Motor Control Lab | Research and Publications

Recent studies in human motor learning have documented the involvement of higher-order somatosensory regions, speciﬁ-cally the rostral parietal cortex (rPC), in learning and retention, with limited participation of primary motor cortex (M1). Theabsence of M1 involvement suggests the recruitment of alternative frontal regions to support learning. The dorsal premotorcortex (PMd) is a primary candidate given its role in movement planning and its anatomical connectivity with rPC. However,the functional contribution of PMd to retention and the nature of its interaction with the parietal cortex—whether they operateindependently or as an integrated network—remains unknown. To address this, we used a visuomotor adaptation task inwhich participants adapted to altered visual feedback. Following the acquisition of the motor memory, continuous theta burststimulation (cTBS) was applied to either M1, rPC, or PMd to assess their contribution to retention. In retention tests, 24 h later,disruption of PMd led to a signiﬁcant impairment, conﬁrming its role in visuomotor learning. Disruption of rPC similarly led toimpairment, whereas disruption of M1 did not. Crucially, when rPC and PMd were disrupted simultaneously to test for inde-pendent effects, the resulting impairment was no greater than when either area was disrupted alone. These ﬁndings thus indi-cate that these areas are functionally interconnected in a motor learning circuit, such that disruption of either node results ina similar impairment. This supports a model of adaptation and learning in humans that relies on a distributed parietal-premotornetwork that is not dependent on M1.

NEW & NOTEWORTHY This study provides direct evidence of motor learning-related plasticity in a cortical circuit involving sen-sory areas in rostral parietal cortex and premotor areas in frontal cortex. Using a visuomotor adaptation paradigm with cTBS, dis-ruption of rostral parietal or dorsal premotor cortex impairs retention, whereas disruption of primary motor cortex does not.Disrupting both regions together causes no additional impairment, suggesting an interconnected learning circuit in humans andadvancing understanding of cortical substrates of motor learning.

Changes to speech offer a quantifiable means to assess speech motor learning, and the resulting memory is thought to be motor in nature. Here, we evaluate this idea and show instead that memory for speech movements has a sensory basis. Speech motor learning, using altered auditory feedback, provides an experimental model to address this question as it involves auditory, somatosensory, and motor components to learning. Transcranial magnetic stimulation was used to disrupt auditory (superior temporal gyrus, STG), posterior somatosensory (S1), or motor (M1) cortex following speech motor learning. Retention tests were conducted 24 h later. It was found that following disruption of either STG or S1, motor memory retention was impaired whereas disruption of M1 led to retention that was no different than that of a no TMS control condition. The effects of disruption were specific to speech motor learning and did not interfere with speech production per se. Taken together, the findings support the notion that plasticity in the sensory cortex, both auditory and somatosensory, is necessary for speech motor learning and memory. In speech, changes to sensory systems enable the production of newly learned movements.

Our knowledge of human sensorimotor learning and memory is predominantly based on the visuospatial workspace and limb movements. Humans also have a remarkable ability to produce and perceive speech sounds. We asked whether the human speech-auditory system could serve as a model to characterize the retention of sensorimotor memory in a workspace that is functionally independent of the visuospatial one. Using adaptation to altered auditory feedback, we investigated the durability of a newly acquired speech-acoustical memory (8- and 24-h delay), its sensitivity to the manner of acquisition (abrupt vs. gradual perturbation), and factors affecting memory retrieval. We observed extensive retention of learning (70%) but found no evi- dence for ofﬂine gains. The speech-acoustical memory was insensitive to the manner of its acquisition. To assess factors affecting memory retrieval, tests were ﬁrst done in the absence of auditory feedback (with masking noise). Under these condi- tions, it appeared there was no memory for prior learning as if after an overnight delay, speakers had returned to their habit- ual speech production modes. However, when speech was reintroduced, resulting in speech error feedback, speakers returned immediately to their fully adapted state. This rapid switch shows that the two modes of speech production (adapted and habitual) can coexist in parallel in sensorimotor memory. The ﬁndings demonstrate extensive persistence of speech- acoustical memory and reveal context-speciﬁc memory retrieval processes in speech-motor learning. We conclude that the human speech-auditory system can be used to characterize sensorimotor memory in a workspace that is distinct from the visuospatial workspace. NEW & NOTEWORTHY There is extensive retention of speech-motor learning. Two parallel modes exist in speech motor mem- ory, one with access to everyday habitual speech and the other with access to newly learned speech-acoustical maps. The avail- ability of speech error feedback triggers a switch between these two modes. Properties of sensorimotor memory in the human speech-auditory system are behaviorally similar to, but functionally independent of, their visuospatial counterparts.

Studies using magnetic brain stimulation indicate the involvement of somatosensory regions in the acquisition and retention of newly learned movements. Recent work found an impairment in motor memory when retention was tested shortly after the appli- cation of continuous theta-burst stimulation (cTBS) to the primary somatosensory cortex, compared with stimulation of the primary motor cortex or a control zone. This finding that the somatosensory cortex is involved in motor memory retention whereas the motor cortex is not, if confirmed, could alter our understanding of human motor learning. It would indicate that plasticity in sensory systems underlies newly learned movements, which is different than the commonly held view that adaptation learning involves updates to a motor controller. Here we test this idea. Participants were trained in a visuomotor adaptation task, with visual feedback gradually shifted. Following adaptation, cTBS was applied either to M1, S1, or an occipital cortex control area. Participants were tested for retention 24 h later. It was observed that S1 stimulation led to reduced retention of prior learning, compared with stimulation of M1 or the control area (with no significant difference between M1 and control). In a further control, cTBS was applied to S1 following training with unrotated feedback, in which no learning occurred. This had no effect on movement in the retention test indicating the effects of S1 stimulation on movement are learning specific. The findings are consistent with the S1 participation in the encoding of learning-related changes to movements and in the retention of human motor memory.

Recent studies have indicated somatosensory cortex involvement in motor learning and retention. However, the nature of its contribution is unknown. One possibility is that the somatosensory cortex is transiently engaged during movement. Alternatively, there may be durable learning-related changes which would indicate sensory participation in the encoding of learned movements. These possibilities are dissociated by disrupting the somatosensory cortex following learning, thus targeting learning-related changes which may have occurred. If changes to the somatosensory cortex contribute to retention, which, in effect, means aspects of newly learned movements are encoded there, disruption of this area once learning is complete should lead to an impairment. Participants were trained to make movements while receiving rotated visual feedback. The primary motor cortex (M1) and the primary somatosensory cortex (S1) were targeted for continuous theta-burst stimulation, while stimulation over the occipital cortex served as a control. Retention was assessed using active movement reproduction, or recognition testing, which involved passive movements produced by a robot. Disruption of the somatosensory cortex resulted in impaired motor memory in both tests. Suppression of the motor cortex had no impact on retention as indicated by comparable retention levels in control and motor cortex conditions. The effects were learning specific. When stimulation was applied to S1 following training with unrotated feedback, movement direction, the main dependent variable, was unaltered. Thus, the somatosensory cortex is part of a circuit that contributes to retention, consistent with the idea that aspects of newly learned movements, possibly learning-updated sensory states (new sensory targets) which serve to guide movement, may be encoded there.

This study tests for a function of the somatosensory cortex, that, in addition to its role in processing somatic afferent information, somatosensory cortex contributes both to motor learning and the stabilization of motor memory. Continuous theta-burst magnetic stimulation (cTBS) was applied, before force-field training to disrupt activity in either the primary somatosensory cortex, primary motor cortex, or a control zone over the occipital lobe. Tests for retention and relearning were conducted after a 24 h delay. Analysis of movement kinematic measures and force-channel trials found that cTBS to somatosensory cortex disrupted both learning and subsequent retention, whereas cTBS to motor cortex had little effect on learning but possibly impaired retention. Basic movement variables are unaffected by cTBS suggesting that the stimulation does not interfere with movement but instead disrupts changes in the cortex that are necessary for learning. In all experimental conditions, relearning in an abruptly introduced force field, which followed retention testing, showed extensive savings, which is consistent with previous work suggesting that more cognitive aspects of learning and retention are not dependent on either of the cortical zones under test. Taken together, the findings are consistent with the idea that motor learning is dependent on learning-related activity in the somatosensory cortex. NEW & NOTEWORTHY This study uses noninvasive transcranial magnetic stimulation to test the contribution of somatosensory and motor cortex to human motor learning and retention. Continuous theta-burst stimulation is applied before learning; participants return 24 h later to assess retention. Disruption of the somatosensory cortex is found to impair both learning and retention, whereas disruption of the motor cortex has no effect on learning. The findings are consistent with the idea that motor learning is dependent upon learning-related plasticity in somatosensory cortex.

Speech perception is known to be a multimodal process, relying not only on auditory input but also on the visual system and possibly on the motor system as well. To date there has been little work on the potential involvement of the somatosensory sys- tem in speech perception. In the present review, we identify the somatosensory system as another contributor to speech per- ception. First, we argue that evidence in favor of a motor contribution to speech perception can just as easily be interpreted as showing somatosensory involvement. Second, physiological and neuroanatomical evidence for auditory-somatosensory interac- tions across the auditory hierarchy indicates the availability of a neural infrastructure that supports somatosensory involvement in auditory processing in general. Third, there is accumulating evidence for somatosensory involvement in the context of speech specifically. In particular, tactile stimulation modifies speech perception, and speech auditory input elicits activity in somatosen- sory cortical areas. Moreover, speech sounds can be decoded from activity in somatosensory cortex; lesions to this region affect perception, and vowels can be identified based on somatic input alone. We suggest that the somatosensory involvement in speech perception derives from the somatosensory-auditory pairing that occurs during speech production and learning. By bringing together findings from a set of studies that have not been previously linked, the present article identifies the somato- sensory system as a presently unrecognized contributor to speech perception.

Retention tests conducted after sensorimotor adaptation frequently exhibit a rapid return to baseline performance once the altered sensory feedback is removed. This so-called washout of learning stands in contrast with other demonstrations of retention, such as savings on re-learning and anterograde interference effects of initial learning on new learning. In the present study, we tested the hypothesis that washout occurs when there is a detectable discrepancy in retention tests between visual information on the target position and somatosensory information on the position of the limb. Participants were tested following adaptation to gradually rotated visual feedback (15 degree or 30 degree). Two different types of targets were used for retention testing, a point target in which a perceptual mismatch is possible, and an arc-target that eliminated the mismatch. It was found that, except when point targets were used, retention test movements were stable throughout aftereffect trials, indicating little loss of information. Substantial washout was only observed in tests with a single point target, following adaptation to a large amplitude 30 degree rotation. In control studies designed to minimize the use of explicit strategies during learning, we observed similar patterns of decay when participants moved to point targets that suggests that the effects observed here relate primarily to implicit learning. The results suggest that washout in aftereffect trials following visuomotor adaptation is due to a detectable mismatch between vision and somatosensation. When the mismatch is removed experimentally, there is little evidence of loss of information.NEW & NOTEWORTHY Aftereffects following sensorimotor adaptation are important because they bear on the understanding of the mechanisms that subserve forgetting. We present evidence that information loss previously reported during retention testing occurs only when there is a detectable discrepancy between vision and somatosensation and, if this mismatch is removed, the persistence of adaptation is observed. This suggests that washout during aftereffect trials is a consequence of the experimental design rather than a property of the memory system itself.

This study assesses the involvement in human motor learning, of the ventrolateral prefrontal cortex (BA 9/46v), a somatic region in the middle frontal gyrus. The potential involvement of this cortical area in motor learning is suggested by studies in nonhuman primates which have found anatomic connections between this area and sensorimotor regions in frontal and parietal cortex, and also with basal ganglia output zones. It is likewise sug- gested by electrophysiological studies which have shown that activity in this region is implicated in somatic sensory memory and is also influenced by reward. We directly tested the hypothesis that area 9/46v is in- volved in reinforcement-based motor learning in humans. Participants performed reaching movements to a hidden target and received positive feedback when successful. Before the learning task, we applied continu- ous theta burst stimulation (cTBS) to disrupt activity in 9/46v in the left or right hemisphere. A control group received sham cTBS. The data showed that cTBS to left 9/46v almost entirely eliminated motor learning, whereas learning was not different from sham stimulation when cTBS was applied to the same zone in the right hemisphere. Additional analyses showed that the basic reward-history-dependent pattern of movements was preserved but more variable following left hemisphere stimulation, which suggests an overall deficit in so- matic memory for target location or target directed movement rather than reward processing per se. The re- sults indicate that area 9/46v is part of the human motor learning circuit.

Reinforcement learning has been used as an experimental model of motor skill acquisition, where at times movements are suc- cessful and thus reinforced. One fundamental problem is to understand how humans select exploration over exploitation during learning. The decision could be influenced by factors such as task demands and reward availability. In this study, we applied a clustering algorithm to examine how a change in the accuracy requirements of a task affected the choice of exploration over ex- ploitation. Participants made reaching movements to an unseen target using a planar robot arm and received reward after each successful movement. For one group of participants, the width of the hidden target decreased after every other training block. For a second group, it remained constant. The clustering algorithm was applied to the kinematic data to characterize motor learning on a trial-to-trial basis as a sequence of movements, each belonging to one of the identified clusters. By the end of learning, movement trajectories across all participants converged primarily to a single cluster with the greatest number of suc- cessful trials. Within this analysis framework, we defined exploration and exploitation as types of behavior in which two succes- sive trajectories belong to different or similar clusters, respectively. The frequency of each mode of behavior was evaluated over the course of learning. It was found that by reducing the target width, participants used a greater variety of different clusters and displayed more exploration than exploitation. Excessive exploration relative to exploitation was found to be detrimental to subsequent motor learning. NEW & NOTEWORTHY The choice of exploration versus exploitation is a fundamental problem in learning new motor skills through reinforcement. In this study, we employed a data-driven approach to characterize movements on a trial-by-trial basis with an unsupervised clustering algorithm. Using this technique, we found that changes in task demands and, in particular, in the required accuracy of movements, influenced the ratio of exploration to exploitation. This analysis framework provides an attractive tool to investigate mechanisms of explorative and exploitative behavior while studying motor learning.

Evidences of perceptual changes that accompany motor activity have been limited primarily to audition and somatosensation. Here we asked whether motor learning results in changes to visual motion perception. We designed a reaching task in which participants were trained to make movements along several directions, while the visual feedback was provided by an intrinsically ambiguous moving stimulus directly tied to hand motion. We find that training improves coherent motion perception and that changes in movement are correlated with perceptual changes. No perceptual changes are observed in passive training even when observers were provided with an explicit strategy to facilitate single motion perception. A Bayesian model suggests that movement training promotes the fine-tuning of the internal representation of stimulus geometry. These results emphasize the role of sensorimotor interaction in determining the persistent properties in space and time that define a percept.

