In this article I discuss the various parts of the OT architecture in phonology: its basic structure as an input-output system, Gen, Con and Eval. These aspects are considered in the light of the question what is innate in language (i.e., part of Universal Grammar) and what could be based on acquisition (i.e., experience with language). We will see that all parts of OT can be related to general cognitive learning strategies, obviating the need for any assumption of innateness. It is hoped that future research in this area will bring OT closer to the field of general cognitive science.
Emergent Phonology seeks to minimize the role of Universal Grammar in linguistics by investigating how units such as distinctive features, segments, words, morphemes, and syllables, and other aspects of grammar, such as phonological, morphological or syntactic rules and conditions, emerge in the course of acquisition and language use, rather than as part of an innate language capacity. An obvious candidate for being acquired rather than being innate are the phonological constraints that take a central place in Optimality Theory. In this paper I discuss whether, and if so how, a constraint like *COMPLEX ‘No complex onsets’, which is assumed to be active in the acquisition of English and many other languages, could be acquired on the basis of the data to which the English language-learning child is exposed. If this constraint is acquired, it lessens the burden on any innate capacity, which is hypothesized to contain more general, cognitive strategies—perhaps not exclusive to linguistics.