PROJECT SUMMARY Dexterous manipulation of objects relies on brain computations that integrate neural signals encoding tactile signals on the skin with the proprioceptive state of the hand. In particular, neural ensembles that encode tactile motion on the skin are critical because they provide feedback signals to motor planning areas to indicate whether an object is slipping from the hand. Although we have a good understanding of how tactile motion signals are represented in the brain, this knowledge has been accrued from studies that placed the hand in a fixed position. Indeed, all studies that have investigated tactile motion mechanisms at the single-cell level have done so in animals not performing a motion discrimination task, and with their hands placed in a fixed posture. These studies show that tactile motion can be represented in area 1 by cells that integrate different tactile cues of the object (e.g., direction, speed, and saliency). However, our recent work in humans shows that perception of tactile motion on a finger is modulated by the proprioceptive state of the hand, and the body part location in which motion judgements are made relative to (i.e., the reference frame). These findings indicate that current models of tactile motion require major revisions. That is, neural models of tactile motion should take into account how motion representations in touch are transformed by proprioception and/or reference frame signals. Thus, the overarching goal of this application is to determine the neural areas and mechanisms that generate reference frame-specific representations of tactile motion. Key to determining the single-cell mechanisms that mediate reference frame-specific representations of tactile motion is to record activity in non-human primates (NHPs) discriminating tactile motion stimuli in different reference frames, and with their hands placed in different postures. Unfortunately, our field does not have an established paradigm, or training regime, to study these motion mechanisms in a NHP. In Aim I, we develop a behavioral paradigm to train NHPs to discriminate the motion direction of stimulus on a finger (e.g., index finger) in different reference frames (e.g., relative to the center of the body, or the thumb), while their hands are placed different proprioceptive states (e.g., pronated vs. supinated). In Aim II, we will record single-unit activity in somatosensory cortex (area 1) in trained NHPs to determine the neural computations that generate reference frame-specific representations of tactile motion. Our behavioral experiments will test that perceptual representations of tactile motion are conserved across NHPs and humans, demonstrating that NHPs are a viable species to study neural mechanisms of tactile motion at the single cell level. Our neurophysiology experiments will test whether motion selective neurons in area 1, the tactile analogue of medial temporal (MT) cortex for visual motion, flexibly represent motion in diff...