The development of the human brain continues through to early adulthood. It has been suggested that cortical plasticity during this protracted period of development shapes circuits in associative transmodal regions of the brain. Here we considered how cortical plasticity during development might contribute to the coordinated brain activity required for speech motor learning. Specifically, we examined patterns of brain functional connectivity whose strength covaried with the capacity for speech audio-motor adaptation in children ages 5-12 and in young adults of both sexes. Children and adults showed distinct patterns of the encoding of learning in the brain. Adult performance was associated with connectivity in transmodal regions that integrate auditory and somatosensory information, whereas children rely on basic somatosensory and motor circuits. A progressive reliance on transmodal regions is consistent with human cortical development and suggests that human speech motor adaptation abilities are built on cortical remodeling that is observable in late childhood and is stabilized in adults.

Motor skill retention is typically measured by asking participants to reproduce previously learned movements from memory. The analog of this retention test (recall memory) in human verbal memory is known to under-estimate how much learning is actually retained. Here we asked whether information about previously learned movements, which can no longer be reproduced, is also retained. Following visuomotor adaptation,we used tests of recall that involved reproduction of previously learned movements and tests of recognition in which participants were asked whether a candidate limb displacement, produced by a robot arm held by the subject, corresponded to a movement direction that was experienced during active training. The main finding was that 24 h after training, estimates of recognition memory were about twice as accurate as those of recall memory. Thus, there is information about previously learned movements that is not retrieved using recall testing but can be accessed in tests of recognition. We conducted additional tests to assess whether,24 h after learning, recall for previously learned movements could be improved by presenting passive movements as retrieval cues. These tests were conducted immediately prior to recall testing and involved the passive playback of a small number of movements, which were spread across the workspace and included both adapted and baseline movements, without being marked as such. This technique restored recall memory for movements to levels close to those of recognition memory performance. Thus, somatic information may enable retrieval of otherwise inaccessible motor memories.

One of the puzzles of learning to talk or play a musical instrument is how we learn which movement produces a particular sound: an audiomotor map. The initial stages of map acquisition can be studied by having participants learn arm movements to auditory targets. The key question is what mechanism drives this early learning. Three learning processes from previous literature were tested: map learning may rely on active motor outflow (target), on error correction, and on the correspondence between sensory and motor distances (i.e., that similar movements map to similar sounds). Alternatively, we hypothesized that map learning can proceed without these. Participants made movements that were mapped to sounds in a number of different conditions that each precluded one of the potential learning processes.We tested whether map learning relies on assumptions about topological continuity by exposing participants to a permuted map that did not preserve distances in auditory and motor space. Further groups were tested who passively experienced the targets, kinematic trajectories produced by a robot arm, and auditory feedback as a yoked active participant (hence without active motor outflow). Another group made movements without receiving targets (thus without experiencing errors). In each case we observed substantial learning,therefore none of the three hypothesized processes is required for learning. Instead early map acquisition can occur with free exploration without target error correction, is based on sensory-to-sensory correspondences, and possible even for discontinuous maps. The findings are consistent with the idea that early sensorimotor map formation can involve instance-specific learning. NEW & NOTEWORTHY: This study tested learning of novel sensorimotor maps in a variety of unusual circumstances, including learning a mapping that was permuted in such as way that it fragmented the sensorimotor workspace into discontinuous parts, thus no preserving sensory and motor topology. Participants could learn this mapping, and they could learn without motor outflow or targets. These results point to a robust learning mechanism building on individual instances, inspired from machine learning literature.

Speech learning requires precise motor control, but it likewise requires transient storage of information to enable the adjustment of upcoming movements based on the success or failure of previous attempts.The contribution of somatic sensory memory for limb position has been documented in work on arm movement; however, in speech,the sensory support for speech production comes from both somatosensory and auditory inputs, and accordingly sensory memory for either or both of sounds and somatic inputs might contribute to learning. In the present study, adaptation to altered auditory feed-back was used as an experimental model of speech motor learning.Participants also underwent tests of both auditory and somatic sensory memory. We found that although auditory memory for speech sounds is better than somatic memory for speech-like facial skin deformations, somatic sensory memory predicts adaptation, where as auditory sensory memory does not. Thus even though speech relies substantially on auditory inputs and in the present manipulation adaptation requires the minimization of auditory error, it is somatic inputs that provide the memory support for learning. NEW & NOTEWORTHY: In speech production, almost everyone achieves an exceptionally high level of proficiency. This is remark-able because speech involves some of the smallest and most care-fully timed movements of which we are capable. The present paper demonstrates that sensory memory contributes to speech motor learning. Moreover, we report the surprising result that somatic sensory memory predicts speech motor learning, whereas auditory memory does not.

Auditory speech perception enables listeners to access phonological categories from speech sounds. During speech production and speech motor learning, speakers' experience matched auditory and somatosensory input. Accordingly, access to phonetic units might also be provided by somatosensory information.The present study assessed whether humans can identify vowels using somatosensory feedback, without auditory feedback.A tongue-positioning task was used in which participants were required to achieve different tongue postures within the /e,?, a/ articulatory range, in a procedure that was totally non-speech like, involving distorted visual feedback of tongue shape.Tongue postures were measured using electromagnetic articulography. At the end of each tongue-positioning trial, subjects were required to whisper the corresponding vocal tract configuration with masked auditory feedback and to identify the vowel associated with the reached tongue posture. Masked auditory feedback ensured that vowel categorization was based on somatosensory feedback rather than auditory feedback. A separate group of subjects was required to auditorily classify the whispered sounds.In addition, we modeled the link between vowel categories and tongue postures in normal speech production with a Bayesian classifier based on the tongue postures recorded from the same speakers for several repetitions of the /e,?, a/ vowels during a separate speech production task. Overall, our results indicate that vowel categorization is possible with somatosensory feed-back alone, with an accuracy that is similar to the accuracy of the auditory perception of whispered sounds, and in congruence with normal speech articulation, as accounted for by the Bayesian classifier.

Motor learning is associated with plasticity in both motor and somatosensory cortex. It is known from animal studies that tetanic stimulation to each of these areas individually induces long-term potentiation in its counterpart. In this context it is possible that changes in motor cortex contribute to somatosensory change and that changes in somatosensory cortex are involved in changes in motor areas of the brain. It is also possible that learning-related plasticity occurs in these areas independently. Tobetter understand the relative contribution to human motor learning ofmotor cortical and somatosensory plasticity, we assessed the time course of changes in primary somatosensory and motor cortex excitability during motor skill learning. Learning was assessed using aforce production task in which a target force profile varied from one trial to the next. The excitability of primary somatosensory cortex was measured using somatosensory evoked potentials in response to median nerve stimulation. The excitability of primary motor cortex was measured using motor evoked potentials elicited by single-pulse transcranial magnetic stimulation. These two measures were inter-leaved with blocks of motor learning trials. We found that the earliest changes in cortical excitability during learning occurred in somatosensory cortical responses, and these changes preceded changes inmotor cortical excitability. Changes in somatosensory evoked potentials were correlated with behavioral measures of learning. Changes in motor evoked potentials were not. These findings indicate that plasticity in somatosensory cortex occurs as a part of the earliest stages of motor learning, before changes in motor cortex are observed.NEW & NOTEWORTHY: We tracked somatosensory and motorcortical excitability during motor skill acquisition. Changes in both motor cortical and somatosensory excitability were observed during learning; however, the earliest changes were in somatosensory cortex,not motor cortex. Moreover, the earliest changes in somatosensory cortical excitability predict the extent of subsequent learning; those in motor cortex do not. This is consistent with the idea that plasticity insomatosensory cortex coincides with the earliest stages of human motor learning.

Previous work has shown that motor learning is associated with changes to both movements and to the somatosensory perception of limb position. In an earlier study that motivates the current work, it appeared that following washout trials, movements did not return to baseline but rather were aligned with associated changes to sensed limb position. Here, we provide a systematic test of this relationship, examining the idea that adaptation-related changes to sensed limb position and to the path of the limb are linked, not only after washout trials but at all stages of the adaptation process. We used a force-field adaptation paradigm followed by washout trials in which subjects performed movements without visual feedback of the limb. Tests of sensed limb position were conducted at each phase of adaptation, specifically before and after baseline movements in a null field, after force-field adaptation, and following washout trials in a null field. As in previous work, sensed limb position changed in association with force-field adaptation. At each stage of adaptation, we observed a correlation between the sensed limb position and associated path of the limb. At a group level, there were differences between the clockwise and counter-clockwise conditions. However, whenever there were changes in sensed limb position, movements following washout did not return to baseline. This suggests that adaptation in sensory and motor systems is not independent processes but rather sensorimotor adaptation is linked to sensory change. Sensory change and limb movement remain in alignment throughout adaptation such that the path of the limb is aligned with the altered sense of limb position.

Early stages of sensorimotor map acquisition: learning with free exploration, without active movement or global structure. J Neurophysiol 122: 1708-1720, 2019. First published August 21, 2019; doi:10.1152/jn.00429.2019. One of the puzzles of learning to talk or play a musical instrument is how we learn which movement produces a particular sound: an audiomotor map. The initial stages of map acquisition can be studied by having participants learn arm movements to auditory targets. The key question is what mechanism drives this early learning. Three learning processes from previous literature were tested: map learning may rely on active motor outflow (target), on error correction, and on the correspondence between sensory and motor distances (i.e., that similar movements map to similar sounds). Alternatively, we hypothesized that map learning can proceed without these. Participants made movements that were mapped to sounds in a number of different conditions that each precluded one of the potential learning processes. We tested whether map learning relies on assumptions about topological continuity by exposing participants to a permuted map that did not preserve distances in auditory and motor space. Further groups were tested who passively experienced the targets, kinematic trajectories produced by a robot arm, and auditory feedback as a yoked active participant (hence without active motor outflow). Another group made movements without receiving targets (thus without experiencing errors). In each case we observed substantial learning, therefore none of the three hypothesized processes is required for learning. Instead early map acquisition can occur with free exploration without target error correction, is based on sensory-to-sensory correspondences, and possible even for discontinuous maps. The findings are consistent with the idea that early sensorimotor map formation can involve instance-specific learning. NEW & NOTEWORTHY This study tested learning of novel sensorimotor maps in a variety of unusual circumstances, including learning a mapping that was permuted in such as way that it fragmented the sensorimotor workspace into discontinuous parts, thus not preserving sensory and motor topology. Participants could learn this mapping, and they could learn without motor outflow or targets. These results point to a robust learning mechanism building on individual instances, inspired from machine learning literature.

Newly learned motor skills are initially labile and then consolidated to permit retention. The circuits that enable the consolidation of motor memories remain uncertain. Most work to date has focused on primary motor cortex, and although there is ample evidence of learning-related plasticity in motor cortex, direct evidence for its involvement in memory consolidation is limited. Learning-related plasticity is also observed in somatosensory cortex, and accordingly, it may also be involved in memory consolidation. Here, by using transcranial magnetic stimulation (TMS) to block consolidation, we report the first direct evidence that plasticity in somatosensory cortex participates in the consolidation of motor memory. Participants made movements to targets while a robot applied forces to the hand to alter somatosensory feedback. Immediately following adaptation, continuous theta-burst transcranial magnetic stimulation (cTBS) was delivered to block retention; then, following a 24-hour delay, which would normally permit consolidation, we assessed whether there was an impairment. It was found that when mechanical loads were introduced gradually to engage implicit learning processes, suppression of somatosensory cortex following training almost entirely eliminated retention. In contrast, cTBS to motor cortex following learning had little effect on retention at all; retention following cTBS to motor cortex was not different than following sham TMS stimulation. We confirmed that cTBS to somatosensory cortex interfered with normal sensory function and that it blocked motor memory consolidation and not the ability to retrieve a consolidated motor memory. In conclusion, the findings are consistent with the hypothesis that in adaptation learning, somatosensory cortex rather than motor cortex is involved in the consolidation of motor memory.

Motor learning is associated with plasticity in both motor and somatosensory cortex. It is known from animal studies that tetanic stimulation to each of these areas individually induces long-term potentiation in its counterpart. In this context it is possible that changes in motor cortex contribute to somatosensory change and that changes in somatosensory cortex are involved in changes in motor areas of the brain. It is also possible that learning related plasticity occurs in these areas independently. To better understand the relative contribution to human motor learning of motor cortical and somatosensory plasticity, we assessed the time course of changes in primary somatosensory and motor cortex excitability during motor skill learning. Learning was assessed using a force production task in which a target force profile varied from one trial to the next. The excitability of primary somatosensory cortex was measured using somatosensory evoked potentials in response to median nerve stimulation. The excitability of primary motor cortex was measured using motor evoked potentials elicited by single-pulse transcranial magnetic stimulation. These two measures were interleaved with blocks of motor learning trials. We found that the earliest changes in cortical excitability during learning occurred in somatosensory cortical responses and these changes preceded changes in motor cortical excitability. Changes in somatosensory evoked potentials were correlated with behavioral measures of learning. Changes in motor evoked potentials were not. These findings indicate that plasticity in somatosensory cortex occurs as a part of the earliest stages of motor learning, before changes in motor cortex are observed.

The relationship between neural activation during movement training and the plastic changes that survive beyond movement execution is not well understood. Here we ask whether the changes in resting-state functional connectivity observed following motor learning overlap with the brain networks that track movement error during training. Human participants learned to trace an arched trajectory using a computer mouse in an MRI scanner. Motor performance was quantified on each trial as the maximum distance from the prescribed arc. During learning, two brain networks were observed, one showing increased activations for larger movement error, comprising the cerebellum, parietal, visual, somatosensory, and cortical motor areas, and the other being more activated for movements with lower error, comprising the ventral putamen and the OFC. After learning, changes in brain connectivity at rest were found predominantly in areas that had shown increased activation for larger error during task, specifically the cerebellum and its connections with motor, visual, and somatosensory cortex. The findings indicate that, although both errors and accurate movements are important during the active stage of motor learning, the changes in brain activity observed at rest primarily reflect networks that process errors. This suggests that error-related networks are represented in the initial stages of motor memory formation.

