Connection Science is an interdisciplinary scientific and technical journal that has, since 1989, been a focus for research on connectionist modelling and neural network learning in both living and artificial systems with an emphasis on cognition and AI. Papers on these traditional themes are still strongly encouraged. However, in 2002, in response to exciting new work in evolutionary methods and adaptive robotics, its scope was broadened to include computational research on all biologically inspired adaptive mechanisms as well as all areas of biologically inspired robotics research.

Papers submitted to the journal may be of a practical nature including, but not restricted to, new adaptive methods, novel implementations of existing methods as well as empirical work that has a strong modelling or theoretical component in psychology, biology, neuroscience or biologically inspired robotics. Submissions may also be theoretical or philosophical. Review papers are welcomed and authors are encouraged to consult with the Editor-in-Chief if they are considering a submission.