Role of action observation and action in sequence learning and coding

A complex sequence learning task was used to determine if the type of coding acquired through physical practice (PP), observation of the stimulus (Obs-S), or observation of stimulus and action (Obs-SA) differs between conditions and whether the type of observation influences subsequent learning of the task when physical practice was permitted. Participants in the Obs-S group were permitted to watch the sequentially illuminated stimuli on the screen. In the Obs-SA group participants could see both flexion-extension movements of the model's arm performing the sequence and the sequentially illuminated stimuli on the screen. Participants in the PP group actually performed the 16-element sequence with their dominant right arm. Delayed retention tests and two inter-manual transfer tests were completed following each of two acquisition sessions. First, the data indicated that learning the sequence structure, as revealed by response time per element, occurred similarly irrespective of the initial practice condition. Secondly, the movement sequence appeared to be coded in abstract visual-spatial coordinates resulting in effector-independent performance. Finally, observing the model's action and sequential stimuli allows participants to transfer the perceived aspects of the movement sequence into efficient coordination patterns when additional physical practice is permitted.     

Transfer of movement sequences: bigger is better

Experiment 1 was conducted to determine if proportional transfer from "small to large" scale movements is as effective as transferring from "large to small." We hypothesize that the learning of larger scale movement will require the participant to learn to manage the generation, storage, and dissipation of forces better than when practicing smaller scale movements. Thus, we predict an advantage for transfer of larger scale movements to smaller scale movements relative to transfer from smaller to larger scale movements. Experiment 2 was conducted to determine if adding a load to a smaller scale movement would enhance later transfer to a larger scale movement sequence. It was hypothesized that the added load would require the participants to consider the dynamics of the movement to a greater extent than without the load. The results replicated earlier findings of effective transfer from large to small movements, but consistent with our hypothesis, transfer was less effective from small to large (Experiment 1). However, when a load was added during acquisition transfer from small to large was enhanced even though the load was removed during the transfer test. These results are consistent with the notion that the transfer asymmetry noted in Experiment 1 was due to factors related to movement dynamics that were enhanced during practice of the larger scale movement sequence, but not during the practice of the smaller scale movement sequence. The findings that the movement structure is unaffected by transfer direction but the movement dynamics are influenced by transfer direction is consistent with hierarchal models of sequence production.     

Eye movements during transitive action observation have sequential structure

In this paper, we explore how observers of a reach-to-grasp action can identify and distinguish between the agent and patient (i.e. target) of the action. We investigate the hypothesis that there is a characteristic sequential structure to the observer’s pattern of saccades, with the agent being fixated first, and then the target. We report an experiment which indicates that this sequence of saccades, while not ubiquitous, is overwhelmingly more likely than chance. The experiment also sheds some light on the mechanisms which allow the observer to saccade from the agent of the action to the target.     

Offline improvement occurs for temporal stability but not accuracy following practice of integer and non-integer rhythms

Procedural learning benefits from memory processes occurring outside practice resulting in offline learning. Offline gains have been demonstrated almost exclusively for the ordinal structure of sequential motor tasks. Many skills also demand that the correct serial order of events be appropriately timed. Evidence indicates that the temporal aspect of a procedural skill can be encoded independent of serial order knowledge and governed by at least two distinct neural circuits. The present experiment determined if (a) offline gains emerge for temporal learning, and (b) if such gains occur for timing supervised by distinct timing systems. Participants experienced 216 practice trials of a 7-key press sequence that involved integer- or non-integer timing rhythms. Twenty-four hours after training 30 test trials were administered. Results revealed robust offline enhancement for timing performance of the non-integer based temporal sequences. This improvement was localized to stabilization of the required relative but not absolute time profiles. The neural circuitry central to supporting the performance of non-integer timing sequences is also a principal constituent of what is described as the "cognitive" timing system. Timing governed by this system appears most susceptible to offline gains via consolidation.     

