Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

John T. Green


Operant (instrumental) conditioning is a laboratory method for investigating voluntary behavior and involves training a particular response, such as pressing a lever, to earn a reinforcer. Operant behavior is generally divided into two categories: actions and habits. Actions are goal-directed and controlled by response-outcome (R-O) associations. Habits are stimulus-driven and controlled by stimulus-response associations (S-R). Behavior is determined to be goal-directed or habitual by whether or not it is sensitive (action) or insensitive (habit) to reinforcer/outcome devaluation. Many brain regions have been linked to the learning and/or expression of actions and/or habits. This dissertation investigates a few different brain regions in goal-directed and habitual behavior, and determines more specific roles for the prelimbic cortex, infralimbic cortex, prelimbic cortex to dorsomedial striatum pathway, and Crus I/II of the cerebellum.

Chapter two investigates the prelimbic and infralimbic cortices in goal-directed behavior. We trained rats on a two-response paradigm, where one response was extensively-trained, and a second response was minimally-trained in a separate context. This maintained both responses as goal-directed. In experiment 1, inactivation of the prelimbic cortex at time of test resulted in an attenuation of responding, but only for the minimally-trained response. This implicates the prelimbic cortex in the expression of goal-directed behavior, but only when that goal-directed behavior is minimally-trained. In experiment 2, we repeated the procedure with infralimbic cortex inactivation and found an attenuation of the extensively-trained response. This implicates the infralimbic cortex in the expression of extensively-trained behavior that is goal-directed.

The third chapter examines the role of the prelimbic cortex-to-dorsomedial striatal pathway in minimally-trained operant behavior. Both regions have been implicated in operant behaviors and have strong anatomical connections, but few studies have directly linked them together in the mediation of operant behaviors. After minimal instrumental conditioning, we silenced projections from the prelimbic cortex to the dorsomedial striatum and found that instrumental behavior was reduced, implicating this PL-DMS pathway in the expression of minimally-trained operant responding.

The final chapter examines the role of Crus I/II of the cerebellar cortex in the expression of goal-directed and habitual behavior. The cerebellum is well-characterized as a mediator of motor coordination via its connections with the motor cortex. There is also evidence of connections between Crus I/II and non-motor regions of the prefrontal cortex. Additionally, recent studies have pointed towards a role for Crus I/II in non-motor function. In experiment 1, rats learned one minimally-trained and one extensively-trained response, and both responses were goal-directed. Inactivation of Crus I/II attenuated responding only in rats that had undergone reinforcer devaluation. Residual responding in rats that have undergone reinforcer devaluation is attributed to habit, suggesting that Crus I/II may be involved in habit expression. In a follow-up experiment, we extensively-trained a single response and verified that it was expressed as a habit. This time, Crus I/II inactivation at time of test had no effect. Overall, this complex pattern of results suggests the possibility that Crus I/II of the cerebellar cortex is only engaged in habit expression when two responses are trained, but further experiments will be necessary to verify this.



Number of Pages

205 p.