A simple computer-generated agent was developed of which four behavioral as
pects could be altered. Emotional behavior was defined and simulated in ter
ms of these aspects. Participants matched a list of six emotions to the ani
mated emotional behavior of the agent. Results suggest that it is possible
to identify emotions from a set of simulated modulation patterns. The succe
ssful simulation of a computer-generated "emotional" agent supports a behav
ioral theory of emotion. In addition, it reveals an almost overlooked aspec
t of emotion that may spark further research and shows the fertility of the
approach known as "artificial life" for the field of psychology.