Coding of on-line and pre-planned movement sequences

Recent experiments have demonstrated that complex multi-element movement sequences were coded in visual-spatial coordinates even after extensive practice, while relatively simple spatial-temporal movement sequences are coded in motor coordinates after a single practice session. The purpose of the present experiment was to determine if the control process rather than the difficulty of the sequence played a role in determining the pattern of effector transfer. To accomplish this, different concurrent feedback conditions were provided to two groups of participants during practice of the same movement sequence. The results indicated that when concurrent visual feedback was provided during the production of the movement, which was thought to encourage on-line control, the participants performed transfer tests with the contra-lateral limb better when the visual-spatial coordinates were reinstated than when the motor coordinates were reinstated. When concurrent visual feedback was not provided, which was thought to encourage pre-planned control, the opposite was observed. The data are consistent with the hypothesis that the mode of control dictates the coordinate system used to code the movement sequence rather than sequence difficulty or stage of practice as has been proposed.     

Representation of movement sequences is related to task characteristics

Recent experiments have produced mixed results in terms of performance when, after learning a sequential task, the same visual-spatial coordinates or the same motor coordinates were reinstated on a subsequent effector transfer test. Given the diversity of tasks and especially sequence characteristics used in previous experiments, the cross-experimental comparison makes inferences and unambiguous interpretations difficult. The purpose of the present experiment was to determine in a principled manner how the spatio-temporal structure of a sequence influences the way the sequence is represented. The results indicated that after limited amount of practice relatively more simple sequences (S1) are coded more efficiently in a mirror (motor) representation which requires the same pattern of homologous muscle activation. Conversely, relatively more complex sequences (S2) are more efficiently coded in a visual-spatial coordinate system which requires movements to the same spatial locations as during acquisition. The data are also consistent with the notion that sequences with different spatio-temporal structures rely to a different degree on distinct control mechanisms (pre-planned vs. on-line, respectively).     

Retrieval practice in motor learning

In this study we sought to determine whether testing promotes the generalization of motor skills during the process of encoding and/or consolidation. We used a dynamic arm movement task that required participants to reproduce a spatial-temporal pattern of elbow extensions and flexions with their dominant right arm. Generalization of motor learning was tested by the ability to transfer the original pattern (extrinsic transformation) or the mirrored pattern (intrinsic transformation) to the unpractised left arm. To investigate the testing effects during both encoding and consolidation processing, participants were administered an initial testing session during early practice before being evaluated on a post-practice testing session administered either 10 min (Testing-Encoding group) or 24 hr apart (Testing-Consolidation group), respectively. Control groups were required to perform a post-practice testing session administered after either a 10-min (Control-Encoding group) or 24-hr delay (Control-Consolidation group). The findings revealed that testing produced rapid, within-practice skill improvements, yielding better effector transfer at the 10-min testing for the Testing-Encoding group on both extrinsic and intrinsic transformation tests when compared with the Control-Encoding group. Furthermore, we found better performance for the Testing-Consolidation group at the 24-hr testing for extrinsic and intrinsic transformations of the movement pattern when compared with the Control-Consolidation group. However, our results did not indicate any significant testing advantage on the latent, between-session development of the motor skill representation (i.e., from the 10-min to the 24-hr testing). The testing benefits expressed at the 10-min testing were stabilised but did not extend during the period of consolidation. This indicates that testing contributes to the generalisation of motor skills during encoding but not consolidation.     

Interference within hands: Retrieval-induced forgetting of left and right hand movements

We examined retrieval-induced forgetting of motor sequences that were categorized by the effectors (left or right hand) involved in their execution. This left–right categorization was independent from input locations or input devices. In addition, the acquired motor sequences were arbitrarily assigned to left and right. Participants learned twelve sequential joystick movements as responses to letter stimuli. Half of the sequences pertained to the left, half to the right hand. Subsequent retrieval-practice of half the items of one hand induced forgetting for the non-retrieved rest of the items of that hand in a final recall test. This finding demonstrates that the hands were used to organize the memory storage of motor sequences in a way that gave rise to later interference between commonly stored items, that is, linked to the same hand.     

