Abstract: This chapter discusses the relation between representation and learnability in different approaches to reduplication. We show that Correspondence Theory (McCarthy and Prince 1995) and Raimy (2000) both increase the representational possibilities for phonological theory but that only the Raimy (2000) approach has a natural simplicity metric which helps constrain the hypothesis space a language learner is confronted with. The least complex representation that maps to a surface reduplicated form in the Raimy (2000) theory can be identified based on the number of segments and precedence links. This unique least complex representation can then serve as the null hypothesis for a learner who will only consider more complex representations if confronted with positive evidence. Correspondence Theory on the other hand does not have any natural complexity metric to distinguish between phonetically identical but representationally distinct candidates which leads to an explosion in the hypothesis space for the learner. We conclude that the Raimy (2000) approach to reduplication is more representationally constrained than the Correspondence Theory approach and thus should be preferred.

Keywords: reduplication,, repetition,, Correspondence Theory,, economy of representation,, Temiar,