The study of interceptive actions has been at the heart of movement science for more than half a century, dating back to the seminal work of Whiting on perception and action in ball catching (Sharp & Whiting, 1974; Whiting, 1968, 1969; Whiting et al., 1970; Whiting & Sharp, 1974). Interceptive actions are indeed paradigmatic examples of our behavioral interaction with dynamic elements of the environment. However, notwithstanding the wealth of work on the prime examples of catching and hitting, to date, there is still no operational model available capable of realistically capturing how we manage even the basic requirement of any interceptive action of getting to the right place at the right time. This basic requirement is arguably most purely incorporated in the task of lateral (i.e., direction-constrained) interception, in which the agent’s action is restricted to movement along a fixed interception axis, which is the focus of the present contribution.
The first attempt at modeling lateral manual interception behavior was grounded in the idea that the hand movement would be guided by a (perceived and continuously updated) required velocity, defined as the ratio of the current lateral distance between hand and target over the time remaining until the target reaches the interception axis (Peper et al., 1994; also see Dessing et al., 2002). In this model, the hand velocity is continuously driven toward the currently required velocity, with the latter being modulated by a faster-than-linear activation function (cf. Bullock & Grossberg, 1988). Inclusion of such an activation function was necessary to ensure a smooth initial rise in hand velocity and sufficiently rapid integration of differences between current and required hand velocity in subsequent phases. Although this first-order dynamics type of model allowed for capturing—at least the qualitative aspects of—kinematic phenomena observed in interception behavior (as reported in Montagne et al., 1999; Peper et al., 1994), its initial structure and development over time (Dessing et al., 2004, 2005, 2009) have revealed several limitations. First, inclusion of an activation function is in fact problematic, given this element’s predominant role in the shaping of the resulting movement. Second, model extensions have given rise to an inflation of hypothesized system components, thereby dissolving the parsimony of the initial idea.
For the present purposes, we therefore build on two theoretical contributions addressing pattern formation dynamics for discrete movement. Schöner (1990) proposed that discrete movements could be generated by a dynamical system with two stable point attractors, located at the initial and to-be-reached positions. In his model, the intention to move, instantiated as “behavioral information” (BI), is expressed as an additional component in the equation of motion, allowing for driving the system from the fixed-point regime into a limit cycle regime when BI is turned on. Turning BI off, about halfway through the cycle, brings the system back into the stable fixed-point regime, leading it to relax toward the final point. The Jirsa and Kelso (2005) excitator model also uses BI, in this case to push a bistable attractor point system over a local flow-segregating separatrix into the attraction basin of the final point attractor. It is important to notice that in both proposals this BI only serves to push the system into a desired regime or state; it is not prescriptive of the particular aspects of the movement to be made (Schöner, 1990). In the following, we prefer to refer this contribution to the pattern dynamics as “behavioral seed” (BS) to reserve the term information for perceptual quantities.
In developing his initial proposal in the field of robotics, Schöner subsequently integrated task requirements in the form of (perceived) constraints on behavior, allowing for scaling the movement’s amplitude and duration (Oubbati & Schöner, 2013; Schöner, 1994a, 1994b; Schöner et al., 1995; Schöner & Santos, 2001). By including spatial and temporal constraints into the model, it became capable of generating movements arriving at a perceptually defined place after a perceptually-defined time, as in discrete interception movements. We note, however, that the above-cited contributions, using generic (normal form) state equations, were developed to demonstrate the principles of a nonlinear dynamics account for time and space constrained discrete movements but did not attempt to capture empirically observed end-effector kinematics. As our goal is precisely to model such empirically observed end-effector kinematics, we drew inspiration from this theoretical work but took the somewhat different route of starting from a second-order dynamic of the general form provided by Equation 1 already explored in earlier data-driven modeling studies of rhythmic behavior (e.g., Beek et al., 1996; Beek, Schmidt et al., 1995; Kay et al., 1987, 1991; Mottet & Bootsma, 1999). We note that, using such an approach, Zaal et al. (1999) modeled discrete reaching toward targets that were either stationary or moving (away from the initial hand position) along the reaching movement axis. They thus included a particular type of lateral manual interception, requiring the hand to catch up with the moving target. Following Schöner’s (1990) suggestion of a (partial) limit cycle implication, Zaal et al.’s (1999) Duffing–Rayleigh limit cycle model adequately captured experimentally observed relations among movement amplitude, movement time, and peak velocity. This study, however, did not address the transition from limit cycle to the (subsequent) fixed-point regime and therefore remains incomplete. It is also worth highlighting that, as the to-be-reached target moved along the hand movement axis, the task did not directly constrain either where or when the target was to be intercepted, as is the case for the type of data that we want to model.
