Nervous systems distinguish between self- and other-generated movements by monitoring discrepancies between planned and performed actions. To do so, when motor systems transmit motor commands to muscles, they simultaneously transmit motor copies, or corollary discharges, to sensory areas. There, corollary discharge signals are compared to sensory feedback arising from movements (reafference), which can result in gating of expected feedback. Curiously, in infant rats, twitches—which are self-generated movements produced exclusively and abundantly during active sleep (AS)—differ from wake-movements in that they trigger robust neural activity. Accordingly, we hypothesized that the gating actions of corollary discharge that predict wake reafference are suspended during twitching. In this dissertation, we first demonstrate that twitches, but not wake movements, robustly activate sensorimotor cortex as they do other brain areas. Next, we demonstrate that wake movements can activate the sensorimotor cortex under conditions involving presumed discrepancies between corollary discharge and reafference signals. Lastly, we reveal a neural mechanism in the brainstem that inhibits reafference, but only during wakefulness; this inhibitory mechanism is suppressed during active sleep. All together, our findings provide the first demonstration of a state-dependent neural comparator of planned and performed actions, one that permits the transmission of sensory feedback from self-generated twitches to the developing nervous system.

<p>Many daily motor tasks involve the precise control of both force level and motor timing. The neural mechanisms concurrently managing these movement parameters remain unclear, as the dominant focus of previous literature has been to examine each in isolation. As a result, little is understood regarding the contribution of various sensory modalities to force output and interval production in sequential motor tasks. This thesis uses a sequential force production task to investigate the roles of visual and somatosensory feedback in the timed control of force. In Chapter 2 we find that removal of visual force feedback resulted in specific force output errors, but leaves motor timing behavior relatively unaffected according to predictions of the two-level timing model by Wing and Kristofferson (1973). In Chapter 3, we show that force output errors exhibited in the absence of a visual reference may be related to the processing of reafferent somatosensation from self-generated force pulses. The results of Chapter 4 reveal evidence that force errors exhibited following visual feedback removal are consistent with a shift in the perceived magnitude of force output and that the direction of error may be determined by prior task constraints. In Chapter 5 we find evidence of effector-specificity in the processing of and compensation for reafferent somatosensation. Lastly, in Chapter 6 we find that the interplay between audition and somatosensation in the control of sound level by the vocal effectors resembles that which is observed between vision and somatosensation in the control of force by the distal effectors.</p> / Doctor of Philosophy (PhD)