ABSTRACT
Introduction Attempts to explain language and identify a model comprehensive enough to account for language generation and understanding are rooted in ancient times. In modern approaches, scientists are interested in developing computational techniques for both speech (or character/word sequences) recognition and synthesis. An important class of methods of language processing is based on probabilistic models [1], another one uses neural networks (in particular self-organizing maps) [2]–[4], but the main challenge for any approach is dealing with the nonlinear character of language phenomenon. The main progress achieved so far pertains in principal with a concise articulation of the constraints that language obeys rather than to provide a working model of language performance. However, the recent evidences of neural sciences and the development in the domain of dynamical systems may offer promising perspectives in modeling such nonlinear processes as language processing from the chaos theory point of view. Therefore, starting from the fact that natural language phenomenon can be viewed as a dynamical system, the purpose of this work is to investigate the possibility of modeling the linguistic components by chaotic attractors.