Given that to date there is no complete model capable of generating realistic lateral manual interception movements, we proceed by first identifying what we want our model to be able to capture, as this defines its basic characteristics. To independently vary where and when the target is to be intercepted, the task must have target trajectories that cross the interception axis. Studies using such target trajectories have revealed a variety of influences of target trajectory characteristics on the kinematics of the interceptive actions produced (e.g., Arzamarski et al., 2007; Dessing et al., 2005; Ledouit et al., 2013; Michaels et al., 2006; Montagne et al., 1999, 2000; Peper et al., 1994), including effects of distance to be covered, time pressure, and target trajectory orientation with respect to the interception axis. In most of the above cited empirical studies, targets could cross the interception axis at different distances on the left and on the right of the initial hand position. Our model thus needs to be able to generate both leftward and rightward movements.
Data Set Used for Modeling
For the present modeling purposes, we therefore needed a data set of lateral manual interceptive movements collected for participants being confronted with targets that could reach the interception axis at different locations, on the left and on the right side of the initial end-effector position, after different target motion durations. The data collected in the study of lateral manual interception reported by Ledouit et al. (2013) fitted these requirements. Briefly, in this study, participants moved a handheld stylus (represented on screen by a 0.1-cm wide white line cursor) along a horizontal interception axis situated near the bottom of a digitizing tablet’s screen to intercept virtual targets (hereafter referred to as balls, represented by a 0.8-cm diameter white circle,) moving from top to bottom across the black-background screen toward the interception axis at one of two constant speeds (20 or 32 cm/s), leading to ball motion durations of 1.6 and 1.0 s. From Ledouit et al.’s (2013) data set, we retained the orthogonal ball trajectories (no sideward ball motion component) that crossed the interception axis at lateral distances of −14, −7, +7, and +14 cm from the initial stylus position. Figure 1 presents the ensemble-average kinematic profiles of the recorded lateral interception movements, which the present contribution aims to model.
Ensemble averages of empirically observed hand trajectories from Ledouit et al.’s (2013) data set for interception of uniformly moving balls following trajectories perpendicular to the interception axis, arriving at distances of −14, −7, +7, and +14 cm from the hand starting position after ball flight times of 1.6 and 1.0 s. (A) Position as a function of time, (B) velocity as a function of time, and (C) phase portraits (velocity as a function of position). (Color figure online).
Citation: Motor Control 2025; 10.1123/mc.2024-0036
As can be seen from Figure 1, the distance to be covered and time available both affected the movement kinematics, notably in terms of the timing and magnitude of peak velocity reached and the overall shape of the velocity–time curve.
Model Structure
The model developed for generating these different movements was built in the following way. With balls arriving both left and right of the starting position, we modified Schöner’s (1990) and Jirsa and Kelso’s (2005) lead of using a Duffing function to create stable equilibrium points located at the starting position and at the goal position, with BS pushing the hand to move the system into the attraction basin of the goal position. We note that, in both of these approaches, the origin of the reference frame used is located halfway between the stable equilibrium points that define the starting and goal positions. Movement from the starting position is thus typically unilateral.
