Cascade-correlation (CC) is a generative, feed-forward learning algorithm for artificial neural networks (Fahlman & Lebiere, 1990). It has been used extensively in the simulation of various aspects of cognitive development (Shultz et al., 1995).