Motor learning can be monitored by observing the development of neural correlates of error processing. Among these neural correlates, the error- and feedback-related negativity (Ne/ERN and FRN) represent error processing mechanisms. While the Ne/ERN is more related to error prediction, the FRN is found after an error is manifested. The questions the current study strives to answer are: What information is needed by the system to make error predictions and how is this represented by the Ne/ERN and FRN in a complex motor task? We reduced the information and increased the difficulty level for the prediction in a semivirtual throwing task and found no Ne/ERN but a large FRN when the action result was finally observed (hitting or missing a target). We assume that uncertainty for error prediction was too high (either due to insufficient information or due to lacking prerequisites for prediction), such that error processing had to be mainly based on feedback. The finding is in line with the reinforcement theory of learning, after which Ne/ERN and FRN should behave complementary.
Michael Joch, Mathias Hegele, Heiko Maurer, Hermann Müller and Lisa K. Maurer
Jürgen Konczak, Kai Brommann and Karl Theodor Kalveram
Knowledge of how stiffness, damping, and the equilibrium position of specific limbs change during voluntary motion is important for understanding basic strategies of neuromotor control. Presented here is an algorithm for identifying time-dependent changes in joint stiffness, damping, and equilibrium position of the human forearm. The procedure requires data from only a single trial. The method relies neither on an analysis of the resonant frequency of the arm nor on the presence of an external bias force. Its validity was tested with a simulated forward model of the human forearm. Using the parameter estimations as forward model input, the angular kinematics (model output) were reconstructed and compared to the empirically measured data. Identification of mechanical impedance is based on a least-squares solution of the model equation. As a regularization technique and to improve the temporal resolution of the identification process, a moving temporal window with a variable width was imposed. The method's performance was tested by (a) identifying a priori known hypothetical time-series of stiffness, damping, and equilibrium position, and (b) determining impedance parameters from recorded single-joint forearm movements during a hold and a goal-directed movement task. The method reliably reconstructed the original angular kinematics of the artificial and human data with an average positional error of less than 0.05 rad for movement amplitudes of up to 0.9 rad, and did not yield hypermetric trajectories like previous procedures not accounting for damping.
Silvia C. Lipski, Stefanie Unger, Martine Grice and Ingo G. Meister
Adult speakers have developed precise forward models of articulation for their native language and seem to rely less on auditory sensory feedback. However, for learning of the production of new speech sounds, auditory perception provides a corrective signal for motor control. We assessed adult German speakers’ speech motor learning capacity in the absence of auditory feedback but with clear somatosensory information. Learners were presented with a nonnative singleton-geminate duration contrast of voiceless, unaspirated bilabial plosives /p/ vs. /pp/ which is present in Italian. We found that the lack of auditory feedback had no immediate effect but that deviating productions emerged during the course of learning. By the end of training, speakers with masked feedback produced strong lengthening of segments and showed more variation on their production than speakers with normal auditory feedback. Our findings indicate that auditory feedback is necessary for the learning of precise coordination of articulation even if somatosensory feedback is salient.
Stacey L. Gorniak, Marcos Duarte and Mark L. Latash
We explored possible effects of negative covariation among finger forces in multifinger accurate force production tasks on the classical Fitts’s speed-accuracy trade-off. Healthy subjects performed cyclic force changes between pairs of targets “as quickly and accurately as possible.” Tasks with two force amplitudes and six ratios of force amplitude to target size were performed by each of the four fingers of the right hand and four finger combinations. There was a close to linear relation between movement time and the log-transformed ratio of target amplitude to target size across all finger combinations. There was a close to linear relation between standard deviation of force amplitude and movement time. There were no differences between the performance of either of the two “radial” fingers (index and middle) and the multifinger tasks. The “ulnar” fingers (little and ring) showed higher indices of variability and longer movement times as compared with both “radial” fingers and multifinger combinations. We conclude that potential effects of the negative covariation and also of the task-sharing across a set of fingers are counterbalanced by an increase in individual finger force variability in multifinger tasks as compared with single-finger tasks. The results speak in favor of a feed-forward model of multifinger synergies. They corroborate a hypothesis that multifinger synergies are created not to improve overall accuracy, but to allow the system larger flexibility, for example to deal with unexpected perturbations and concomitant tasks.
Geoffrey T. Burns, Kenneth M. Kozloff and Ronald F. Zernicke
, 1990 ), throwing ( Morriss, Bartlett, & Fowler, 1997 ), and dancing ( Murgia, 1995 ). Each approach has advantages and drawbacks. The forward modeling of movements allows us to control variables and manipulate systems, revealing precise relations to the cost and performance criteria. This facilitates
Cornelia Frank, Taeho Kim and Thomas Schack
observational practice alone. The authors concluded that observational practice did either not lead to an update of internal models related to the task, or to differential updating of internal models. Along these lines, the authors discuss the possibility that different forward models evolve during different
Adilson Santos Andrade de Sousa, Marilia A. Correia, Breno Quintella Farah, Glauco Saes, Antônio Eduardo Zerati, Pedro Puech-Leao, Nelson Wolosker, Gabriel G. Cucato and Raphael M. Ritti-Dias
relationship between barriers to physical activity and sex was analyzed by binary logistic regression. Male and negative answers to the barriers were taken as references. The stepwise backward–forward modeling was performed, and two models were presented. The first model had age, ankle–brachial index, BMI
Cheryl M. Glazebrook
include both forward models that allow the nervous system to predict sensory feedback and inverse models in which the motor commands needed to achieve a goal are formulated ( Wolpert et al., 2003 ). As inverse models are developed, we are more accurately able to plan our movements, whereas as forward
Sara M. Scharoun, David A. Gonzalez, Eric A. Roy and Pamela J. Bryden
behavioral sciences . New York : Routledge . Contreras-Vidal , J.L. ( 2006 ). Development of forward models for hand localization and movement control in 6- to 10-year-old children . Human Movement Science, 25 ( 4 ), 634 – 645 . PubMed doi:10.1016/j.humov.2006.07.006 10.1016/j.humov.2006
Ing-Shiou Hwang, Chia-Ling Hu, Wei-Min Huang, Yi-Ying Tsai and Yi-Ching Chen
.M. , & Miall , R.C. ( 1996 ). Forward models for physiological motor control . Neural Networks, 9 ( 8 ), 1265 – 1279 . PubMed ID: 12662535 doi: 10.1016/S0893-6080(96)00035-4 Woods , D.L. , Wyma , J.M. , Yund , E.W. , Herron , T.J. , & Reed , B. ( 2015 ). Age-related slowing of response