To circumvent this, we parametrized the Duffing function to switch from a monostable regime with the equilibrium point located at the starting position (here coinciding with the origin of the reference frame) to a bistable stable regime with equilibrium points located at the required distance to the left and right of the starting position. As a result of this switch, the starting position becomes an unstable equilibrium point. A (relatively small) signed BS contribution is used to push the system in the required direction, to the left or to the right. We note that this left–right feature of our model is also needed if we want to include reversal movements (interception on the right characterized by an initial leftward movement component, or vice versa, as reported by Montagne et al., 1999, under particular target trajectory conditions) in future model applications.
Exemplary vector field representation of the phase space of the bistable attractor point regime of the model. The origin (0, 0) is a repellor, whereas in this example (−14, 0) and (+14, 0) are attractors. Pushing the system rightward from its (0, 0) initial location leads it to follow the stabilized trajectory to (+14, 0); pushing the system leftward from its (0, 0) initial location leads it to follow the stabilized trajectory to (−14, 0). Note that the areas of interest are the first and third quadrants.
Citation: Motor Control 2025; 10.1123/mc.2024-0036
To dampen the system during movement and allow for the variations in the shape of the velocity profile observed in the empirical data, we included both nonconservative Rayleigh and Van der Pol terms. Such hybrid Rayleigh and Van der Pol damping is quite common in dynamical modeling of upper limb movements (Kay et al., 1987; Mottet & Bootsma, 1999; Schmidt et al., 2015).
Phase portraits of (rightward) movements generated by the dynamical system captured by Equation 3, demonstrating the influences of Rayleigh and Van der Pol damping terms on the kinematic shapes. For all simulations, coefficient a was set to 0.5 and xg to 14, while coefficients of the Rayleigh term (d) and the Van der Pol term (e) were independently varied. The top dashed gray curve corresponds to d = e = 0.1. Dotted gray to red variations reflect effects of increasing e (0.1, 0.2, 0.4, 0.6, 0.8, and 1.0) for d = 0.1. Gray to orange variations reflect effects of increasing d (0.1, 0.2, 0.4, 0.6, 0.8, and 1.0) for e = 0.1. Arrows track changes in the location of peak velocity. (Color figure online).
Citation: Motor Control 2025; 10.1123/mc.2024-0036
Scaling the Model to the Data
We recall that our data set contains interception movements with different amplitudes (ball arrival positions at 7 and 14 cm, to the left and to the right of the starting position) and different temporal constraints (balls reaching the interception location in 1.6 and 1.0 s). The model must therefore be scaled to these constraints. Following Peper et al. (1994), we assume that, at the moment of onset of the movement, information is available specifying both future ball arrival position on the interception axis (with the goal location, i.e., specified ball arrival position denoted as xb) and time until the ball reaches this position (with specified ball flight time denoted as Tf).
Scaling the model to the data is therefore to be done by modulation of the coefficients of model terms by xb and Tf. The magnitude of the Duffing coefficient a influences the duration of the movement as larger a’s are associated with larger initial accelerations and hence shorter movement durations. The perceived time constraint Tf may thus be incorporated by modulating coefficient a by Tf. Given the condition-dependent kinematic features of the observed interceptive movements (Figure 1), further modulating influences of space–time constraints xb and Tf on model parameters are also expected.
Fitting the Model to the Data
For the present purposes, we assumed that information about the future ball arrival position on the interception axis and ball flight time became available after 150 ms of ball motion across the screen, as empirical hand acceleration tended to rapidly increase thereafter. For each ball motion condition, we therefore switched the initial goal location xg = 0 to the perceived ball arrival location xg = xb and simultaneously turned BS on at that time. BS was subsequently turned off 300 ms later. With BS thus being activated at 150 ms after trial onset and deactivated 300 ms later, BS is in fact a function of time, hereafter denoted as BS(t).
