizing can be explained, and should be analyzed, without reference to markedness constraints.