Exp Brain Res (2008) 188:275–288 DOI 10.1007/s00221-008-1360-6

R ES EA R C H A R TI CLE

The time course of amplitude speciWcation in brief interceptive actions Welber Marinovic · Annaliese Plooy · James R. Tresilian

Received: 28 May 2007 / Accepted: 19 March 2008 / Published online: 16 April 2008 © Springer-Verlag 2008

Abstract The interception of fast moving objects typically allows the object to be seen for only a short period of time. This limits the time available to prepare the movement. To deal with short preparation intervals, performers are likely to prepare a motor program in advance. Although motor preparation may begin before the target is seen, accuracy requires that certain program parameters are determined from observations of the target. In the experiments reported here we sought to determine the last moment at which information about the distance to move (amplitude) can be incorporated into a program. We employed an empirical protocol that allowed us to examine whether new amplitude information is incorporated discretely or continuously into the program during short intervals prior to movement onset (MO)—the preparation interval. Participants were trained to hit targets at two diVerent distances with movements of a speciWc duration (180 ms): targets were moving in “Experiment 1” and stationary in “Experiment 2”. This method permitted an estimate of MO time. Preparation intervals were manipulated by delivering a stimulus cue for movement amplitude at varying times prior to the estimated MO. Results demonstrated that amplitude information could be eVectively incorporated into the program provided the preparation interval was greater than about 200 ms. In addition, the results indicated that amplitude was speciWed predominantly in a discrete

W. Marinovic (&) · A. Plooy · J. R. Tresilian Perception and Motor Systems Laboratory, School of Human Movement Studies, The University of Queensland, St Lucia 4072, Australia e-mail: [email protected] J. R. Tresilian Department of Psychology, University of Warwick, Coventry CV4 7AL, UK

manner, though the number of responses directed towards a central default amplitude suggest that the distance between targets was near to a threshold for continuous speciWcation. Keywords Human · Interception · Motor control · Motor preparation · Movement

Introduction Interception of a moving object often involves preparing elements of performance—such as movement direction, speed and amplitude—in advance of starting to move. In ball sports, a receiving player’s preparation for an upcoming interception may involve the use of information obtained from an opponent’s behavior before the ball has started to move in their direction (e.g., Abernethy 1990a, b; Morya et al. 2003; Savelsbergh et al. 2002). Information obtained from previous attempts can also contribute to interceptive preparation: for example, in an experimental setting, people have been found to make use of the characteristics of moving targets experienced on previous trials (de Lussanet et al. 2001; 2002; Gray 2002a, b). Once the object is moving towards them, a person can obtain information from its motion about where and when to make the interception (Regan and Gray 2000; Tresilian 1999) and so further preparation occur during this period. Thus, preparation of interceptive actions can begin well in advance of starting to move and continue, in principle, right up until the moment descending commands are issued. In the studies reported here, we examine the preparation of interception during the 0.5 s or so immediately prior to starting to move during which information is available from the moving target only. This complements previous studies that have investigated advance preparation prior to the start of

123

276

target motion based on cues from an opposing player in ball sports and/or on task history (e.g., Abernethy 1990a, b; Morya et al. 2003; Savelsbergh et al. 2002). Although 0.5 a second may seem a very short period of time, it is not uncommon for players of ball sports to have only about this amount of time to view a moving ball before actually making the interception (e.g., Watts and Bahill 1990). Since this is the only time during which information about the actual trajectory of the object can be obtained, it is likely to be crucial for making precise and accurate interceptions. Furthermore, since many interceptions of this sort involve very brief movements, often lasting less than 200 ms (Bootsma and Vanwieringen 1990; Marinovic et al. 2004; Schmidt and Lee 2005; Watts and Bahill 1990), the possibility for making eVective feedback-based corrections once the movement has begun is extremely limited. This makes it likely that such interceptions are largely determined by motor programs prepared in advance of starting to move (Schmidt and Lee 2005; Tyldesley and Whiting 1975). This is not to claim that all interceptive actions are controlled in this way. There are alternative models that do not involve motor programs but instead continuously transform sensory input into motor output (e.g., Dessing et al. 2002; Peper et al. 1994; Tresilian 1994), which may be more appropriate for longer duration interceptions. However, these models have little, if anything, to say about advance preparation and we do not consider them further here. In order for a program-governed movement to be accurate, the parameters of the program must be determined, at least in part, from observations of the actual trajectory of the moving target. Standard models of motor preparation in aimed movement tasks involve at least two processes (Rosenbaum 1980; Schmidt and Lee 2005): a response selection process in which the program for the chosen type of movement is activated and a parameter speciWcation process in which characteristics of the required movements (such as amplitude, duration and direction) are speciWed. We have recently proposed a version of this type of model for rapid interceptive actions (Tresilian 2005) in which the sequence of events are as illustrated in Fig. 1a. First, the performer selects a particular program corresponding to the type of interception they intend to make and can use advance information and prior expectations to set its parameters. The parameters can be altered or Wne-tuned as more information becomes available. Most important in this respect is information derived from (visual) observation of the moving target itself. This information can only be useful if obtained prior to the moment the program starts to generate commands, which occurs when a speciWc, criterion value of the target’s time to arrival at the chosen interception point (TTCcrit) is reached. The time interval between the moment when the target is Wrst seen and the