In fitting the model to the data, we concurrently explored the presence of systematic modulatory influences of movement amplitude
Fitting was performed using a nonlinear gradient-based optimization method (fmincon MATLAB function). To prevent the algorithm from getting stuck in local position error minima, we performed multiple fits, exploring the effects of varying initial settings and lower and upper boundaries of the different parameters.
As expected, time available to reach the interception location (Tf) exerted an inverse proportional modulatory influence on the parameter of the Duffing term, such that shorter ball flight times gave rise to higher initial accelerations. A similar (inverse proportional) influence on the Duffing term was observed for the required amplitude of movement (
The optimization procedure revealed that Rayleigh and Van der Pol damping terms indeed played complementary roles. With larger movement distances requiring less damping, the Van der Pol term was inverse proportionally modulated by
Finally, optimization indicated that the BS(t), providing a temporary push in the required direction of magnitude xb, was inverse proportionally modulated by the time constraint Tf. Thus, BS(t) in fact became proportional to xb/Tf, which corresponds to the (perceptually specified) initial required velocity, as defined by Peper et al. (1994).
Figure 4 presents the trajectories generated by this parametrized model for each of the four ball arrival positions under each of the two ball flight durations.
Final model (Equation 5) simulation results for interceptive movements of balls following trajectories perpendicular to the interception axis, arriving at distances of −14, −7, +7, and +14 cm for the hand starting position after ball flight times of 1.6 and 1.0 s, with coefficient settings a′ = 15.3223, d′ = 0.0164, e′ = 5.3504, and f′ = –1.2601. (A) Position as a function of time, (B) velocity as a function of time, and (C) phase portraits (velocity as a function of position). (Color figure online).
Citation: Motor Control 2025; 10.1123/mc.2024-0036
Discussion
As can be seen from Figure 4, parametrized with the required movement direction (sign of xb in BS), movement amplitude
Of course, the model did not perfectly reproduce the empirically observed movements. A first reason for this is that the observed leftward and rightward interception movements were not fully symmetrical, as can be seen from Figure 1. Quantitatively, this symmetrical discrepancy amounted to an RMSD of 0.80 cm. We note that this amounts to some 30% of the overall (2.60 cm) RMSD of the model fit, indicating the model indeed captured the common kinematic characteristics quite well. We did not seek to adapt the model to be able to account for the left–right asymmetry observed in the selected data set, notably because for now we want it to remain as simple and general as possible. For the same reasons, we also did not include between-condition variations in the moment of onset of movement, which could explain the model’s somewhat slower-than-observed rise in velocity for the shorter movement amplitudes. This could be mitigated via a later onset of BS in the simulations of these particular conditions. The finding that the magnitude of BS, providing an initial push in the required direction, was in fact best set proportional to the ratio of xb over Tf is interesting as it thereby creates a soft connection between the current model and Peper et al.’s (1994) initial required velocity model.
As mentioned earlier, our choice to not include a linear damping term in the to-be-fitted model structure was based on the potential redundancy (in terms of fitting) of incorporating three damping terms. Indeed, fitting the model (Equation 4) with an additional linear damping term may inappropriately return a negative coefficient for this term, indicating energy injection rather than dissipation (see Mottet & Bootsma, 1999, p. 243, for a discussion of similar fitting problems). For the record, we note that fitting the model with combinations of linear and either Rayleigh or Van der Pol damping did not allow for capturing the empirically observed kinematic characteristics and their variations under the four xb × Tf combinations to a sufficiently satisfactory extent.