Recent studies using visuomotor adaptation and sequence learning tasks have assessed the involvement of working memory in the visuospatial domain. The capacity to maintain previously performed movements in working memory is perhaps even more important in reinforcement-based learning to repeat accurate movements and avoid mistakes. Using this kind of task in the present work, we tested the relationship between somatosensory working memory and motor learning. The first experiment involved separate memory and motor learning tasks. In the memory task, the participant's arm was displaced in different directions by a robotic arm, and the participant was asked to judge whether a subsequent test direction was one of the previously presented directions. In the motor learning task, participants made reaching movements to a hidden visual target and were provided with positive feedback as reinforcement when the movement ended in the target zone. It was found that participants that had better somatosensory working memory showed greater motor learning. In a second experiment, we designed a new task in which learning and working memory trials were interleaved, allowing us to study participants memory for movements they performed as part of learning. As in the first experiment, we found that participants with better somatosensory working memory also learned more. Moreover, memory performance for successful movements was better than for movements that failed to reach the target. These results suggest that somatosensory working memory is involved in reinforcement motor learning and that this memory preferentially keeps track of reinforced movements. NEW & NOTEWORTHY The present work examined somatosensory working memory in reinforcement-based motor learning. Working memory performance was reliably correlated with the extent of learning. With the use of a paradigm in which learning and memory trials were interleaved, memory was assessed for movements performed during learning. Movements that received positive feedback were better remembered than movements that did not. Thus working memory does not track all movements equally but is biased to retain movements that were rewarded.

Background. Passive robot-generated arm movements in conjunction with proprioceptive decision making and feedback modulate functional connectivity (FC) in sensory motor networks and improve sensorimotor adaptation in normal individuals. This proof-of-principle study investigates whether these effects can be observed in stroke patients. Methods. A total of 10 chronic stroke patients with a range of stable motor and sensory deficits (Fugl-Meyer Arm score [FMA] 0-65, Nottingham Sensory Assessment [NSA] 10-40) underwent resting-state functional magnetic resonance imaging before and after a single session of robot-controlled proprioceptive training with feedback. Changes in FC were identified in each patient using independent component analysis as well as a seed region-based approach. FC changes were related to impairment and changes in task performance were assessed. Results. A single training session improved average arm reaching accuracy in 6 and proprioception in 8 patients. Two networks showing training-associated FC change were identified. Network C1 was present in all patients and network C2 only in patients with FM scores >7. Relatively larger C1 volume in the ipsilesional hemisphere was associated with less impairment (r = 0.83 for NSA, r = 0.73 for FMA). This association was driven by specific regions in the contralesional hemisphere and their functional connections (supramarginal gyrus with FM scores r = 0.82, S1 with NSA scores r = 0.70, and cerebellum with NSA score r = ?0.82). Conclusion. A single session of robot-controlled proprioceptive training with feedback improved movement accuracy and induced FC changes in sensory motor networks of chronic stroke patients. FC changes are related to functional impairment and comprise bilateral sensory and motor network nodes.

Adaptation to an abrupt change in the dynamics of the interaction between the arm and the physical environment has been reported as occurring more rapidly but with less retention than adaptation to a gradual change in interaction dynamics. Faster adaptation to an abrupt change in interaction dynamics appears inconsistent with kinematic error sensitivity which has been shown to be greater for small errors than large errors. However, the comparison of adaptation rates was based on incomplete adaptation. Furthermore, the metric which was used as a proxy of the changing internal state, namely the linear regression between the force disturbance and the compensatory force (the adaptation index), does not distinguish between internal state inaccuracy resulting from amplitude or temporal errors. To resolve the apparent inconsistency, we compared the evolution of the internal state during complete adaptation to an abrupt and gradual change in interaction dynamics. We found no difference in the rate at which the adaptation index increased during adaptation to a gradual compared to an abrupt change in interaction dynamics. In addition, we separately examined amplitude and temporal errors using different metrics, and found that amplitude error was reduced more rapidly under the gradual than the abrupt condition, whereas temporal error (quantified by smoothness) was reduced more rapidly under the abrupt condition. We did not find any significant change in phase lag during adaptation under either condition. Our results also demonstrate that even after adaptation is complete, online feedback correction still plays a significant role in the control of reaching.

One of the puzzles of learning to talk or play a musical instrument is how we learn which movement produces a particular sound: an audiomotor map. Existing research has used mappings that are already well learned such as controlling a cursor using a computer mouse. By contrast, the acquisition of novel sensorimotor maps was studied by having participants learn arm movements to auditory targets. These sounds did not come from different directions but, like speech, were only distinguished by their frequencies. It is shown that learning involves forming not one but two maps: a point map connecting sensory targets with motor commands and an error map linking sensory errors to motor corrections. Learning a point map is possible even when targets never repeat. Thus, although participants make errors, there is no opportunity to correct them because the target is different on every trial, and therefore learning cannot be driven by error correction. Furthermore, when the opportunity for error correction is provided, it is seen that acquiring error correction is itself a learning process that changes over time and results in an error map. In principle, the error map could be derived from the point map, but instead, these two maps are independently acquired and jointly enable sensorimotor control and learning. A computational model shows that this dual encoding is optimal and simulations based on this architecture predict that learning the two maps results in performance improvements comparable with those observed empirically.

As one learns to dance or play tennis, the desired somatosensory state is typically unknown. Trial and error is important as motor behavior is shaped by successful and unsuccessful movements. As an experimental model, we designed a task in which human participants make reaching movements to a hidden target and receive positive reinforcement when successful. We identified somatic and reinforcement-based sources of plasticity on the basis of changes in functional connectivity using resting-state fMRI before and after learning. The neuroimaging data revealed reinforcement-related changes in both motor and somatosensory brain areas in which a strengthening of connectivity was related to the amount of positive reinforcement during learning. Areas of prefrontal cortex were similarly altered in relation to reinforcement, with connectivity between sensorimotor areas of putamen and the reward-related ventromedial prefrontal cortex strengthened in relation to the amount of successful feedback received. In other analyses, we assessed connectivity related to changes in movement direction between trials, a type of variability that presumably reflects exploratory strategies during learning. We found that connectivity in a network linking motor and somatosensory cortices increased with trial-to-trial changes in direction. Connectivity varied as well with the change in movement direction following incorrect movements. Here the changes were observed in a somatic memory and decision making network involving ventrolateral prefrontal cortex and second somatosensory cortex. Our results point to the idea that the initial stages of motor learning are not wholly motor but rather involve plasticity in somatic and prefrontal networks related both to reward and exploration.

In the present paper, we present evidence for the idea that speech motor learning is accompanied by changes to the neural coding of both auditory and somatosensory stimuli. Participants in our experiments undergo adaptation to altered auditory feedback, an experimental model of speech motor learning which like visuo-motor adaptation in limb movement, requires that participants change their speech movements and associated somatosensory inputs to correct for systematic real-time changes to auditory feedback. We measure the sensory effects of adaptation by examining changes to auditory and somatosensory event-related responses. We find that adaptation results in progressive changes to speech acoustical outputs that serve to correct for the perturbation. We also observe changes in both auditory and somatosensory event-related responses that are correlated with the magnitude of adaptation. These results indicate that sensory change occurs in conjunction with the processes involved in speech motor adaptation.

There is accumulating evidence from behavioral, neurophysiological, and neuroimaging studies that the acquisition of motor skills involves both perceptual and motor learning. Perceptual learning alters movements, motor learning, and motor networks of the brain. Motor learning changes perceptual function and the sensory circuits of the brain. Here, we review studies of both human limb movement and speech that indicate that plasticity in sensory and motor systems is reciprocally linked. Taken together, this points to an approach to motor learning in which perceptual learning and sensory plasticity have a fundamental role. Trends Sensorimotor adaptation results in changes to sensory systems and sensory networks in the brain. Perceptual learning modifies sensory systems and directly alters the motor networks of the brain. Perceptual changes associated with sensorimotor adaptation are durable and occur in parallel with motor learning.

Cortical processing associated with orofacial somatosensory function in speech has received limited experimental attention due to the difficulty of providing precise and controlled stimulation. This article introduces a technique for recording somatosensory event-related potentials (ERP) that uses a novel mechanical stimulation method involving skin deformation using a robotic device. Controlled deformation of the facial skin is used to modulate kinesthetic inputs through excitation of cutaneous mechanoreceptors. By combining somatosensory stimulation with electroencephalographic recording, somatosensory evoked responses can be successfully measured at the level of the cortex. Somatosensory stimulation can be combined with the stimulation of other sensory modalities to assess multisensory interactions. For speech, orofacial stimulation is combined with speech sound stimulation to assess the contribution of multi-sensory processing including the effects of timing differences. The ability to precisely control orofacial somatosensory stimulation during speech perception and speech production with ERP recording is an important tool that provides new insight into the neural organization and neural representations for speech.

The early stages of motor skill acquisition are often marked by uncertainty about the sensory and motor goals of the task, as is the case in learning to speak or learning the feel of a good tennis serve. Here we present an experimental model of this early learning process, in which targets are acquired by exploration and reinforcement rather than sensory error. We use this model to investigate the relative contribution of motor and sensory factors to human motor learning. Participants make active reaching movements or matched passive movements to an unseen target using a robot arm. We find that learning through passive movements paired ith reinforcement is comparable with learning associated with active movement, both in terms of magnitude and durability, with improvements due to training still observable at a 1 week retest. Motor learning is also accompanied by changes in somatosensory perceptual acuity. No stable changes in motor performance are observed for participants that train, actively or passively, in the absence of reinforcement, or for participants who are given explicit information about target position in the absence of somatosensory experience. These findings indicate that the somatosensory system dominates learning in the early stages of motor skill acquisition.

Recent studies of human speech motor learning suggest that learning is accompanied by changes in auditory perception. But what drives the perceptual change? Is it a consequence of changes in the motor system? Or is it a result of sensory inflow during learning? Here, subjects participated in a speech motor-learning task involving adaptation to altered auditory feedback and they were subsequently tested for perceptual change. In two separate experiments, involving two different auditory perceptual continua, we show that changes in the speech motor system that accompany learning drive changes in auditory speech perception. Specifically, we obtained changes in speech perception when adaptation to altered auditory feedback led to speech production that fell into the phonetic range of the speech perceptual tests. However, a similar change in perception was not observed when the auditory feedback that subjects' received during learning fell into the phonetic range of the perceptual tests. This indicates that the central motor outflow associated with vocal sensorimotor adaptation drives changes to the perceptual classification of speech sounds.

The perception of speech is notably malleable in adults, yet alterations in perception seem to have little impact on speech production. However, we hypothesized that speech perceptual training might immediately influence speech motor learning. To test this, we paired a speech perceptual-training task with a speech motor-learning task. Subjects performed a series of perceptual tests designed to measure and then manipulate the perceptual distinction between the words head and had. Subjects then produced head with the sound of the vowel altered in real time so that they heard themselves through headphones producing a word that sounded more like had. In support of our hypothesis, the amount of motor learning in response to the voice alterations depended on the perceptual boundary acquired through perceptual training. The studies show that plasticity in adults' speech perception can have immediate consequences for speech production in the context of speech learning.

PURPOSE:
Somatosensory information associated with speech articulatory movements affects the perception of speech sounds and vice versa, suggesting an intimate linkage between speech production and perception systems. However, it is unclear which cortical processes are involved in the interaction between speech sounds and orofacial somatosensory inputs. The authors examined whether speech sounds modify orofacial somatosensory cortical potentials that were elicited using facial skin perturbations.
METHOD:
Somatosensory event-related potentials in EEG were recorded in 3 background sound conditions (pink noise, speech sounds, and nonspeech sounds) and also in a silent condition. Facial skin deformations that are similar in timing and duration to those experienced in speech production were used for somatosensory stimulation.
RESULTS:
The authors found that speech sounds reliably enhanced the first negative peak of the somatosensory event-related potential when compared with the other 3 sound conditions. The enhancement was evident at electrode locations above the left motor and premotor area of the orofacial system. The result indicates that speech sounds interact with somatosensory cortical processes that are produced by speech-production-like patterns of facial skin stretch.
CONCLUSION:
Neural circuits in the left hemisphere, presumably in left motor and premotor cortex, may play a prominent role in the interaction between auditory inputs and speech-relevant somatosensory processing.

As we begin to acquire a new motor skill, we face the dual challenge of determining and refining the somatosensory goals of our movements and establishing the best motor commands to achieve our ends. The two typically proceed in parallel, and accordingly it is unclear how much of skill acquisition is a reflection of changes in sensory systems and how much reflects changes in the brain's motor areas. Here we have intentionally separated perceptual and motor learning in time so that we can assess functional changes to human sensory and motor networks as a result of perceptual learning. Our subjects underwent fMRI scans of the resting brain before and after a somatosensory discrimination task. We identified changes in functional connectivity that were due to the effects of perceptual learning on movement. For this purpose, we used a neural model of the transmission of sensory signals from perceptual decision making through to motor action. We used this model in combination with a partial correlation technique to parcel out those changes in connectivity observed in motor systems that could be attributed to activity in sensory brain regions. We found that, after removing effects that are linearly correlated with somatosensory activity, perceptual learning results in changes to frontal motor areas that are related to the effects of this training on motor behavior and learning. This suggests that perceptual learning produces changes to frontal motor areas of the brain and may thus contribute directly to motor learning.

Motor learning often involves situations in which the somatosensory targets of movement are initially, poorly defined, as for example, in learning to speak or learning the feel of a proper tennis serve. Under these conditions, motor skill acquisition presumably requires perceptual as well as motor learning. That is, it engages both the progressive shaping of sensory targets and associated changes in motor performance. In the present paper, we test the idea that perceptual learning alters somatosensory function and in so doing produces changes to motor performance and sensorimotor adaptation. Subjects in these experiments undergo perceptual training in which a robotic device passively moves the arm on one of a set of fan shaped trajectories. Subjects are required to indicate whether the robot moved the limb to the right or the left and feedback is provided. Over the course of training both the perceptual boundary and acuity are altered. The perceptual learning is observed to improve both the rate and extent of learning in a subsequent sensorimotor adaptation task and the benefits persist for at least 24 hours. The improvement in the present studies is obtained regardless of whether the perceptual boundary shift serves to systematically increase or decrease error on subsequent movements. The beneficial effects of perceptual training are found to be substantially dependent upon reinforced decision-making in the sensory domain. Passive-movement training on its own is less able to alter subsequent learning in the motor system. Overall, this study suggests perceptual learning plays an integral role in motor learning.

