Rhymes as a window into grammar

Child language displays phenomena that are uncommon in adult languages. Consonant Harmony, i.e. long distance consonant harmony of Place of Articulation, is one such phenomenon. There are several Optimality theoretical accounts of Consonant Harmony, all involving constraints that are highly ranked at the early stages of development, such as REPEAT (Pater 1997) or AGREE (Pater & Werle 2001, 2003), but with development these constraints are demoted to a lower position such that their presence is no longer felt in later stages of acquisition. As this phenomenon is specific to child-language and virtually unknown in adult languages, the question is whether the constraints that play such a dominant role in the early stages of acquisition are child-specific and made redundant in the course of acquisition or are the constraints still available under specific circumstances in phenomenon usually characterized as ‘the emergence of the unmarked’ (TETU) (McCarthy & Prince 1994). Thus, the central question that we want to investigate in this paper is whether these lowranked constraints can ever surface after the initial stages of acquisition. We focus here on an approach to account for the phenomenon of Consonant Harmony, developed by Fikkert & Levelt (2004). They view Consonant Harmony as an epiphenomenon of the general development of Place of Articulation (PoA) features in children’s early productions. In their study, the development of PoA from five children between the ages of 1;0 and 2;11 was investigated. They coded all target words and all forms produced by the children