The operational dynamical model for lateral manual interception behavior developed in the present contribution seeks to capture the pattern formation principles at work in such spatiotemporally constrained discrete movements. As such, to paraphrase Beek, Peper, et al. (1995) it is a mathematical, phenomenological model that does not seek to provide an account of the observed phenomena in terms of their underlying causes. Rather, it allows for capturing—and further exploring—the regularities in the time evolution of the action system under different environmental conditions, here captured by the space–time constraints on the interceptive action. In this regard it is important to bear in mind that the goal of the participants in the Ledouit et al. (2013) study of lateral manual interception—from which the data modeled here were extracted—was to simply intercept the ball, making it closer to catching (e.g., Dessing et al., 2005; Michaels et al., 2006; Montagne et al., 1999; Peper et al., 1994) than to hitting. (e.g., Bootsma & Van Wieringen, 1990; Smeets & Brenner, 1995; Tresilian & Lonergan, 2002; Tresilian & Houseman, 2005). Catching and hitting tasks indeed differ in several aspects, perhaps most clearly brought out by the velocity profiles of the (hand) movements deployed. Our catching task is characterized by an initial rise and subsequent fall in the movement velocity over time, which can even approach zero at the moment of interception for longer ball flight durations (see Figure 1). Hitting tasks, on the other hand, are typically characterized by a movement velocity that continuously increases over time, with peak velocity being reached close to the moment of contact. As the movement behaviors are thus quite different, whether hitting movements can be captured by a model structure of the type developed in the present contribution for now remains an open question.
In the present contribution, we moved from proof of concept (Schöner, 1990; Jirsa & Kelso, 2005) to the production of realistic movement patterns for direction-constrained manual interception. Although the model developed was demonstrated to adequately capture ensemble-average movement patterns under different space and time constraints, motor performance is typically also characterized by some degree of variability, both in the movement pattern and in the outcome. In order to tentatively explore the model’s potential in accommodating such variability, we probed the effect of range-delimited random variations in model coefficients over repeated simulations of the four combinations of ball flight time (1.0 and 1.6 s) and ball arrival position (7 and 14 cm). Figure 5 presents an example of 500 movement trajectories thus generated (using Equation 5) for each condition with a 30% between-trial noise on coefficients a′, d′, and e′. Visual inspection of these families of movement trajectories revealed the typical (rising and falling) pattern of between-trial variability as a function of time for goal-directed movements (Darling & Cooke, 1987; Darling et al., 1988; Hansen et al., 2008).
Results of 500 runs of the parameterized model (Equation 5), with each run producing movement trajectories (thin lines, together creating surfaces) for all four combinations of movement duration and amplitude (for reasons of readability, rightward for 1.6-s duration and leftward for 1.0-s durations). Coefficients a′, d′, and e′ vary over runs as a result of multiplying each with a random number between 0.7 and 1.3, thereby adding 30% noise to the coefficients. The thick lines in each panel represented the model without noise. Note that the number of trajectories above and below the thick (no-noise) model trajectory are approximately equal.
Citation: Motor Control 2025; 10.1123/mc.2024-0036
Taking this analysis one step further, we also examined the outcome variability. The original data from Ledouit et al. (2013) revealed a pattern of spatial endpoint variability (captured by spatial variable error at the moment of ball arrival) corresponding to the well-known speed–accuracy trade-off (see Plamondon & Alimi, 1997, for a review): spatial variable error was largest for interceptive actions requiring the highest movement speed (i.e., 14-cm distance to be covered in 1.0 s) and smallest for interceptive actions requiring the lowest movement speed (i.e., 7-cm distance to be covered in 1.6 s), with the other two intermediate movement speed conditions producing intermediate variable error magnitudes. The exemplary model simulations shown in Figure 5 in fact reproduced this pattern: spatial variable errors calculated on the original data and on the model simulations revealed close to identical patterns, r(4) = .993, p = .007; R2 = .985. The model dynamics thus inherently incorporate the speed–accuracy trade-off. Future work within this research program will address the general pertinence of the model by testing it on different lateral manual interception data sets. We also intend to use the model to explore its potential in capturing two particularly intriguing findings in lateral interception: the angle-of-approach effect and, relatedly, the movement-reversal effect. The former effect resides in the presence of systematic differences in hand movement kinematics when intercepting rectilinearly moving targets arriving at the same interception position after the same flight duration while coming from different starting positions (Arzamarski et al., 2007; Ledouit et al., 2013; Montagne et al., 1999; Peper et al., 1994). The latter effect, of movement starting toward the left before intercepting the ball on the right, or vice versa, occurs under particular ball trajectory conditions when participants initiate their interceptive movement early on after movement onset (Montagne et al., 1999). For the present purposes, from the Ledouit et al.’s (2013) data set, we only retained the ball trajectories that were orthogonally oriented with respect to the interception axis. However, this data set also contains ball trajectories with different angles of approach to the same interception location and can therefore be used to explore the model’s capability (e.g., by having the goal location evolve over time) of reproducing the angle-of-approach effect. It could, in a similar way, also be capable of generating reversal movements.
