Thesis (Ph.D.) - Indiana University, Psychology, 2009The neuromodulatory neurotransmitter dopamine (DA) plays a complex, but central role in the learning and performance of stimulus-response (S-R) behaviors. Studies have implicated DA's role in reward-driven learning and also its role in setting the overall level of vigor or frequency of response. Here, a neurocomputational model is developed which models DA's influence on a set of brain regions believed to be involved in the learning and execution of S-R tasks, including frontal cortex, basal ganglia, and cingulate cortex. An `actor' component of the model is trained, using `babble' (random behavior selection) and `critic' (rewarding and punishing) components of the model, to perform acceptance/rejection responses upon presentation of color stimuli in the context of recently presented auditory tones. The model behaves like an autonomous organism learning (and relearning) through `trial-and-error'. The focus of the study, the impact of hypo- and hyper-normal DA activity on this model, is investigated by three different dopaminergic pathways--two striatal and one prefrontal cortical--being manipulated independently during the learning and performance of the color response task. Hypo-DA conditions, analogous to Parkinsonism, cause slowing and reduction of frequency of learned responses, and, at extremes, degrade the learning (either initial or reversal) of the task. Hyper-DA conditions, analogous to psychostimulant effects, cause more rapid response times, but also can lead to perseveration of incorrect learning of response on the task. The presence of these effects often depends on which DA-ergic pathway is manipulated, however, which has implications for interpretation of the pharmacological experimental data. The proposed model embodies an integrative theory of dopamine function which suggests that the base rate of DA cell activity encodes the overall `activity-oriented motivation' of the organism, with hunger and/or expectation of reward driving both response vigor and tendency to generate an explorative `babble' response. This more `tonic' feature of DA functionality coexists naturally with the more extensively-studied `phasic' reward-learning features. The model may provide better insights on the role of DA system dysfunction in the cognitive and motivational symptoms of disorders such as Parkinsonism, psychostimulant abuse, ADHD, OCD, and schizophrenia, accounting for deficits in both learning and performance of tasks.