Cognitive underpinnings of contextual interference during motor learning

The reported study examined the cognitive processes underlying contextual interference (CI) in motor learning. This experiment was designed to assess the combined influence of practice schedule (blocked or random) and task similarity (similar or dissimilar) on acquisition and retention performance. Participants (N = 60) learned a set of three variations of a timing task according to a similar (900, 1000 and 1100 ms) or dissimilar parameter condition (700, 1000 and 1300 ms) with either a blocked or random practice order: this resulted in 4 experimental groups. Performance in delayed retention demonstrated a typical CI effect due to the schedule of practice for the dissimilar parameter condition with the random practice group outperforming the blocked practice group. Conversely, no blocked-random difference was found for the similar parameter condition. These findings lend support for the reconstruction hypothesis by showing that supplementing random practice with additional intertask elaboration (i.e., similar parameter condition) did not facilitate subsequent retention performance.     

Covariation of attentional cost and stability provides further evidence for two routes to learning new coordination patterns

This study investigated how learning a new bimanual coordination pattern affects the attentional resources allotted by the CNS to maintain it throughout the acquisition process. The repertoire of the existing stable coordination patterns was individually evaluated before and after practice in order to detect expected changes with learning. Bistable participants, who initially exhibited stable and accurate coordination patterns at 0° and 180° of relative phase, practiced a 90° pattern, whereas multistable participants, who already mastered the 90° pattern, practiced 135° pattern instead. In a typical dual-task paradigm, all participants had to simultaneously perform a reaction time task that assessed the associated attentional cost. Beyond an overall increase in accuracy, the results revealed a significant decrease in the attentional cost for bistable participants, accompanying the stabilization of the 90° pattern with learning, but not for multistable participants, as the 135° pattern barely stabilized. Pattern stability and attentional cost co-evolve during learning and the process follows two different routes depending on the interplay between the task and the learner’s coordination abilities before practice.     

Increased dynamic regulation of postural tone through Alexander Technique training

Gurfinkel and colleagues (2006) recently found that healthy adults dynamically modulate postural muscle tone in the body axis during anti-gravity postural maintenance and that this modulation is inversely correlated with axial stiffness. Our objective in the present study was to investigate whether dynamic modulation of axial postural tone can change through training. We examined whether teachers of the Alexander Technique (AT), who undergo “long-term” (3-year) training, have greater modulation of axial postural tone than matched control subjects. In addition, we performed a longitudinal study on the effect of “short-term” (10-week) AT training on the axial postural tone of individuals with low back pain (LBP), since short term AT training has previously been shown to reduce LBP. Axial postural tone was quantified by measuring the resistance of the neck, trunk and hips to small (±10°), slow (1°/s) torsional rotation during stance. Modulation of tone was determined by the torsional resistance to rotation (peak-to-peak, phase-advance, and variability of torque) and axial muscle activity (EMG). Peak-to-peak torque was lower (∼50%), while phase-advance and cycle-to-cycle variability were enhanced for AT teachers compared to matched control subjects at all levels of the axis. In addition, LBP subjects decreased trunk and hip stiffness following short-term AT training compared to a control intervention. While changes in static levels of postural tone may have contributed to the reduced stiffness observed with the AT, our results suggest that dynamic modulation of postural tone can be enhanced through long-term training in the AT, which may constitute an important direction for therapeutic intervention.     

Learning new movement patterns: a study on good and poor writers comparing learning conditions emphasizing spatial, timing or abstract characteristics

In the earliest stages of motor-skill learning cognitive, visuo-spatial and dynamic processes play an important role. Which of these should be addressed first when children need to learn a new complex movement sequence? This study compares three learning methods in a within-subject design by having 18 good and 18 poor 8-year-old writers master unfamiliar, letter-like patterns by (1) tracing a trajectory on a screen, (2) tracking a moving target (pursuit), and (3) performing the pattern using written explicit instructions. Following each 10-trial learning phase, the children completed a short test phase. Besides errors and kinematic data, Dynamic Time Warping (DTW) was used to calculate the deviation for each pattern from the ideal shape (DTW-distance). As predicted, the number of errors and DTW-distance were very low during the learning phase of the tracing and pursuit conditions and higher in the explicit condition. Conversely, in the test phase, tracing yielded the highest DTW-distance and the explicit condition the lowest DTW-distance and error percentages. The results were remarkably similar for the good and poor writers. The poor learning results of the tracing condition and the good results of the explicit condition have important implications for the teaching of handwriting and remedial therapy.     