Acknowledgment
This project received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement no. 956,003.
References
Arzamarski, R., Harrison, S.J., Hajnal, A., & Michaels, C.F. (2007). Lateral ball interception: Hand movements during linear ball trajectories. Experimental Brain Research, 177(3), 312–323.
Beek, P.J., & Beek, W.J. (1988). Tools for constructing dynamical models of rhythmic movement. Human Movement Science, 7(2–4), 301–342.
Beek, P.J., Peper, C.E., & Stegeman, D.F. (1995). Dynamical models of movement coordination. Human Movement Science, 14(4–5), 573–608.
Beek, P.J., Rikkert, W.E.I., & Van Wieringen, P.C.W. (1996). Limit cycle properties of rhythmic forearm movements. Journal of Experimental Psychology: Human Perception and Performance, 22(5), 1077–1093.
Beek, P.J., Schmidt, R.C., Morris, A.W., Sim, M.-Y., & Turvey, M.T. (1995). Linear and nonlinear stiffness and friction in biological rhythmic movements. Biological Cybernetics, 73(6), 499–507.
Bizzi, E., Accornero, N., Chapple, W., & Hogan, N. (1982). Arm trajectory formation in monkeys. Experimental Brain Research, 46(1), 139–143.
Bongers, R.M., Fernandez, L., & Bootsma, R.J. (2009). Linear and logarithmic speed–accuracy trade-offs in reciprocal aiming result from task-specific parameterization of an invariant underlying dynamics. Journal of Experimental Psychology: Human Perception and Performance, 35(5), 1443–1457.
Bootsma, R.J., & van Wieringen, P.C.W. (1990). Timing an attacking forehand drive in table tennis. Journal of Experimental Psychology: Human Perception and Performance, 16(1), 21–29.
Bullock, D., & Grossberg, S. (1988). Neural dynamics of planned arm movements: Emergent invariants and speed–accuracy properties during trajectory formation. Psychological Review, 95(1), 49–90.
Darling, W.G., Cole, K.J., & Abbs, J.H. (1988). Kinematic variability of grasp movements as a function of practice and movement speed. Experimental Brain Research, 73(2), 225–235.
Darling, W.G., & Cooke, J.D. (1987). Changes in the variability of movement trajectories with practice. Journal of Motor Behavior, 19(3), 291–309.
Dessing, J.C., Beek, P.J., Caljouw, S.R., & Peper, C.L.E. (2004). A dynamical neural network for hitting an approaching object. Biological Cybernetics, 91(6), 377–387.
Dessing, J.C., Bullock, D., Peper, C.E., & Beek, P.J. (2002). Prospective control of manual interceptive actions: Comparative simulations of extant and new model constructs. Neural Networks, 15(2), 163–179.
Dessing, J.C., Oostwoud Wijdenes, L., Peper, C.L.E., & Beek, P.J. (2009). Adaptations of lateral hand movements to early and late visual occlusion in catching. Experimental Brain Research, 192(4), 669–682.
Dessing, J.C., Peper, C.E., Bullock, D., & Beek, P.J. (2005). How position, velocity, and temporal information combine in the prospective control of catching: Data and model. Journal of Cognitive Neuroscience, 17(4), 668–686.
Feldman, A.G. (1986). Once more on the equilibrium-point hypothesis (λ model) for motor control. Journal of Motor Behavior, 18(1), 17–54.