Observing the actions of others has been shown to affect motor learning, but does it have effects on sensory systems as well? It has been recently shown that motor learning that involves actual physical practice is also associated with plasticity in the somatosensory system. Here, we assessed the idea that observational learning likewise changes somatosensory function. We evaluated changes in somatosensory function after human subjects watched videos depicting motor learning. Subjects first observed video recordings of reaching movements either in a clockwise or counterclockwise force field. They were then trained in an actual force-field task that involved a counterclockwise load. Measures of somatosensory function were obtained before and after visual observation and also following force-field learning. Consistent with previous reports, video observation promoted motor learning. We also found that somatosensory function was altered following observational learning, both in direction and in magnitude, in a manner similar to that which occurs when motor learning is achieved through actual physical practice. Observation of the same sequence of movements in a randomized order did not result in somatosensory perceptual change. Observational learning and real physical practice appear to tap into the same capacity for sensory change in that subjects that showed a greater change following observational learning showed a reliably smaller change following physical motor learning. We conclude that effects of observing motor learning extend beyond the boundaries of traditional motor circuits, to include somatosensory representations.

Motor learning in the context of arm reaching movements has been frequently investigated using the paradigm of force-field learning. It has been recently shown that changes to somatosensory perception are likewise associated with motor learning. Changes in perceptual function may be the reason that when the perturbation is removed following motor learning, the hand trajectory does not return to a straight line path even after several dozen trials. To explain the computational mechanisms that produce these characteristics, we propose a motor control and learning scheme using a simplified two-link system in the horizontal plane:We represent learning as the adjustment of desired joint-angular trajectories so as to achieve the reference trajectory of the hand. The convergence of the actual hand movement to the reference trajectory is proved by using a Lyapunov-like lemma, and the result is confirmed using computer simulations. The model assumes that changes in the desired hand trajectory influence the perception of hand position and this in turn affects movement control. Our computer simulations support the idea that perceptual change may come as a result of adjustments to movement planning with motor learning.

Motor learning is reflected in changes to the brain's functional organization as a result of experience. We show here that these changes are not limited to motor areas of the brain and indeed that motor learning also changes sensory systems. We test for plasticity in sensory systems using somatosensory evoked potentials (SEPs). A robotic device is used to elicit somatosensory inputs by displacing the arm in the direction of applied force during learning. We observe that following learning there are short latency changes to the response in somatosensory areas of the brain that are reliably correlated with the magnitude of motor learning: subjects who learn more show greater changes in SEP magnitude. The effects we observe are tied to motor learning. When the limb is displaced passively, such that subjects experience similar movements but without experiencing learning, no changes in the evoked response are observed. Sensorimotor adaptation thus alters the neural coding of somatosensory stimuli.

A complex interplay has been demonstrated between motor and sensory systems. We showed recently that motor learning leads to changes in the sensed position of the limb (Ostry DJ, Darainy M, Mattar AA, Wong J, Gribble PL. J Neurosci 30: 5384-5393, 2010). Here, we document further the links between motor learning and changes in somatosensory perception. To study motor learning, we used a force field paradigm in which subjects learn to compensate for forces applied to the hand by a robotic device. We used a task in which subjects judge lateral displacements of the hand to study somatosensory perception. In a first experiment, we divided the motor learning task into incremental phases and tracked sensory perception throughout. We found that changes in perception occurred at a slower rate than changes in motor performance. A second experiment tested whether awareness of the motor learning process is necessary for perceptual change. In this experiment, subjects were exposed to a force field that grew gradually in strength. We found that the shift in sensory perception occurred even when awareness of motor learning was reduced. These experiments argue for a link between motor learning and changes in somatosensory perception, and they are consistent with the idea that motor learning drives sensory change.

The idea that humans learn and maintain accurate speech by carefully monitoring auditory feedback is widely held. But this view neglects the fact that auditory feedback is highly correlated with somatosensory feedback during speech production. Somatosensory feedback from speech movements could be a primary means by which cortical speech areas monitor the accuracy of produced speech. We tested this idea by placing the somatosensory and auditory systems in competition during speech motor learning. To do this, we combined two speech-learning paradigms to simultaneously alter somatosensory and auditory feedback in real time as subjects spoke. Somatosensory feedback was manipulated by using a robotic device that altered the motion path of the jaw. Auditory feedback was manipulated by changing the frequency of the first formant of the vowel sound and playing back the modified utterance to the subject through headphones. The amount of compensation for each perturbation was used as a measure of sensory reliance. All subjects were observed to correct for at least one of the perturbations, but auditory feedback was not dominant. Indeed, some subjects showed a stable preference for either somatosensory or auditory feedback during speech.

Does motor learning generalize to new situations that are not experienced during training, or is motor learning essentially specific to the training situation? In the present experiments, we use speech production as a model to investigate generalization in motor learning. We tested for generalization from training to transfer utterances by varying the acoustical similarity between these two sets of utterances. During the training phase of the experiment, subjects received auditory feedback that was altered in real time as they repeated a single consonant vowel-consonant utterance. Different groups of subjects were trained with different consonant-vowel-consonant utterances, which differed from a subsequent transfer utterance in terms of the initial consonant or vowel. During the adaptation phase of the experiment, we observed that subjects in all groups progressively changed their speech output to compensate for the perturbation (altered auditory feedback). After learning, we tested for generalization by having all subjects produce the same single transfer utterance while receiving unaltered auditory feedback. We observed limited transfer of learning, which depended on the acoustical similarity between the training and the transfer utterances. The gradients of generalization observed here are comparable to those observed in limb movement. The present findings are consistent with the conclusion that speech learning remains specific to individual instances of learning.

Here we describe two studies linking perceptual change with motor learning. In the first, we document persistent changes in somatosensory perception that occur following force field learning. Subjects learned to control a robotic device that applied forces to the hand during arm movements. This led to a change in the sensed position of the limb that lasted at least 24 h. Control experiments revealed that the sensory change depended on motor learning. In the second study, we describe changes in the perception of speech sounds that occur following speech motor learning. Subjects adapted control of speech movements to compensate for loads applied to the jaw by a robot. Perception of speech sounds was measured before and after motor learning. Adapted subjects showed a consistent shift in perception. In contrast, no consistent shift was seen in control subjects and subjects that did not adapt to the load. These studies suggest that motor learning changes both sensory and motor function.

Motor learning changes the activity of cortical motor and subcortical areas of the brain, but does learning affect sensory systems as well? We examined inhumansthe effects of motor learning using fMRI measures of functional connectivity under resting conditions and found persistent changes in networks involving both motor and somatosensory areas of the brain. We developed a technique that allows us to distinguish changes in functional connectivity that can be attributed to motor learning from those that are related to perceptual changes that occur in conjunction with learning. Using this technique, we identified a new network in motor learning involving second somatosensory cortex, ventral premotor cortex, and supplementary motor cortex whose activation is specifically related to perceptual changes that occur in conjunction with motor learning. We also found changes in a network comprising cerebellar cortex, primary motor cortex, and dorsal premotor cortex that were linked to the motor aspects of learning. In each network, we observed highly reliable linear relationships between neuroplastic changes and behavioral measures of either motor learning or perceptual function. Motor learning thus results in functionally specific changes to distinct resting-state networks in the brain.

The brain easily generates the movement that is needed in a given situation. Yet surprisingly, the results of experimental studies suggest that it is difficult to acquire more than one skill at a time. To do so, it has generally been necessary to link the required movement to arbitrary cues. In the present study, we show that speech motor learning provides an informative model for the acquisition of multiple sensorimotor skills. During training, subjects were required to repeat aloud individual words in random order while auditory feedback was altered in real-time in different ways for the different words. We found that subjects can quite readily and simultaneously modify their speech movements to correct for these different auditory transformations. This multiple learning occurs effortlessly without explicit cues and without any apparent awareness of the perturbation. The ability to simultaneously learn several different auditory-motor transformations is consistent with the idea that, in speech motor learning, the brain acquires instance-specific memories. The results support the hypothesis that speech motor learning is fundamentally local.

Motor learning is dependent on kinesthetic information that is obtained both from cutaneous afferents and from muscle receptors. In human arm movement, information from these two kinds of afferents is largely correlated. The facial skin offers a unique situation in which there are plentiful cutaneous afferents and essentially no muscle receptors and, accordingly, experimental manipulations involving the facial skin may be used to assess the possible role of cutaneous afferents in motor learning. We focus here on the information for motor learning provided by the deformation of the facial skin and the motion of the lips in the context of speech. We used a robotic device to slightly stretch the facial skin lateral to the side of the mouth in the period immediately preceding movement. We found that facial skin stretch increased lip protrusion in a progressive manner over the course of a series of training trials. The learning was manifest in a changed pattern of lip movement, when measured after learning in the absence of load. The newly acquired motor plan generalized partially to another speech task that involved a lip movement of different amplitude. Control tests indicated that the primary source of the observed adaptation was sensory input from cutaneous afferents. The progressive increase in lip protrusion over the course of training fits with the basic idea that change in sensory input is attributed to motor performance error. Sensory input, which in the present study precedes the target movement, is credited to the target-related motion, even though the skin stretch is released prior to movement initiation. This supports the idea that the nervous system generates motor commands on the assumption that sensory input and kinematic error are in register.

Movements are inherently variable. When we move to a particular point in space, a cloud of final limb positions is observed around the target. Previously we noted that
patterns of variability at the end of movement to a circular target were not circular, but instead reflected patterns of limb stiffness in directions where limb stiffness was high, variability in end position was low, and vice versa. Here we examine the determinants of variability at movement end in more detail. To do this, we have subjects move the handle of a robotic device from different starting positions into a circular target. We use position servocontrolled displacements of the robot's handle to measure limb stiffness at the end of movement and we also record patterns of end position variability. To examine the effect of change in posture on movement variability, we use a visual motor transformation in which we change the limb configuration and also the actual movement target, while holding constant the visual display. We find that, regardless of movement direction, patterns of variability at the end of movement vary systematically with limb configuration and are also related to patterns of limb stiffness, which are likewise configuration dependent. The result suggests that postural configuration determines the base level of movement variability, on top of which control mechanisms can act to further alter variability.

Studies on generalization show the nature of how learning is encoded in the brain. Previous studies have shown rather limited generalization of dynamics learning across changes in movement direction, a finding that is consistent with the idea that learning is primarily local. In contrast, studies show a broader pattern of generalization across changes in movement amplitude, suggesting a more general form of learning. To understand this difference, we performed an experiment in which subjects held a robotic manipulandum and made movements to targets along the body midline. Subjects were trained in a velocitydependent force field while moving to a 15 cm target. After training, subjects were tested for generalization using movements to a 30 cm target. We used force channels in conjunction with movements to the 30 cm target to assess the extent of generalization. Force channels restricted lateral movements and allowed us to measure force production during generalization. We compared actual lateral forces to the forces expected if dynamics learning generalized fully. We found that, during the test for generalization, subjects produced reliably less force than expected. Force production was appropriate for the portion of the transfer movement in which velocities corresponded to those experienced with the 15 cm target. Subjects failed to produce the expected forces when velocities exceeded those experienced in the training task. This suggests that dynamics learning generalizes little beyond the range of one's experience. Consistent with this result, subjects who trained on the 30 cm target showed full generalization to the 15 cm target. We performed two additional experiments that show that interleaved trials to the 30 cm target during training on the 15 cm target can resolve the difference between the current results and those reported previously.

Motor learning is dependent upon plasticity in motor areas of the brain, but does it occur in isolation, or does it also result in changes to sensory systems? We examined changes to somatosensory function that occur in conjunction with motor learning. We found that even after periods of training as brief as 10 min, sensed limb position was altered and the perceptual change persisted for 24 h. The perceptual change was reflected in subsequent movements; limb movements following learning deviated from the prelearning trajectory by an amount that was not different in magnitude and in the same direction as the perceptual shift. Crucially, the perceptual change was dependent upon motor learning. When the limb was displaced passively such that subjects experienced similar kinematics but without learning, no sensory change was observed. The findings indicate that motor learning affects not only motor areas of the brain but changes sensory function as well.

Is plasticity in sensory and motor systems linked? Here, in the context of speech motor learning and perception, we test the idea sensory function is modified by motor learning and, in particular, that speech motor learning affects a speaker's auditory map. We assessed speech motor learning by using a robotic device that displaced the jaw and selectively altered somatosensory feedback during speech. We found that with practice speakers progressively corrected for the mechanical perturbation and after motor learning they also showed systematic changes in their perceptual classification of speech sounds. The perceptual shift was tied to motor learning. Individuals that displayed greater amounts of learning also showed greater perceptual change. Perceptual change was not observed in control subjects that produced the same movements, but in the absence of a force field, nor in subjects that experienced the force field but failed to adapt to the mechanical load. The perceptual effects observed here indicate the involvement of the somatosensory system in the neural processing of speech sounds and suggest that speech motor learning results in changes to auditory perceptual function.

Speech production involves some of the most precise and finely timed patterns of human movement. Here, in the context of jaw movement in speech, we show that spatial precision in speech production is systematically associated with the regulation of impedance and in particular, with jaw stiffness a measure of resistance to displacement. We estimated stiffness and also variability during movement using a robotic device to apply brief force pulses to the jaw. Estimates of stiffness were obtained using the perturbed position and force trajectory and an estimate of what the trajectory would be in the absence of load. We estimated this reference trajectory using a new technique based on Fourier analysis. A moving-average (MA) procedure was used to estimate stiffness by modeling restoring force as the moving average of previous jaw displacements. The stiffness matrix was obtained from the steady state of the MA model. We applied this technique to data from 31 subjects whose jaw movements were perturbed during speech utterances and kinematically matched nonspeech movements. We observed systematic differences in stiffness over the course of jaw-lowering and jaw-raising movements that were correlated with measures of kinematic variability. Jaw stiffness was high and variability was low early and late in the movement when the jaw was elevated. Stiffness was low and variability was high in the middle of movement when the jaw was lowered. Similar patterns were observed for speech and nonspeech conditions. The systematic relationship between stiffness and variability points to the idea that stiffness regulation is integral to the control of orofacial movement variability.

Previous studies have demonstrated anisotropic patterns of hand impedance under static conditions and during movement. Here we show that the pattern of kinematic error observed in studies of dynamics learning is associated with this anisotropic impedance pattern. We also show that the magnitude of kinematic error associated with this anisotropy dictates the amount of motor learning and, consequently, the extent to which dynamics learning generalizes. Subjects were trained to reach to visual targets while holding a robotic device that applied forces during movement. On infrequent trials, the load was removed and the resulting kinematic error was measured. We found a strong correlation between the pattern of kinematic error and the anisotropic pattern of hand stiffness. In a second experiment subjects were trained under force-field conditions to move in two directions: one in which the dynamic perturbation was in the direction of maximum arm impedance and the associated kinematic error was low and another in which the perturbation was in the direction of low impedance where kinematic error was high. Generalization of learning was assessed in a reference direction that lay intermediate to the two training directions. We found that transfer of learning was greater when training occurred in the direction associated with the larger kinematic error. This suggests that the anisotropic patterns of impedance and kinematic error determine the magnitude of dynamics learning and the extent to which it generalizes.

