LEARNING THE PHONOLOGY OF A LANGUAGE: AN OPTIMALITY THEORY APPROACH

Abstract

Tesar & Srnolensky's book is a valuable reference for traditional approaches to leamability within an Optimality Theory framework. It summarises the reflections of two of the 'founder members' of the discipline. However, other researchers have developed altemative algorithms based on the previous work presented in Tesar & Smolensky which, in our opinion, are more realisfic insofar as they can cope with variation and developmental instability (Boersrna & Hayes 2001). Another interesting question, which affects all computational approaches to leamability, is whether such theories are really grounded or not. Now we know that computers can actually work out the constraint ranking of a language starting frorn some (but not al]) initial hierarchies, provided that they are given sufficient overt information. Does that really mean that this is the way the human mind works?. We cannot be satisfied with a simple statement of the type 'if the hurnan mind performed these operations, it would acquire a language'. The only possible answer is that further research on phonological acquisition must be carried out in order to test whether RIPICD and EDCD are indeed at work in phonological acquisition by human beings. Tesar & Srnolensky's Learnability in Opfimality Theory is possibly a must for phonologists. But insofar as the developrnent of OT seems to have wider implications to the extent of having become a revolution in linguistic theory, it is also recomended for linguists in general and specially for applied linguists with some interest in phonological acquisition.