Gonzalez, D.L., & Piro, O. (1987). Global bifurcations and phase portrait of an analytically solvable nonlinear oscillator: Relaxation oscillations and saddle-node collisions. Physical Review A, 36(9), 4402–4410.
Haken, H., Kelso, J.A.S., & Bunz, H. (1985). A theoretical model of phase transitions in human hand movements. Biological Cybernetics, 51(5), 347–356.
Hansen, S., Elliott, D., & Khan, M.A. (2008). Quantifying the variability of three-dimensional aiming movements using ellipsoids. Motor Control, 12(3), 241–251.
Hasan, Z. (1986). Optimized movement trajectories and joint stiffness in unperturbed, inertially loaded movements. Biological Cybernetics, 53(6), 373–382.
Hogan, N. (1984). An organizing principle for a class of voluntary movements. Journal of Neuroscience, 4(11), 2745–2754. http://www.jneurosci.org/content/4/11/2745.short
Jirsa, V.K., & Kelso, J.A.S. (2005). The excitator as a minimal model for the coordination dynamics of discrete and rhythmic movement generation. Journal of Motor Behavior, 37(1), 35–51.
Jordan, D.W., & Smith, P. (1987). Nonlinear ordinary differential equations. Clarendon Press.
Kadar, E.E., Schmidt, R.C., & Turvey, M.T. (1993). Constants underlying frequency changes in biological rhythmic movements. Biological Cybernetics, 68(5), 421–430.
Kawato, M., Maeda, Y., Uno, Y., & Suzuki, R. (1990). Trajectory formation of arm movement by cascade neural network model based on minimum torque-change criterion. Biological Cybernetics, 62(4), 275–288.
Kay, B.A., Kelso, J.A., Saltzman, E.L., & Schöner, G. (1987). Space–time behavior of single and bimanual rhythmical movements: Data and limit cycle model. Journal of Experimental Psychology: Human Perception and Performance, 13(2), 178–192.
Kay, B.A., Saltzman, E.L., & Kelso, J.A. (1991). Steady-state and perturbed rhythmical movements: A dynamical analysis. Journal of Experimental Psychology: Human Perception and Performance, 17(1), 183–197.
Ledouit, S., Casanova, R., Zaal, F.T.J.M., & Bootsma, R.J. (2013). Prospective control in catching: The persistent angle-of-approach effect in lateral interception. PLoS One, 8(11), Article e80827.
McIntyre, J., & Bizzi, E. (1993). Servo hypotheses for the biological control of movement. Journal of Motor Behavior, 25(3), 193–202.
Michaels, C.F., Jacobs, D.M., & Bongers, R.M. (2006). Lateral interception II: Predicting hand movements. Journal of Experimental Psychology: Human Perception and Performance, 32(2), 459–472.
Montagne, G., Fraisse, F., Ripoll, H., & Laurent, M. (2000). Perception-action coupling in an interceptive task: First-order time-to-contact as an input variable. Human Movement Science, 19(1), 59–72.
Montagne, G., Laurent, M., Durey, A., & Bootsma, R.J. (1999). Movement reversals in ball catching. Experimental Brain Research, 129(1), 87–92.
Mottet, D., & Bootsma, R.J. (1999). The dynamics of goal-directed rhythmical aiming. Biological Cybernetics, 80(4), 235–245.
Nelson, W.L. (1983). Physical principles for economies of skilled movements. Biological Cybernetics, 46(2), 135–147.
Ostry, D.J., & Feldman, A.G. (2003). A critical evaluation of the force control hypothesis in motor control. Experimental Brain Research, 153(3), 275–288.
Oubbati, F., & Schöner, G. (2013). Autonomous timed movement based on attractor dynamics in a ball hitting task. In Proceedings of the 5th International Conference on agents and artificial intelligence, February 15–18, 2013, Barcelona, Spain (pp. 304–311).
