Dynamic Modelling
A major challenge for phonological theory is to give a systematic, principled account of the interaction between the qualitative units of phonological form (e.g., segments, tone categories, prosodic units and positions) and quantitative phonetic properties of speech articulation and acoustics. The quantitative investigation of prosody in less-studied languages has already begun to reveal an even more richly intertwined web connecting quantitative and qualitative aspects of structure than was previously known. The complexities of these interactions make it particularly crucial to develop analytical tools in which qualitative and quantitative descriptions are intrinsically integratedand in which predictions can be made regarding their joint behavior. The basis for such tools can be found in the mathematics of nonlinear dynamics, a formal language that can express both quantitative and qualitative properties of complex systems. Such systems have been employed to model several aspects of cognition (Haken, 1977; Kelso, 1995; Port & van Gelder,1995; Tuller, Case, Ding & Kelso, 1994) and have recently been applied to modeling syllable structure (Goldstein et al., 2006; Nam et al., 2009), phonological processes such as Hungarian vowel harmony and German final devoicing (Gafos & Benus, 2006), the allomorphy of the regular English past tense (Goldstein, 2011) and aspects of prosody (Saltzman et al., 2008; Nava, 2010).