In this report, we summarize two attempts to ascertain whether structure arises in Hopfield brain models subject to Hebbian learning. The first attempt uses graph measures to derive numerical scores for networks, and then apply data mining methods to separate the networks in parameter space. The second attempt uses one- and two-event chains to relate synapse strength to connection information. Learning methods are applied to brains of different types to see if changes can be detected. No Hebbian learning method caused tree-like structure to develop.