Somatosensory signals from the facial skin and muscles of the vocal tract provide a rich source of sensory input in speech production. We show here that the somatosensory system is also involved in the perception of speech. We use a robotic device to create patterns of facial skin deformation that would normally accompany speech production. We find that when we stretch the facial skin while people listen to words, it alters the sounds they hear. The systematic perceptual variation we observe in conjunction with speech-like patterns of skin stretch indicates that somatosensory inputs affect the neural processing of speech sounds and shows the involvement of the somatosensory system in the perceptual processing in speech.

Speech production, like other sensorimotor behaviors, relies on multiple sensory inputs-audition, proprioceptive inputs from muscle spindles and cutaneous inputs from mechanoreceptors in the skin and soft tissues of the vocal tract. However, the capacity for intelligible speech by deaf speakers suggests that somatosensory input alone may contribute to speech motor control and perhaps even to speech learning. We assessed speech motor learning in cochlear implant recipients who were tested with their implants turned off. A robotic device was used to alter somatosensory feedback by displacing the jaw during speech. We found that implant subjects progressively adapted to the mechanical perturbation with training. Moreover, the corrections that we observed were for movement deviations that were exceedingly small, on the order of millimeters, indicating that speakers have precise somatosensory expectations. Speech motor learning is substantially dependent on somatosensory input.

Coactivation of antagonist muscles is readily observed early in motor learning, in interactions with unstable mechanical environments and in motor system pathologies. Here we present evidence that the nervous system uses coactivation control far more extensively and that patterns of cocontraction during movement are closely tied to the specific requirements of the task. We have examined the changes in cocontraction that follow dynamics learning in tasks that are thought to involve finely sculpted feedforward adjustments to motor commands. We find that, even following substantial training, cocontraction varies in a systematic way that depends on both movement direction and the strength of the external load. The proportion of total activity that is due to cocontraction nevertheless remains remarkably constant. Moreover, long after indices of motor learning and electromyographic measures have reached asymptotic levels, cocontraction still accounts for a significant proportion of total muscle activity in all phases of movement and in all load conditions. These results show that even following dynamics learning in predictable and stable environments, cocontraction forms a central part of the means by which the nervous system regulates movement.

In the present study, we recorded the kinematics of grasping movements in order to measure the possible interference caused by digits printed on the visible face of the objects to grasp. The aim of this approach was to test the hypothesis that digit magnitude processing shares common mechanisms with object size estimate during grasping. In the first stages of reaching, grip aperture was found to be larger consequent to the presentation of digits with a high value rather than a low one. The effect of digit magnitude on grip aperture was more pronounced for large objects. As the hand got closer to the object, the influence of digit magnitude decreased and grip aperture progressively reflected the actual size of the object. We concluded that number magnitude may interact with grip aperture while programming the grasping movements.

The idea that the brain controls movement using a neural representation of limb dynamics has been a dominant hypothesis in motor control research for well over a decade. Speech movements offer an unusual opportunity to test this proposal by means of an examination of transfer of learning between utterances that are to varying degrees matched on kinematics. If speech learning results in a generalizable dynamics representation, then, at the least, learning should transfer when similar movements are embedded in phonetically distinct utterances. We tested this idea using three different pairs of training and transfer utterances that substantially overlap kinematically. We find that, with these stimuli, speech learning is highly contextually sensitive and fails to transfer even to utterances that involve very similar movements. Speech learning appears to be extremely local, and the specificity of learning is incompatible with the idea that speech control involves a generalized dynamics representation.

It is known that humans can modify the impedance of the musculoskeletal periphery, but the extent of this modification is uncertain. Previous studies on impedance control under static conditions indicate a limited ability to modify impedance, whereas studies of impedance control during reaching in unstable environments suggest a greater range of impedance modification. As a first step in accounting for this difference, we quantified the extent to which stiffness changes from posture to movement even when there are no destabilizing forces. Hand stiffness was estimated under static conditions and at the same position during both longitudinal (near to far) and lateral movements using a position-servo technique. A new method was developed to predict the hand "reference" trajectory for purposes of estimating stiffness. For movements in a longitudinal direction, there was considerable counterclockwise rotation of the hand stiffness ellipse relative to stiffness under static conditions. In contrast, a small counterclockwise rotation was observed during lateral movement. In the modeling studies, even when we used the same modeled cocontraction level during posture and movement, we found that there was a substantial difference in the orientation of the stiffness ellipse, comparable with that observed empirically. Indeed, the main determinant of the orientation of the ellipse in our modeling studies was the movement direction and the muscle activation associated with movement. Changes in the cocontraction level and the balance of cocontraction had smaller effects. Thus even when there is no environmental instability, the orientation of stiffness ellipse changes during movement in a manner that varies with movement direction.

Humans routinely make movements to targets that have different accuracy requirements in different directions. Examples extend from everyday occurrences such as grasping the handle of a coffee cup to the more refined instance of a surgeon positioning a scalpel. The attainment of accuracy in situations such as these might be related to the nervous system's capacity to regulate the limb's resistance to displacement, or impedance. To test this idea, subjects made movements from random starting locations to targets that had shape-dependent accuracy requirements. We used a robotic device to assess both limb impedance and patterns of movement variability just as the subject reached the target. We show that impedance increases in directions where required accuracy is high. Independent of target shape, patterns of limb stiffness are seen to predict spatial patterns of movement variability. The nervous system is thus seen to modulate limb impedance in entirely predictable environments to aid in the attainment of reaching accuracy.

The capacity for skill development over multiple training episodes is fundamental to human motor function. We have studied the process by which skills evolve with training by progressively modifying a series of motor learning tasks that subjects performed over a 1-mo period. In a series of empirical and modeling studies, we show that performance undergoes repeated modification with new learning. Each in a series of prior training episodes contributes such that present performance reflects a weighted average of previous learning. Moreover, we have observed that the relative weighting of skills learned wholly in the past changes with time. This suggests that the neural substrate of skill undergoes modification after consolidation.

Studies on plasticity in motor function have shown that motor learning generalizes, such that movements in novel situations are affected by previous training. It has been shown that the pattern of generalization for visuomotor rotation learning changes when training movements are made to a wide distribution of directions. Here we have found that for dynamics learning, the shape of the generalization gradient is not similarly modifiable by theextent of training within the workspace. Subjects learned to control a robotic device during training and we measured how subsequent movements in a reference direction were affected. Our results show that as the angular separation between training and test directions increased, the extent of generalization was reduced. When training involved multiple targets throughout the workspace, the extent of generalization was no greater than following training to the nearest target alone. Thus a wide range of experience compensating for a dynamics perturbation provided no greater benefit than localized training. Instead, generalization was complete when training involved targets that bounded the reference direction. This suggests that broad generalization of dynamics learning to movements in novel directions depends on interpolation between instances of localized learning.

Speech production is dependent on both auditory and somatosensory feedback. Although audition may appear to be the dominant sensory modality in speech production, somatosensory information plays a role that extends from brainstem responses to cortical control. Accordingly, the motor commands that underlie speech movements may have somatosensory as well as auditory goals. Here we provide evidence that, independent of the acoustics, somatosensory information is central to achieving the precision requirements of speech movements. We were able to dissociate auditory and somatosensory feedback by using a robotic device that altered the jaw's motion path, and hence proprioception, without affecting speech acoustics. The loads were designed to target either the consonant- or vowel-related portion of an utterance because these are the major sound categories in speech. We found that, even in the absence of any effect on the acoustics, with learning subjects corrected to an equal extent for both kinds of loads. This finding suggests that there are comparable somatosensory precision requirements for both kinds of speech sounds. We provide experimental evidence that the neural control of stiffness or impedance--the resistance to displacement--provides for somatosensory precision in speech production.

We used a robotic device to test the idea that impedance control involves a process of learning or adaptation that is acquired over time and permits the voluntary control of the pattern of stiffness at the hand. The tests were conducted in statics. Subjects were trained over the course of three successive days to resist the effects of one of three different kinds of mechanical loads, single axis loads acting in the lateral direction, single axis loads acting in the forward/backward direction and isotropic loads that perturbed the limb in eight directions about a circle. We found that subjects in contact with single axis loads voluntarily modified their hand stiffness orientation such that changes to the direction of maximum stiffness mirrored the direction of applied load. In the case of isotropic loads, a uniform increase in endpoint stiffness was observed. Using a physiologically realistic model of two-joint arm movement, the experimentally determined pattern of impedance change could be replicated by assuming that coactivation of elbow and double joint muscles was independent of coactivation of muscles at the shoulder. Moreover, using this pattern of coactivation control we were able to replicate an asymmetric pattern of rotation of the stiffness ellipse that was observed empirically. The present findings are consistent with the idea that arm stiffness is controlled through the use of at least two independent cocontraction commands.

Recent studies of human arm movement have suggested that the control of stiffness may be important both for maintaining stability and for achieving differences in movement accuracy. In the present study, we have examined the voluntary control of postural stiffness in 3D in the human jaw. The goal is to address the possible role of stiffness control in both stabilizing the jaw and in achieving the differential precision requirements of speech sounds. We previously showed that patterns of kinematic variability in speech are systematically related to the stiffness of the jaw. If the nervous system uses stiffness control as a means to regulate kinematic variation in speech, it should also be possible to show that subjects can voluntarily modify jaw stiffness. Using a robotic device, a series of force pulses was applied to the jaw to elicit changes in stiffness to resist displacement. Three orthogonal directions and three magnitudes of forces were tested. In all conditions, subjects increased the magnitude of jaw stiffness to resist the effects of the applied forces. Apart from the horizontal direction, greater increases in stiffness were observed when larger forces were applied. Moreover, subjects differentially increased jaw stiffness along a vertical axis to counteract disturbances in this direction. The observed changes in the magnitude of stiffness in different directions suggest an ability to control the pattern of stiffness of the jaw. The results are interpreted as evidence that jaw stiffness can be adjusted voluntarily, and thus may play a role in stabilizing the jaw and in controlling movement variation in the orofacial system.

Previous studies have used transfer of learning over workspace locations as a means to determine whether subjects code information about dynamics in extrinsic or intrinsic coordinates. Transfer has been observed when the torque associated with joint displacement is similar between workspace locations rather than when the mapping between hand displacement and force is preserved which is consistent with muscle- or joint based encoding. In the present study, we address the generality of an intrinsic coding of dynamics and examine how generalization occurs when the pattern of torques varies over the workspace. In two initial experiments, we examined transfer of learning when the direction of a force field was fixed relative to an external frame of reference. While there were no beneficial effects of transfer following training at a single location (Experiment 1 and 2), excellent performance was observed at the center of the workspace following training at two lateral locations (Experiment 2). Experiment 3 and associated simulations assessed the characteristics of this generalization. In these studies, we examined the patterns of transfer observed following adaptation to force fields that were composed of two subfields that acted in opposite directions. The experimental and simulated data are consistent with the idea that information about dynamics is encoded in intrinsic coordinates. The nervous system generalizes dynamics learning by interpolating between sets of control signals, each locally adapted to different patterns of torques.

Substantial neurophysiological evidence points to the posterior parietal cortex (PPC) as playing a key role in the coordinate transformation necessary for visually guided reaching. Our goal was to examine the role of PPC in the context of learning new dynamics of arm movements. We assessed this possibility by stimulating PPC with transcranial magnetic stimulation (TMS) while subjects learned to make reaching movements with their right hand in a velocity-dependent force field. We reasoned that, if PPC is necessary to adjust the trajectory of the arm as it interacts with a novel mechanical system, interfering with the functioning of PPC would impair adaptation. Single pulses of TMS were applied over the left PPC 40 msec after the onset of movement during adaptation. As a control, another group of subjects was stimulated over the visual cortex. During early stages of learning, the magnitude of the error (measured as the deviation of the hand paths) was similar across groups. By the end of the learning period, however, error magnitudes decreased to baseline levels for controls but remained significantly larger for the group stimulated over PPC. Our findings are consistent with a role of PPC in the adjustment of motor commands necessary for adapting to a novel mechanical environment.

We used a robotic device to test the idea that impedance control involves a process of learning or adaptation that is acquired over time and permits the voluntary control of the pattern of stiffness at the hand. The tests were conducted in statics. Subjects were trained over the course of three successive days to resist the effects of one of three different kinds of mechanical loads, single axis loads acting in the lateral direction, single axis loads acting in the forward/backward direction and isotropic loads that perturbed the limb in eight directions about a circle. We found that subjects in contact with single axis loads voluntarily modified their hand stiffness orientation such that changes to the direction of maximum stiffness mirrored the direction of applied load. In the case of isotropic loads, a uniform increase in endpoint stiffness was observed. Using a physiologically realistic model of two-joint arm movement, the experimentally determined pattern of impedance change could be replicated by assuming that coactivation of elbow and double joint muscles was independent of coactivation of muscles at the shoulder. Moreover, using this pattern of coactivation control we were able to replicate an asymmetric pattern of rotation of the stiffness ellipse that was observed empirically. The present findings are consistent with the idea that arm stiffness is controlled through the use of at least two independent cocontraction commands.

Recently, Shadmehr and colleagues (Criscimagna-Hemminger et al. 2003) reported a pattern of generalization of force-field adaptation between arms that differs from the pattern that occurs across different configurations of the same arm. While the intralimb pattern of generalization points to an intrinsic encoding of dynamics, the interlimb transfer described by these authors indicates that information about force is represented in a frame of reference external to the body. In the present study, subjects adapted to a viscous curl-field in two experimental conditions. In one condition, the field was introduced suddenly and produced clear deviations in hand paths; in the second condition, the field was introduced gradually so that at no point during the adaptation process could subjects observe or had to correct for a substantial kinematic error. In the first case, a pattern of interlimb transfer consistent with Criscimagna-Hemminger et al. was observed, whereas no transfer of learning between limbs occurred in the second condition. The findings suggest that there is limited transfer of fine compensatory force adjustment between limbs. Transfer when it does occur may be largely the results of a "cognitive" strategy that arises as a result of the sudden introduction of load and associated kinematic error.