The influence of writing practice on letter recognition in preschool children: a comparison between handwriting and typing

A large body of data supports the view that movement plays a crucial role in letter representation and suggests that handwriting contributes to the visual recognition of letters. If so, changing the motor conditions while children are learning to write by using a method based on typing instead of handwriting should affect their subsequent letter recognition performances. In order to test this hypothesis, we trained two groups of 38 children (aged 3-5 years) to copy letters of the alphabet either by hand or by typing them. After three weeks of learning, we ran two recognition tests, one week apart, to compare the letter recognition performances of the two groups. The results showed that in the older children, the handwriting training gave rise to a better letter recognition than the typing training.     

Implicit transfer of spatial structure in visuomotor sequence learning

Implicit learning and transfer in sequence learning are essential in daily life. Here, we investigated the implicit transfer of visuomotor sequences following a spatial transformation. In the two experiments, participants used trial and error to learn a sequence consisting of several button presses, known as the m × n task (Hikosaka et al., 1995). After this learning session, participants learned another sequence in which the button configuration was spatially transformed in one of the following ways: mirrored, rotated, and random arrangement. Our results showed that even when participants were unaware of the transformation rules, accuracy of transfer session in the mirrored and rotated groups was higher than that in the random group (i.e., implicit transfer occurred). Both those who noticed the transformation rules and those who did not (i.e., explicit and implicit transfer instances, respectively) showed faster performance in the mirrored sequences than in the rotated sequences. Taken together, the present results suggest that people can use their implicit visuomotor knowledge to spatially transform sequences and that implicit transfers are modulated by a transformation cost, similar to that in explicit transfer.     

Model for a flexible motor memory based on a self-active recurrent neural network

Using recent recurrent network architecture based on the reservoir computing approach, we propose and numerically simulate a model that is focused on the aspects of a flexible motor memory for the storage of elementary movement patterns into the synaptic weights of a neural network, so that the patterns can be retrieved at any time by simple static commands. The resulting motor memory is flexible in that it is capable to continuously modulate the stored patterns. The modulation consists in an approximately linear inter- and extrapolation, generating a large space of possible movements that have not been learned before. A recurrent network of thousand neurons is trained in a manner that corresponds to a realistic exercising scenario, with experimentally measured muscular activations and with kinetic data representing proprioceptive feedback. The network is “self-active” in that it maintains recurrent flow of activation even in the absence of input, a feature that resembles the “resting-state activity” found in the human and animal brain. The model involves the concept of “neural outsourcing” which amounts to the permanent shifting of computational load from higher to lower-level neural structures, which might help to explain why humans are able to execute learned skills in a fluent and flexible manner without the need for attention to the details of the movement.     

Consistently modeling the same movement strategy is more important than model skill level in observational learning contexts

The experiment undertaken was designed to elucidate the impact of model skill level on observational learning processes. The task was bimanual circle tracing with a 90° relative phase lead of one hand over the other hand. Observer groups watched videos of either an instruction model, a discovery model, or a skilled model. The instruction and skilled model always performed the task with the same movement strategy, the right-arm traced clockwise and the left-arm counterclockwise around circle templates with the right-arm leading. The discovery model used several movement strategies (tracing-direction/hand-lead) during practice. Observation of the instruction and skilled model provided a significant benefit compared to the discovery model when performing the 90° relative phase pattern in a post-observation test. The observers of the discovery model had significant room for improvement and benefited from post-observation practice of the 90° pattern. The benefit of a model is found in the consistency with which that model uses the same movement strategy, and not within the skill level of the model. It is the consistency in strategy modeled that allows observers to develop an abstract perceptual representation of the task that can be implemented into a coordinated action. Theoretically, the results show that movement strategy information (relative motion direction, hand lead) and relative phase information can be detected through visual perception processes and be successfully mapped to outgoing motor commands within an observational learning context.     

