Locality and harmony: Perspectives from artificial grammar learning

Harmony is one of the most researched phenomena in phonological theory, mainly due to its iterative, nonlocal nature. In this review, I summarize several of the key issues in harmony that can be addressed using an artificial grammar learning approach, including differences between consonant and vowel harmony, transparent and opaque neutral vowels, computational complexity of attested and unattested harmony patterns, and morphological and phonetic influences on harmony. The review demonstrates how the artificial grammar learning paradigm can offer unique insights into the representation of vowel harmony and provides suggestions for ways in which aspects of harmony can be manipulated to offer further insights into the still unanswered questions on harmony and phonological representations.