The "ba, ba, ba" sound universal to babies' babbling around 7 months captures scientific attention because it provides insights into the mechanisms underlying language acquisition and vestiges of its evolutionary origins. Yet the prevailing mystery is what is the biological basis of babbling, with one hypothesis being that it is a non-linguistic motoric activity driven largely by the baby's emerging control over the mouth and jaw, and another being that it is a linguistic activity reflecting the babies' early sensitivity to specific phonetic-syllabic patterns. Two groups of hearing babies were studied over time (ages 6, 10, and 12 months), equal in all developmental respects except for the modality of language input (mouth versus hand): three hearing babies acquiring spoken language (group 1: "speech-exposed") and a rare group of three hearing babies acquiring sign language only, not speech (group 2: "sign-exposed"). Despite this latter group's exposure to sign, the motoric hypothesis would predict similar hand activity to that seen in speech-exposed hearing babies because language acquisition in sign-exposed babies does not involve the mouth. Using innovative quantitative Optotrak 3-D motion-tracking technology, applied here for the first time to study infant language acquisition, we obtained physical measurements similar to a speech spectrogram, but for the hands. Here we discovered that the specific rhythmic frequencies of the hands of the sign-exposed hearing babies differed depending on whether they were producing linguistic activity, which they produced at a low frequency of approximately 1 Hz, versus non-linguistic activity, which they produced at a higher frequency of approximately 2.5 Hz the identical class of hand activity that the speech-exposed hearing babies produced nearly exclusively. Surprisingly, without benefit of the mouth, hearing sign-exposed babies alone babbled systematically on their hands. We conclude that babbling is fundamentally a linguistic activity and explain why the differentiation between linguistic and non-linguistic hand activity in a single manual modality (one distinct from the human mouth) could only have resulted if all babies are born with a sensitivity to specific rhythmic patterns at the heart of human language and the capacity to use them.

The ability to formulate explicit mathematical models of motor systems has played a central role in recent progress in motor control research. As a result of these modeling efforts and in particular the incorporation of concepts drawn from control systems theory, ideas about motor control have changed substantially. There is growing emphasis on motor learning and particularly on predictive or anticipatory aspects of control that are related to the neural representation of dynamics. Two ideas have become increasingly prominent in mathematical modeling of motor function forward internal models and inverse dynamics. The notion of forward internal models which has drawn from work in adaptive control arises from the recognition that the nervous system takes account of dynamics in motion planning. Inverse dynamics, a complementary way of adjusting control signals to deal with dynamics, has proved a simple means to establish the joint torques necessary to produce desired movements. In this paper, we review the force control formulation in which inverse dynamics and forward internal models play a central role. We present evidence in its favor and describe its limitations. We note that inverse dynamics and forward models are potential solutions to general problems in motor control how the nervous system establishes a mapping between desired movements and associated control signals, and how control signals are adjusted in the context of motor learning, dynamics and loads. However, we find little empirical evidence that specifically supports the inverse dynamics or forward internal model proposals per se. We further conclude that the central idea of the force control hypothesis that control levels operate through the central specification of forces is flawed. This is specifically evident in the context of attempts to incorporate physiologically realistic muscle and reflex mechanisms into the force control model. In particular, the formulation offers no means to shift between postures without triggering resistance due to postural stabilizing mechanisms.

The hypothesis that speech goals are defined acoustically and maintained by auditory feedback is a central idea in speech production research. An alternative proposal is that speech production is organized in terms of control signals that subserve movements and associated vocal-tract configurations. Indeed, the capacity for intelligible speech by deaf speakers suggests that somatosensory inputs related to movement play a role in speech production-but studies that might have documented a somatosensory component have been equivocal. For example, mechanical perturbations that have altered somatosensory feedback have simultaneously altered acoustics. Hence, any adaptation observed under these conditions may have been a consequence of acoustic change. Here we show that somatosensory information on its own is fundamental to the achievement of speech movements. This demonstration involves a dissociation of somatosensory and auditory feedback during speech production. Over time, subjects correct for the effects of a complex mechanical load that alters jaw movements (and hence somatosensory feedback), but which has no measurable or perceptible effect on acoustic output. The findings indicate that the positions of speech articulators and associated somatosensory inputs constitute a goal of speech movements that is wholly separate from the sounds produced.

Humans produce speech by controlling a complex biomechanical apparatus to achieve desired speech sounds. We show here that kinematic variability in speech may be influenced by patterns of jaw stiffness. A robotic device was used to deliver mechanical perturbations to the jaw to quantify its stiffness in the mid-sagittal plane. Measured jaw stiffness was anisotropic. Stiffness was greatest along a protrusion-retraction axis and least in the direction of jaw raising and lowering. Consistent with the idea that speech movements reflect directional asymmetries in jaw stiffness, kinematic variability during speech production was found to be high in directions in which stiffness is low and vice versa. In addition, for higher jaw elevations, stiffness was greater and kinematic variability was less. The observed patterns of kinematic variability were not specific to speech similar patterns appeared in speech and nonspeech movements. The empirical patterns of stiffness were replicated by using a physiologically based model of the jaw. The simulation studies support the idea that the pattern of jaw stiffness is affected by musculo-skeletal geometry and muscle-force-generating abilities with jaw geometry being the primary determinant of the orientation of the stiffness ellipse.

Recent studies have demonstrated the ability of subjects to adjust the control of limb movements to counteract the effects of self-generated loads. The degree to which subjects change control signals to compensate for these loads is a reflection of the extent to which forces affecting movement are represented in motion planning. Here, we have used empirical and modeling studies to examine whether the nervous system compensates for loads acting on the jaw during speech production. As subjects walk, loads to the jaw vary with the direction and magnitude of head acceleration. We investigated the patterns of jaw motion resulting from these loads both in locomotion alone and when locomotion was combined with speech production. In locomotion alone, jaw movements were shown to vary systematically in direction and magnitude in relation to the acceleration of the head. In contrast, when locomotion was combined with speech, variation in jaw position during both consonant and vowel production was substantially reduced. Overall, we have demonstrated that the magnitude of load associated with head acceleration during locomotion is sufficient to produce a systematic change in the position of the jaw. The absence of variation in jaw position during locomotion with speech is thus consistent with the idea that in speech, the control of jaw motion is adjusted in a predictive manner to offset the effects of head acceleration.

Neuroimaging techniques may aid in the identification of areas of the human brain that are involved in tactile shape perception. Bodegard et al. (2001) relate differences in the properties of tactile stimuli to differences in areas of cortical activation to infer tactile processing in the somatosensory network.

A significant problem in motor control is how information about movement error is used to modify control signals to achieve desired performance. A potential source of movement error and one that is readily controllable experimentally relates to limb dynamics and associated movement-dependent loads. In this paper, we have used a position control model to examine changes to control signals for arm movements in the context of movement-dependent loads. In the model, based on the equilibrium-point hypothesis, equilibrium shifts are adjusted directly in proportion to the positional error between desired and actual movements. The model is used to simulate multi-joint movements in the presence of both "internal" loads due to joint interaction torques, and externally applied loads resulting from velocity-dependent force fields. In both cases it is shown that the model can achieve close correspondence to empirical data using a simple linear adaptation procedure. An important feature of the model is that it achieves compensation for loads during movement without the need for either coordinate transformations between positional error and associated corrective forces, or inverse dynamics calculations.

During multi-joint limb movements such as reaching, rotational forces arise at one joint due to the motions of limb segments about other joints. We report the results of three experiments in which we assessed the extent to which control signals to muscles are adjusted to counteract these ``interaction torques''. Human subjects performed single- and multi-joint pointing movements involving shoulder and elbow motion, and movement parameters related to the magnitude and direction of interaction torques were systematically manipulated. We examined electromyographic (EMG) activity of shoulder and elbow muscles, and specifically, the relationship between EMG activity and joint interaction torque. A first set of experiments examined single-joint movements. During both single-joint elbow (Experiment 1) and shoulder (Experiment 2) movements, phasic EMG activity was observed in muscles spanning the stationary joint (shoulder muscles in Experiment 1 and elbow muscles in Experiment 1). This muscle activity preceded movement, and varied in amplitude with the magnitude of upcoming interaction torque (the load resulting from motion of the non-stationary limb segment). In a third experiment, subjects performed multi-joint movements involving simultaneous motion at the shoulder and elbow. Movement amplitude and velocity at one joint were held constant, while the direction of movement about the other joint was varied. When the direction of elbow motion was varied (flexion vs extension) and shoulder kinematics were held constant, EMG activity in shoulder muscles varied depending on the direction of elbow motion (and hence the sign of the interaction torque arising at the shoulder). Similarly, EMG activity in elbow muscles varied depending on the direction of shoulder motion, for movements in which elbow kinematics were held constant. The results from all three experiments support the idea that central control signals to muscles are adjusted, in a predictive manner, to compensate for interaction torques loads arising at one joint which depend on motion about other joints.

External loads arising due to the orientation of body segments relative to gravity can affect the achievement of movement goals. The degree to which subjects adjust control signals to compensate for these loads is a reflection of the extent to which forces affecting motion are represented neurally. In the present study we assessed whether subjects, when speaking, compensate for loads due to the orientation of the head relative to gravity. We used a mathematical model of the jaw to predict the effects of control signals that are not adjusted for changes to head orientation. The simulations predicted a systematic change in sagittal plane jaw orientation and horizontal position resulting from changes to the orientation of the head. We conducted an empirical study in which subjects were tested under the same conditions. With one exception, empirical results were consistent with the simulations. In both simulation and empirical studies, the jaw was rotated closer to occlusion and translated in an anterior direction when the head was in the prone orientation. When the head was in the supine orienation, the jaw was rotated away from occlusion. The findings suggest that the nervous system does not completely compensate for changes in head orientation relative to gravity. A second study was conducted to assess possible changes in acoustical patterns due to changes in head orientation. The frequencies of the first (F1) and second (F2) formants associated with the steady-state portion of vowels were measured. As in the kinematic study, systematic differences in the values of F1 and F2 were observed with changes in head orientation. Thus the acoustical analysis further supports the conclusion that control signals are not completely adjusted to offset forces arising due to changes in orientation.

The aim of this study was to examine the possibility of independent muscle coactivation at the shoulder and elbow. Subjects performed rapid point-to-point movements in a horizontal plane, from different initial limb configurations to a single target. EMG activity was measured from flexor and extensor muscles that act at the shoulder (pectoralis clavicular head and posterior deltoid) and elbow (biceps long head and triceps lateral head) and flexor and extensor muscles that act at both joints (biceps short head and triceps long head). Muscle coactivation was assessed by measuring tonic levels of electromyographic (EMG) activity after limb position stabilized following movement end. It was observed that tonic EMG levels following movements to the same target varied as a function of the amplitude of shoulder and elbow motion. Moreover, for the movements tested here, the coactivation of shoulder and elbow muscles was found to be independent tonic EMG activity of shoulder muscles increased in proportion to shoulder movement, but was unrelated to elbow motion, whereas elbow and double-joint muscle coactivation varied with the amplitude of elbow movement, and were uncorrelated with shoulder motion. In addition, tonic EMG levels were higher for movements in which the shoulder and elbow rotated in the same direction, as compared to those in which the joints rotated in opposite directions. In this respect, muscle coactivation may reflect a simple strategy to compensate for forces introduced by multijoint limb dynamics.

The lambda model of the equilibrium-point hypothesis (Feldman & Levin, 1995) is an approach to motor control which, like physics, is based on a logical system coordinating empirical data. The model has gone through an interesting period. On one hand, several nontrivial predictions of the model have been successfully verified in recent studies. In addition, the explanatory and predictive capacity of the model has been enhanced by its extension to multimuscle and multijoint systems. On the other hand, claims have recently appeared suggesting that the model should be abandoned. The present paper focuses on these claims and concludes that they are unfounded. Much of the experimental data that have been used to reject the model are actually consistent with it.

It has been proposed that the control signals underlying voluntary human arm movement have a "complex" non-monotonic time-varying form, and a number of empirical findings have been offered in support of this idea (Gomi and Kawato, 1996, Latash and Gottlieb, 1991). In this paper we address three such findings using a model of two-joint arm motion based on the lambda version of the equilibrium-point hypothesis. The model includes six one- and two-joint muscles, reflexes, modeled control signals, muscle properties and limb dynamics. First, we address the claim that "complex" equilibrium trajectories are required in order to account for non-monotonic joint impedance patterns observed during multi-joint movement (Gomi and Kawato, 1996). Using constant-rate shifts in the neurally specified equilibrium of the limb, and constant cocontraction commands, we obtain patterns of predicted joint stiffness during simulated multi-joint movements which match the non-monotonic patterns reported empirically. We then use the algorithm proposed by Gomi and Kawato (1996) to compute a hypothetical equilibrium trajectory from simulated stiffness, viscosity and limb kinematics. Like that reported by Gomi and Kawato (1996), the resulting trajectory was non-monotonic, first leading then lagging the position of the limb. Second, we address the claim that high levels of stiffness are required to generate rapid single-joint movements when simple equilibrium shifts are used. We compare empirical measurements of stiffness during rapid single-joint movements (Bennett, 1993) with the predicted stiffness of movements generated using constant-rate equilibrium shifts and constant cocontraction commands. Single-joint movements are simulated at a number of speeds, and the procedure used by Bennett (1993) to estimate stiffness is followed. We show that when the magnitude of the cocontraction command is scaled in proportion to movement speed, simulated joint stiffness varies with movement speed in a manner comparable to that reported by Bennett (1993). Third, we address the related claim that non-monotonic equilibrium shifts are required to generate rapid single-joint movements. Using constant-rate equilibrium shifts and constant cocontraction commands, rapid single-joint movements are simulated in the presence of external torques. We use the procedure reported by Latash and Gottlieb (1991) to compute hypothetical equilibrium trajectories from simulated torque and angle measurements during movement. As in Latash and Gottlieb (1991), a non-monotonic function is obtained, even though the control signals used in the simulations are constant-rate changes in the equilibrium position of the limb. Differences between the "simple" equilibrium trajectory proposed in the present paper and those which are derived from the procedures used by Gomi and Kawato (1996) and Latash and Gottlieb (1991) arise from their use of simplified models of force-generation.