Peper, L., Bootsma, R.J., Mestre, D.R., & Bakker, F.C. (1994). Catching balls: How to get the hand to the right place at the right time. Journal of Experimental Psychology: Human Perception and Performance, 20(3), 591–612.
Plamondon, R., & Alimi, A.M. (1997). Speed/accuracy trade-offs in target-directed movements. Behavioral and Brain Sciences, 20, 279–249.
Polit, A., & Bizzi, E. (1978). Processes controlling arm movements in monkeys. Science, 201(4362), 1235–1237.
Schmidt, R.C., Corey, D., Fitzpatrick, P., & Riley, M.A. (2015). The oscillatory basis of Fitts’ law. In B.G. Bardy, R.J. Bootsma, & Y. Guiard (Eds.), Studies in perception and action III: Eighth international conference on perception and action, July 9–14, 1995, Marseille, France (1st ed., pp. 95–98). Routledge.
Schöner, G. (1990). A dynamic theory of coordination of discrete movement. Biological Cybernetics, 63(4), 257–270.
Schöner, G. (1994a). 17—From interlimb coordination to trajectory formation: Common dynamical principles. In S.P. Swinnen, H. Heuer, J. Massion, & P. Casaer (Eds.), Interlimb coordination: Neural, dynamical, and cognitive constraints (pp. 339–368). Academic Press.
Schöner, G. (1994b). Dynamic theory of action-perception patterns: The time-before-contact paradigm. Human Movement Science, 13(3–4), 415–439.
Schöner, G., Dose, M., & Engels, C. (1995). Dynamics of behavior: Theory and applications for autonomous robot architectures. Robotics and Autonomous Systems, 16(2–4), 213–245.
Schöner, G., & Santos, C. (2001). Control of movement time and sequential action through attractor dynamics: A simulation study demonstrating object interception and coordination. In SIRS 2001 Proceedings of the 9th International symposium on intelligent robotic systems, July 18–20, 2001, Toulouse, France (pp. 15–24). https://repositorium.sdum.uminho.pt/bitstream/1822/4327/1/paper_25jan2001.pdf
Schweighofer, N., Arbib, M.A., & Kawato, M. (1998). Role of the cerebellum in reaching movements in humans. I. Distributed inverse dynamics control. European Journal of Neuroscience, 10(1), 86–94.
Sharp, R.H., & Whiting, H.T.A. (1974). Exposure and occluded duration effects in a ball-catching skill. Journal of Motor Behavior, 6(3), 139–147.
Smeets, J.B., & Brenner, E. (1995). Perception and action are based on the same visual information: Distinction between position and velocity. Journal of Experimental Psychology: Human Perception and Performance, 21(1), 19–31.
Tresilian, J., & Lonergan, A. (2002). Intercepting a moving target: Effects of temporal precision constraints and movement amplitude. Experimental Brain Research, 142(2), 193–207.
Tresilian, J.R., & Houseman, J.H. (2005). Systematic variation in performance of an interceptive action with changes in the temporal constraints. The Quarterly Journal of Experimental Psychology Section A, 58(3), 447–466.
Whiting, H.T.A. (1968). Training in a continuous ball throwing and catching task. Ergonomics, 11(4), 375–382.
Whiting, H.T.A. (1969). Acquiring ball skills: A psychological interpretation. Bell.
Whiting, H.T.A., Gill, E.B., & Stephenson, J.M. (1970). Critical time intervals for taking in flight information in a ball-catching task. Ergonomics, 13(2), 265–272.
Whiting, H.T.A., & Sharp, R.H. (1974). Visual occlusion factors in a discrete ball-catching task. Journal of Motor Behavior, 6(1), 11–16.
Zaal, F.T.J.M., Bootsma, R.J., & van Wieringen, P.C.W. (1999). Dynamics of reaching for stationary and moving objects: Data and model. Journal of Experimental Psychology: Human Perception and Performance, 25(1), 149–161.
