An information-processing explanation of one-trial and incremental learning