A model of the midsagittal plane motion of the tongue, jaw, hyoid bone, and larynx is presented, based on the lambda version of equilibrium point hypothesis. The model includes muscle properties and realistic geometrical arrangement of muscles, modeled neural inputs and reflexes, and dynamics of soft tissue and bony structures. The focus is on the organization of control signals underlying vocal tract motions and on the dynamic behavior of articulators. A number of muscle synergies or "basic motions" of the system are identified. In particular, it is shown that systematic sources of variation in an x-ray data base of midsagittal vocal tract motions can be accounted for, at the muscle level, with six independent commands, each corresponding to a direction of articulator motion. There are two commands for the jaw, (corresponding to sagittal plane jaw rotation and jaw protrusion), one command controlling larynx height, and three commands for the tongue, (corresponding to forward and backward motion of the tongue body, arching and flattening of the tongue dorsum, and motion of the tongue tip). It is suggested that all movements of the system can be approximated as linear combinations of such basic motions. In other words, individual movements and sequences of movements can be accounted for by a simple additive control model. The dynamics of individual commands are also assessed. It is shown that the dynamic effects are not neglectable in speechlike movements because of the different dynamic behaviors of soft and bony structures.

With the development of precise three dimensional motion measurement systems and powerful computers for three dimensional graphical visualization, it is possible to record and fully reconstruct human jaw motion. In this paper, we describe a visualization system for displaying three dimensional jaw movements in speech. The system is designed to take as input jaw motion data obtained from one or multi-dimensional recording systems. In the present application, kinematic records of jaw motion were recorded using an optoelectronic measurement system (Optotrak). The corresponding speech signal was recorded using an analog input channel. The three orientation angles and three positions which describe the motion of the jaw as a rigid skeletal structure were derived from the empirical measurements. These six kinematic variables, which in mechanical terms account fully for jaw motion kinematics, act as inputs that drive a real-time three dimensional animation of a skeletal jaw and upper skull. The visualization software enables the user to view jaw motion from any orientation and to change the viewpoint during the course of an utterance. Selected portions of an utterance may be re-played and the speed of the visual display may be varied. The user may also display, along with the audio track, individual kinematic degrees of freedom or several degrees of freedom in combination. The system is presently being used as an educational tool and for research into audio-visual speech recognition.

The kinematics of human jaw movements were assessed in terms of the three orientation angles and three positions that characterize the motion of the jaw as a rigid body. The analysis focused on the identification of the jaw's independent movement dimensions, and was based on an examination of jaw motion paths that were plotted in various combinations of linear and angular coordinate frames. Overall, both behaviors were characterized by independent motion in four degrees of freedom. In general, when jaw movements were plotted to show orientation in the sagittal plane as a function of horizontal position, relatively straight paths were observed. In speech, the slopes and intercepts of these paths varied depending on the phonetic material. The vertical position of the jaw was observed to shift up or down so as to displace the overall form of the sagittal plane motion path of the jaw. Yaw movements were small but independent of pitch, vertical and horizontal position. In mastication, the slope and intercept of the relationship between pitch and horizontal position were affected by the type of food and its size. However, the range of variation was less than that observed in speech. When vertical jaw position was plotted as a function of horizontal position, the basic form of the path of the jaw was maintained but could be shifted vertically. In general, larger bolus diameters were associated with lower jaw positions throughout the movement. The timing of pitch and yaw motion differed. The most common pattern involved changes in pitch angle during jaw opening followed by a phase predominated by lateral motion (yaw). Thus, in both behaviors there was evidence of independent motion in pitch, yaw, horizontal position and vertical position. This is consistent with the idea that motions in these degrees of freedom are independently controlled.

We investigated phasic and tonic stretch reflexes in human jaw opener muscles, which have few, if any, muscle spindles. Jaw unloading reflexes were recorded for both opener and closer muscles. Surface electromyographic (EMG) activity was obtained from left and right digastric and superficial masseter muscles, and jaw orientation and torques were recorded. Unloading of jaw opener muscles elicited a short-latency decrease in EMG activity (averaging 20 ms) followed by a short duration silent period in these muscles and sometimes a short burst of activity in their antagonists. Similar behavior in response to unloading was observed for spindle-rich jaw closer muscles although the latency of the silent period was statistically shorter than that observed for jaw opener muscles (averaging 13 ms). Control studies suggest that the jaw opener reflex was not due to inputs from either cutaneous or periodontal mechanoreceptors. In the unloading response of the jaw openers, the tonic level of EMG activity observed after transition to the new jaw orientation was monotonically related to the residual torque and orientation. This is consistent with the idea that the tonic stretch reflex may mediate the change in muscle activation. In addition, the values of the static net joint torque and jaw orientation after the dynamic phase of unloading were related by a monotonic function resembling the invariant characteristic recorded in human limb joints. The torque-angle characteristics associated with different initial jaw orientations were similar in shape but spatially shifted, consistent with the idea that voluntary changes in jaw orientation may be associated with a change in a single parameter, which may be identified as the threshold of the tonic stretch reflex. It is suggested that functionally significant phasic and tonic stretch reflexes may not be mediated exclusively by muscle spindle afferents. Thus, the hypothesis that central modifications in the threshold of the tonic stretch reflex underlie the control of movement may be applied to the jaw system.

1. When subjects trace patterns such as ellipses, the instantaneous velocity of movements is related to the instantaneous curvature of the trajectories according to a power law movements tend to slow down when curvature is high and speed up when curvature is low. It has been proposed that this relationship is centrally planned. 2. The arm's muscle properties and dynamics can significantly affect kinematics. Even under isometric conditions, muscle mechanical properties can affect the development of muscle forces and torques. Without a model which accounts for these effects, it is difficult to distinguish between kinematic patterns which are attributable to central control and patterns which arise due to dynamics and muscle properties and are not represented in the underlying control signals. 3. In this paper we address the nature of the control signals that underlie movements which obey the power law. We use a numerical simulation of arm movement control based on the lambda version of the equilibrium-point hypothesis. We demonstrate that simulated elliptical and circular movements, and elliptical force trajectories generated under isometric conditions, obey the power law even though there was no relation between curvature and speed in the modeled control signals. 4. We suggest that limb dynamics and muscle mechanics specifically, the spring-like properties of muscles can contribute significantly to the emergence of the power law relationship in kinematics. Thus without a model that accounts for these effects, care must be taken when making inferences about the nature of neural control.

The vocal tract's motion during speech is a complex patterning of the movement of many different articulators according to many different time functions. Understanding this myriad of gestures is important to a number of different disciplines including automatic speech recognition, speech and language pathologies, speech motor control, and experimental phonetics. Central issues are the accurate description of the shape of the vocal tract and determining how each articulator contributes to this shape. A problem facing all of these research areas is how to cope with the multivariate data from speech production experiments. In this paper techniques are described that provide useful tools for describing multivariate functional data such as the measurement of speech movements. The choice of data analysis procedures has been motivated by the need to partition the articulator movement in various ways: end effects separated from shape effects, partitioning of syllable effects, and the splitting of variation within an articulator site from variation from between sites. The techniques of functional data analysis seem admirably suited to the analyses of phenomena such as these. Familiar multivariate procedures such as analysis of variance and principal components analysis have their functional counterparts, and these reveal in a way more suited to the data the important sources of variation in lip motion. Finally, it is found that the analyses of acceleration were especially helpful in suggesting possible control mechanisms. The focus is on using these speech production data to understand the basic principles of coordination. However, it is believed that the tools will have a more general use.

A model is presented of sagittal plane jaw and hyoid motion based on the lambda model of motor control. The model which is implemented as a computer simulation includes central neural control signals, position and velocity dependent reflexes, reflex delays, and muscle properties such as the dependence of force on muscle length and velocity. The model has seven muscles (or muscle groups) attached to the jaw and hyoid as well as separate jaw and hyoid bone dynamics. According to the model, movements result from changes in neurophysiological control variables which shift the equilibrium state of the motor system. One such control variable is an independent change in the membrane potential of alpha-motoneurones (MNs); this variable establishes a threshold muscle length (lambda) at which MN recruitment begins. Motor functions may be specified by various combinations of lambdas. One combination of lambdas is associated with the level of coactivation of muscles. Others are associated with motions in specific degrees of freedom. Using the model, we study the mapping between control variables specified at the level of degrees of freedom and control variables corresponding to individual muscles. We demonstrate that commands can be defined involving linear combinations of lambda change which produce essentially independent movements in each of the four kinematic degrees of freedom represented in the model (jaw orientation, jaw position, vertical and horizontal hyoid position). These linear combinations are represented by vectors in lambda space which may be scaled in magnitude. The vector directions are constant over the jaw / hyoid workspace and result in essentially the same motion from any workspace position. The demonstration that it is not necessary to adjust control signals to produce the same movements in different parts of the workspace supports the idea that the nervous system need not take explicit account of musculo-skeletal geometry in planning movements.

Coarticulation in speech production is a phenomenon in which the articulator movements for a given speech sound vary systematically with the surrounding sounds and their associated movements. Although these variations may appear to be centrally planned, without explicit models of the speech articulators, the kinematic patterns which are attributable to central control cannot be distinguished from those which arise due to dynamics and are not represented in the underlying control signals. In the present paper, we address the origins of coarticulation by comparing the results of empirical and modeling studies of jaw motion in speech. The simulated kinematics of sagittal plane jaw rotation and horizontal jaw translation are compared to the results of empirical studies in which subjects produce speech-like sequences at a normal rate and volume. The simulations examine both "anticipatory" and "carryover" coarticulatory effects. In both cases, the results show that even when no account is taken of context at the level of central control, kinematic patterns vary in amplitude and duration as a function of the magnitude of the preceding or following movement in the same manner as one observes empirically in coarticulation. Since at least some coarticulatory effects may arise from muscle mechanics and dynamics and not from central control, these factors must be considered before drawing inferences about control in coarticulation.

To explore the extent to which functional systems within the human posterior parietal cortex and the superior temporal sulcus are involved in the perception of action, we measured cerebral metabolic activity in human subjects by positron emission tomography during the perception of simulations of biological motion with point-light displays. The experimental design involved comparisons of activity during the perception of goal-directed hand action, whole body motion, object motion, and random motion. The results demonstrated that the perception of scripts of goal-directed hand action implicates the cortex in the intraparietal sulcus and the caudal part of the superior temporal sulcus, both in the left hemisphere. By contrast, the rostrocaudal part of the right superior temporal sulcus and adjacent temporal cortex, and limbic structures such as the amygdala, are involved in the perception of signs conveyed by expressive body movements.

In this paper, we address a number of issues in speech research in the context of the equilibrium point hypothesis of motor control. The hypothesis suggests that movements arise from shifts in the equilibrium position of the limb or the speech articulator. The equilibrium is a consequence of the interaction of central neural commands, reflex mechanisms, muscle properties and external loads, but it is under the control of central neural commands. These commands act to shift the equilibrium via centrally specified signals acting at the level of the motoneurone (MN) pool. In the context of a model of sagittal plane jaw and hyoid motion based on the lambda version of the equilibrium point hypothesis, we consider the implications of this hypothesis for the notion of articulatory targets. We suggest that simple linear control signals may underlie smooth articulatory trajectories. We explore as well the phenomenon of intra-articulator coarticulation in jaw movement. We suggest that even when no account is taken of upcoming context, that apparent anticipatory changes in movement amplitude and duration may arise due to dynamics. We also present a number of simulations that show in different ways how variability in measured kinematics can arise in spite of constant magnitude speech control signals.

The present study quantifies electromyographic (EMG) magnitude, timing, and duration in one and two degree of freedom elbow movements involving combinations of flexion / extension and pronation / supination. The aim is to understand the organization of commands subserving motion in individual and multiple degrees of freedom. The muscles tested in this study fell into two categories with respect to agonist burst magnitude those whose burst magnitude varied with motion in a second degree of freedom at the elbow, and those whose burst magnitude depended on motion in one degree of freedom only. In multiarticular muscles contributing to motion in two degrees of freedom at the elbow, we found that the magnitude of the agonist burst was greatest for movements in which a muscle acted as agonist in both degrees of freedom. The burst magnitudes for one degree of freedom movements were, in turn, greater than for movements in which the muscle was agonist in one degree of freedom and antagonist in the other. It was also found that for movements in which a muscle acted as agonist in two degrees of freedom, the burst magnitude was, in the majority of cases, not different than the sum of the burst magnitudes in the component movements. When differences occured, the burst magnitude for the combined movement was greater than the sum of the components. Other measures of EMG activity such as burst onset time and duration were not found to vary in a systematic manner with motion in these two degrees of freedom. It was also seen that several muscles which produced motion in one degree of freedom at the elbow, including triceps brachii (long head), triceps brachii (lateral head), and pronator quadratus displayed first agonist bursts whose magnitude did not vary with motion in a second degree of freedom. However, for the monoarticular elbow flexors brachialis and brachioradialis, agonist burst magnitude was affected by pronation or supination. Lastly, it was observed that during elbow movements in which muscles acted as agonist in one degree of freedom and antagonist in the other, the muscle activity often displayed both agonist and antagonist components in the same movement. It was found that, for pronator teres and biceps brachii, the timing of the bursts was such that there was activity in these muscles concurrent with activity in both pure agonists and pure antagonists. The empirical summation of EMG burst magnitudes and the presence in a single muscle of both agonist and antagonist bursts within a movement suggest that central commands associated with motion in individual degrees of freedom at the elbow may be superimposed to produce two degree of freedom elbow movements.

The human jaw moves in three spatial dimensions, and its motion is filly specified by three orientation angles and three positions. Using OPTOTRAK, we characterize the basic motions in these six degrees of freedom and their interrelations during speech. As has been reported previously, the principle components of jaw motion fall primarily within the midsagittal plane, where the jaw rotates downward and translates forward during opening movements and follows a similar path during closing. In general, the relation between sagittal plane rotation and horizontal translation (protrusion) is linear. However, speakers display phoneme-specific differences in the slope of this relation and its position within the rotation-translation space. Furthermore, instances of pure rotation and pure translation are observed. These findings provide direct support for the claim that jaw rotation and translation are independently controlled (Flanagan, Ostry & Feldman, 1990). Rotations out of the midsagittal plane are also observed. Yaw about the longitudinal body axis is approximately three degrees and roll usually less than two degrees. The remaining non-sagittal component, lateral translation, is small in magnitude and uncorrelated with other motions.

