Differences in resting activities are not the sole reason for the ultimate differences in activities. To see this, let’s focus on the activity of two other word nodes: move and mate. To add these words to the activity plot, click on the Exceptions to threshold tab of the plot settings form, and add these words to the selected factors underneath always, then click redraw.

Looking at the left-hand side of the plot, we can see that move has a clear advantage over mate at the beginning of the trial. This is what we would expect, because move is a higher frequency word than mate, and so it has a higher resting activity. However, over the course of the stimulus presentation this activity difference becomes smaller and then reverses, so that mate is more strongly activated than move. This may at first be puzzling – both word nodes are orthographic neighbours of the stimulus, so presumably they receive equivalent feedforward input from the letter layer? So how does the mate node managed to overcome its initial disadvantage?

Next: The gang effect