Proprioception plays a different role for sensorimotor adaptation to different distortions

If proprioceptive feedback is degraded by agonist-antagonist muscle vibration, then adaptation to rotated vision remains intact while adaptation to a velocity-dependent force field worsens. Here we evaluate whether this differential effect of vibration is related to the physical nature of the distortion - visual versus mechanical - or to their kinematic coupling to the subjects’ hand - velocity versus position dependent. Subjects adapted to a velocity-dependent visual distortion, to a position-dependent force, or to a velocity-dependent force; one half of the subjects adapted with, and the other half without agonist-antagonist vibration at the wrist, elbow, and shoulder. We found, as before, that vibration slowed down adaptation to a velocity-dependent force. However, vibration did not modify adaptation to the other two distortions, nor did it influence the aftereffects of any distortion. From this we conclude that intact proprioception supports strategic compensatory processes when proprioceptive signals agree with visual ones, and provide relevant (dynamic) information not available to the visual system.     

The role of eye movements in motor sequence learning

An experiment that utilized a 16-element movement sequence was designed to determine the impact of eye movements on sequence learning. The participants were randomly assigned to two experimental groups: a group that was permitted to use eye movements (FREE) and a second group (FIX) that was instructed to fixate on a marker during acquisition (ACQ). A retention test (RET) was designed to provide a measure of learning, and two transfer tests were designed to determine the extent to which eye movements influenced sequence learning. The results demonstrated that both groups decreased the response time to produce the sequence, but the participants in the FREE group performed the sequence more quickly than participants of the FIX group during the ACQ, RET and the two transfer tests. Furthermore, continuous visual control of response execution was reduced over the course of learning. The results of the transfer tests indicated that oculomotor information regarding the sequence can be stored in memory and enhances response production.     

Eye movements are functional during face learning

In a free viewing learning condition, participants were allowed to move their eyes naturally as they learned a set of new faces. In a restricted viewing learning condition, participants remained fixated in a single central location as they learned the new faces. Recognition of the learned faces was then tested following the two learning conditions. Eye movements were recorded during the free viewing learning condition, as well as during recognition. The recognition results showed a clear deficit following the restricted viewing condition, compared with the free viewing condition, demonstrating that eye movements play a functional role during human face learning. Furthermore, the features selected for fixation during recognition were similar following free viewing and restricted viewing learning, suggesting that the eye movements generated during recognition are not simply a recapitulation of those produced during learning.     

Transversus abdominis is part of a global not local muscle synergy during arm movement

The trunk muscle transversus abdominis (TrA) is thought to be controlled independently of the global trunk muscles. Methodological issues in the 1990s research such as unilateral electromyography and a limited range of arm movements justify a re-examination of this theory. The hypothesis tested is that TrA bilateral co-contraction is a typical muscle synergy during arm movement. The activity of 6 pairs of trunk and lower limb muscles was recorded using bilateral electromyography during anticipatory postural adjustments (APAs) associated with the arm movements. The integrated APA electromyographical signals were analyzed for muscle synergy using Principle Component Analysis. TrA does not typically bilaterally co-contract during arm movements (1 out of 6 participants did). APA muscle activity of all muscles during asymmetrical arm movements typically reflected a direction specific diagonal pattern incorporating a twisting motion to transfer energy from the ground up. This finding is not consistent with the hypothesis that TrA plays a unique role providing bilateral, feedforward, multidirectional stiffening of the spine. This has significant implications to the theories underlying the role of TrA in back pain and in the training of isolated bilateral co-contraction of TrA in the prophylaxis of back pain.