1. The kinematics of sagittal-plane jaw motion were assessed in mastication and speech. The movement paths were described in joint coordinates, in terms of the component rotations and translations. The analysis focused on the relationship between rotation and horizontal translation. Evidence was presented that these can be separately controlled.
2. In speech, jaw movements were studied during consonant-vowel utterances produced at different rates and volumes. In mastication, bolus placement, compliance and size as well as chewing rate were manipulated. Jaw movements were recorded using the University of Wisconsin X-ray microbeam system. Jaw rotation and translation were calculated on the basis of the motion of X-ray tracking pellets on the jaw.
3. The average magnitudes of jaw rotation and translation were greater in mastication than in speech. In addition, in speech, it was shown that the average rotation magnitude may vary independent of the horizontal translation magnitude. In mastication, the average magnitude of vertical jaw translation was not dependent on the magnitudes of jaw rotation or horizonal jaw translation.
4. The magnitude of rotation and horizontal jaw translation tended to be correlated when examined on a trial by trial basis. Some subjects also showed a correlation between jaw rotation and vertical jaw translation. However, the proportion of variance accounted for was greater for all subjects in the case of rotation and horizontal translation.
5. Joint space paths in both mastication and speech were found to be straight. The pattern was observed at normal and fast rates of speech and mastication and for loud speech as well. Straight line paths were also observed when subjects produced utterances that had both the syllabic structure and the intonation pattern of speech. The findings suggest that control may be organized in terms of an equilibrium jaw orientation and an equilibrium jaw position.
6. Departures from linearity were also observed. These were typically associated with differences during jaw closing in the end time of rotation and translation. Asynchronies were not observed at the start of jaw closing in either mastication or speech and the movement paths were typically linear within this region.

We investigated the coordination of mono- and bi-articular muscles during movements involving one or more degrees of freedom at the elbow. Subjects performed elbow flexion (or extension) alone, forearm pronation (or supination) alone, and combinations of the two. In bi-articular muscles such as biceps and pronator teres, the amplitude of agonist EMG activity was dependent on motion in both degrees of freedom. Agonist burst amplitudes for combined movements were approximately the sum of the agonist burst amplitudes for movements in the individual degrees of freedom. Activity levels in individual degrees of freedom were, in turn, greater than activity levels observed when a muscle acted as agonist in one degree of freedom and antagonist in the other. Other muscles such as triceps, brachialis, and pronator quadratus, acted primarily during motion in a single degree of freedom. The relative magnitude and the timing of activity between sets of muscles also changed with motion in a second degree of freedom. These patterns are comparable to those reported previously in isometric studies.

The vocal tract's motion during speech is a complex patterning of the movement of many different articulators according to many different time functions. Understanding this myriad of gestures is important to a number of different disciplines including automatic speech recognition, speech and language pathologies, speech motor control, and experimental phonetics. Central issues are the accurate description of the shape of the vocal tract and determining how each articulator contributes to this shape. A problem facing all of these research areas is how to cope with the multivariate data from speech production experiments. In this paper techniques are described that provide useful tools for describing multivariate functional data such as the measurement of speech movements. The choice of data analysis procedures has been motivated by the need to partition the articulator movement in various ways: end effects separated from shape effects, partitioning of syllable effects, and the splitting of variation within an articulator site from variation from between sites. The techniques of functional data analysis seem admirably suited to the analyses of phenomena such as these. Familiar multivariate procedures such as analysis of variance and principal components analysis have their functional counterparts, and these reveal in a way more suited to the data the important sources of variation in lip motion. Finally, it is found that the analyses of acceleration were especially helpful in suggesting possible control mechanisms. The focus is on using these speech production data to understand the basic principles of coordination. However, it is believed that the tools will have a more general use.

The three-dimensional kinematics of the hindlimb back-wipe were examined in spinal frogs. The component movements were identified and the relationship between stimulus position and hindlimb configuration was assessed. The planes of motion of the hindlimb were examined throughout the movement. The back-wipe comprises three essential phases: a placing phase (I), in which the foot is drawn over the back of the frog and placed in a position near to the stimulus; a pre-whisk phase (II), in which the endpoint of the foot moves away from the stimulus; and a whisk/extension phase (III), in which the stimulus is removed. The pre-whisk phase contributes to force production for the whisk/extension (III). In the placing phase a systematic relationship was found between limb endpoint position and stimulus position in the rostro-caudal direction. The hip, knee and metatarsal joint angles were related to the position of the endpoint in the rostro-caudal direction. However, different frogs tended to adopt different strategies to remove the stimulus. In one strategy, when the knee angle was strongly related to the rostro-caudal stimulus position, the metatarsal angle was weakly related and vice versa. Other strategies were observed as well. There was no adjustment in limb endpoint position for stimulus placement in the medial-lateral direction. Consistent with this finding, the point on the foot at which stimulus contact occurred changed systematically as a function of medial-lateral stimulus placement. Thus, in order to remove the stimulus in different medial-lateral positions, the frog used a different part of the foot rather than moving the foot in the direction of the stimulus. In two frogs a relationship was observed between the elevation of the femur and the medial-lateral stimulus position. The motion planes of the hindlimb were studied by examining the instantaneous plane of motion of the endpoint and the planes of motion of adjacent limb segments. The motion of the endpoint was found not to be planar in any phase of the wipe. In contrast, planar motion of the femur and tibia was observed for all phases. Systematic changes in the orientation of these planes characterized the different phases. The position of the hindlimb was found to be variable prior to the placing phase. This variability was not related to stimulus position. However, in trials with multiple wipes, once an initial limb configuration was assumed, the limb returned to this configuration before each wipe in the sequence. Evidence for motor equivalence was sought in two ways.

The determinants of the motion path of the hindlimb were explored in both intact and spinal frogs. In the spinal preparations the kinematic properties of withdrawal and crossed-extension reflexes were studied. In the intact frog the kinematics of withdrawal and swimming movements were examined. Frog hindlimb paths were described in joint angle (intrinsic) coordinates rather than limb endpoint (extrinsic) coordinates. 2. To study withdrawal and crossed-extension reflexes, the initial angles at the hip, knee, and ankle were varied. Withdrawal and crossed extension were recorded in three dimension (3-D) with the use of an infra-red spatial imaging system. Swimming movements against currents of different speeds were obtained with high-speed film. 3. Three strategies were considered related to the form of the hypothesized equilibrium paths specified by the nervous system: all trajectories lie on a single line in angular coordinates; all trajectories are directed toward a common final position; and all trajectories have the same direction independent of initial joint configuration. 4. Joint space paths in withdrawal were found to be straight and parallel independent of the initial joint configuration. The hip and knee were found to start simultaneously and in 75% of the conditions tested to reach maximum velocity simultaneously. Hip-knee maximum velocity ratios were similar in magnitude over differences in initial joint angles. This is consistent with the observation of parallel paths and supports the view that the nervous system specifies a single direction for equilibrium trajectories. 5. Straight line paths with slopes similar to those observed in withdrawal in the spinal preparation were found in swimming movements in the intact frog. Straight line paths in joint space are consistent with the idea that swimming and withdrawal are organized and controlled in a joint-level coordinate system. The similarities observed between spinal and intact preparations suggest that a common set of constructive elements underlies these behaviors. 6. Path curvature was introduced when joint limits were approached toward the end of the movement. Depending on the initial joint angles, the joint movements ended at different times. When initial joint angles were unequal, joints moving from smaller initial angles reached their functional limits earlier and stopped first. 7. In withdrawal and crossed extension in the spinal frog, velocity profiles at a given joint were similar over the initial portion of the curve for movements of different amplitude. This is consistent with the idea that withdrawal and crossed-extension movements of different amplitude are produced by a constant rate of shift of the equilibrium position.

The study of jaw movement in humans is a primary source of information about the relationship between voluntary movement and more primitive motor functions. This study focused on the geometric form of the velocity function, as measured by linear voltage displacement transducer. Movement amplitudes, maximum velocities and durations were greater in mastication than in speech. Nevertheless, there were detailed similarities in the shape of the normalized velocity functions. In jaw-closing movements, the normalized functions were similar in form over differences in rate, movement amplitude (speech movements) and the compliance of the bolus (mastication). In opening movements, the functions for mastication and speech were again similar over differences in amplitude and compliance. However, they differed in shape for fast and slow movements. Normalized acceleration and deceleration durations were approximately equal in rapid movements, whereas, for slower movements, deceleration took substantially

The velocity curves of human arm and speech movements were examined as a function of amplitude and rate in both continuous and discrete movement tasks. Evidence for invariance under scalar transformation was assessed and a quantitative measure of the form of the curve was used to provide information on the implicit cost function in the production of voluntary movement. Arm, tongue and jaw movements were studied separately. The velocity curves of tongue and jaw movement were found to differ in form as a function of movement duration but were similar for movements of different amplitude. In contrast, the velocity curves for elbow movements were similar in form over differences in both amplitude and duration. Thus, the curves of arm movement, but not those of tongue or jaw movement, were geometrically equivalent in form. Measurements of the ratio of maximum to average velocity in arm movement were compared with the theoretical values calculated for a number of criterion functions. For continuous movements, the data corresponded best to values computed for the minimum energy criterion; for discrete movement, values were in the range of those predicted for the minimum jerk and best stiffness criteria. The source of a rate dependent asymmetry in the form of the velocity curve of speech movements was assessed in a control study in which subjects produced simple raising and lowering movements of the jaw without talking. The velocity curves of the non-speech control gesture were similar in form to those of jaw movement in speech. These data, in combination with similar findings for human jaw movement in mastication, suggest that the asymmetry is not a direct consequence of the requirements of the task. The biomechanics and neural control of the orofacial system may be possible sources of this effect.

Medial movements of the lateral pharyngeal wall at the level of the velopharyngeal port were examined by using a computerized ultrasound system. Subjects produced CVNVC sequences involving all combinations of the vowels /a/ and /u/ and the nasal consonants /n/ and /m/. The effects of both vowels on the CVN and NVC gestures (opening and closing of the velopharyngeal port, respectively) were assessed in terms of movement amplitude, duration, and movement onset time. The amplitude of both opening and closing gestures of the lateral pharyngeal wall was less in the context of the vowel /u/ than the vowel /a/. In addition, the onset of the opening gesture towards the nasal consonant was related to the identity of both the initial and the final vowels. The characteristics of the functional coupling of the velum and lateral pharyngeal wall in speech are discussed.

The control of individual speech gestures was investigated by examining laryngeal and tongue movements during vowel and consonant production. A number of linguistic manipulations known to alter the durational characteristics of speech (i.e., speech rate, lexical stress, and phonemic identity) were tested. In all cases a consistent pattern was observed in the kinematics of the laryngeal and tongue gestures. The ratio of maximum instantaneous velocity to movement amplitude, a kinematic index of mass-normalized stiffness, was found to increase systematically as movement duration decreased. Specifically, the ratio of maximum velocity to movement amplitude varied as a function of a parameter, C, times the reciprocal of movement duration. The conformity of the data to this relation indicates that durational change is accomplished by scalar adjustment of a base velocity form. These findings are consistent with the idea that kinematic change is produced by the specification of articulator stiffness.

A computerized pulsed-ultrasound system was used to monitor tongue dorsum movements during the production of consonant-vowel sequences in which speech rate, vowel, and consonant were varied. The kinematics of tongue movement were analyzed by measuring the lowering gesture of the tongue to give estimates of movement amplitude, duration, and maximum velocity. All three subjects in the study showed reliable correlations between the amplitude of the tongue dorsum movement and its maximum velocity. Further, the ratio of the maximum velocity to the extent of the gesture, a kinematic indicator of articulator stiffness, was found to vary inversely with the duration of the movement. This relationship held both within individual conditions and across all conditions in the study such that a single function was able to accommodate a large proportion of the variance due to changes in movement duration. As similar findings have been obtained both for abduction and adduction gestures of the vocal folds and for rapid voluntary limb movements, the data suggest that a wide range of changes in the duration of individual movements might all have a similar origin. The control of movement rate and duration through the specification of biomechanical characteristics of speech articulators is discussed.

Intra-articulator anticipatory and carryover coarticulation were assessed in both temporal and spatial terms. Three subjects produced VCV sequences with velar stop consonants and back vowels. Pulsed ultrasound was used to examine the vertical displacement, duration, and maximum velocity of the tongue dorsum raising (VC transition) and lowering (CV transition) gestures. Anticipatory coarticulation was primarily temporal for two subjects, with decreases in the duration of the VC transition accompanying increases in displacement for the CV transition. Carryover coarticulation was primarily spatial for all three subjects, with decreases in CV displacement and maximum velocity accompanying increases in VC displacement. It is suggested that these intra-articulator patterns can be accounted for in terms of an interaction between the raising gesture and a vowel-specific onset time of the lowering gesture towards the vowel. The implications of this kinematic characterization are discussed.

The kinematics of tongue dorsum movements in speech were studied with pulsed ultrasound to assess similarities in the voluntary control of the speech articulators and the limbs. The stimuli were consonant--vowel syllables in which speech rate and stress were varied. The kinematic patterns for tongue dorsum movements were comparable to those observed in the rapid movement of the arms and hands. The maximum velocity of tongue dorsum raising and lowering was correlated with the extent of the gesture. The slope of the relationship differed for stressed and unstressed vowels but was unaffected by differences in speech rate. At each stress level the correlation between displacement and peak velocity was accompanied by a relatively constant interval from the initiation of the movement to the point of maximum velocity. The data are discussed with reference to systems that can be described with second-order differential equations. The increase in the slope of the displacement/peak-velocity relationship for unstressed versus stressed vowels is suggestive of a tonic increase in articulator stiffness. Variations in displacement are attributed to the level of phasic activity in the muscles producing the gesture.

A computerized system for the measurement of tongue dorsum movements with pulsed echo ultrasound is described. The presentation focuses on technical and methodological considerations in the on-line acquisition of vertical tongue movement information, its digital processing and display. Problems associated with transducer placement, peak detection, data averaging, and curve fitting are considered, and validation procedures based on x ray and indicators of measurement reliability are reported. The discussion centers on advantages and disadvantages of the technique and its applications.

Typewriting, musical instrument playing, spoken language, and dance involve sophisticated motor skills and associated symbol schemes. Researchers in cognition have been interested in these abilities because they enable the study of relations between the structure of motor behavior and the organization of the associated formal system. Typewriting, in particular, is of interest because of the remarkable rate and complexity of finger and hand movements involved and because its performance is readily quantifiable. However, if typewriting is to be used to study either cognitive or motor organization, the factors contributing to its temporal structure must be identified. In this chapter I present the findings of several studies that examine variables that influence the pattern of interkey times in typing. In addition to providing evidence on the constituents of control in typing, the studies provide a basis for the examination of proposals by Ostry (1980) and Sternberg, Monsell, Knoll, and Wright (1978) that certain aspects of typing control are inherently tied to the execution of the sequence. The suggestions arise from observations that patterns of initial latency and interkey time are not changed by the introduction of a delay between stimulus presentation and a response signal. The inability to take advantage of a preparation interval seems to indicate that the programming of typing movements is intimately linked to their execution.