123

Exp Brain Res (2008) 188:275–288

time when TTCcrit is reached is called the Wnal stage of motor preparation in Fig. 1a. This stage is critical because it is the only period during which information about the actual motion of the target can be used to determine the movement: it is clearly of interest to investigate the processes involved. The experiments reported here examined the eVects of providing the information needed for correct program parameter speciWcation after the target was Wrst seen. An experimental technique well-suited to investigating the time-course of parameter speciWcation immediately prior to movement initiation is the timed response paradigm (Ghez et al. 1990; Schouten and Bekker 1967). In this paradigm, participants initiate a response at a particular moment in time and information about exactly what response they are required to make is made available to them by means of an indicational stimulus presented at any time prior to this moment. The time between presentation of the stimulus and the initiation of the response is called the stimulus–response (S–R) interval. This interval represents the time the person has available to prepare the response on the basis of the stimulus information and is the primary independent variable. Using this protocol it is possible to estimate the last moment at which information can be incorporated into the program and to determine the way in which this incorporation is achieved (Ghez et al. 1997, 1990). Ghez and colleagues have identiWed two modes of incorporation which they term the continuous mode and the discrete mode. The modes primarily diVer in what is predicted to occur across randomly presented target trials as the S–R interval increases. In the continuous mode, a parameter value change occurs as a continuous shift from an initially single value on trials with short preparation intervals to values that progressively conform with the appropriate target locations as the preparation interval increases. In the discrete mode, a program parameter is speciWed as having either one value or another. As the preparation interval increases so does the percentage of responses toward the correct target. The two experiments reported in this paper sought to investigate the speciWcation of movement amplitude during the Wnal stage of preparation in an interceptive task. The task was to hit a target that could either require a small amplitude movement (‘near’ target) or a larger amplitude movement (‘far’ target). We were interested in both the last moment at which incorporation of new amplitude information into the program could be made and the mode in which the change was accomplished and we followed the methods introduced by Ghez and colleagues to make these assessments. In the Wrst experiment, the task was an interceptive hitting task similar to that used in earlier studies (Tresilian and Lonergan 2002) and a modiWed version of the standard timed response protocol was employed. In the standard

Exp Brain Res (2008) 188:275–288

277

Fig. 1 Temporal sequence of events taking place during the interception of a moving object in the program model adapted from Tresilian (2005). a When the TTC of the moving target (TTCtgt) is equal to TTCcrit it takes some time to process the visual information (perceptual transmission time, PT) and trigger the MP and another time for the motor commands to reach the muscles (transmission time, TT). PT + TT is the visuomotor delay. The time between movement onset and the time when the target is hit determines movement time (MT).

b Sequence of events in the timed response protocol of “Experiment 1”. An indicational stimulus that indicated the required movement amplitude is presented after the targets become visible but prior to movement onset (estimated to be 180 ms prior to the target reaching the hitting location). The S–R interval is the time between presentation of the indicational stimulus and the actual movement onset. The S–E interval is the time between stimulus presentation and the expected movement onset

protocol the time of response initiation is determined by requiring participants start their movement in synchrony with the last of a series of tones (Ghez et al. 1990; Schouten and Bekker 1967). This is not suitable for interception tasks and so we adopted the protocol used in a previous study (Tresilian and Plooy 2006a) in which participants are trained to make interceptive movements of a speciWc movement time (in the experiment reported here this time was 180 ms) permitting the time of movement onset to be estimated as 180 ms prior to the target’s arrival at the hitting location. A stimulus indicating which movement amplitude would be required was provided at variable times prior to the estimated onset time. These times correspond to diVerent S–R intervals, but since the S–R interval is the time between the stimulus and the actual onset it is not necessarily exactly the same as the interval between the stimulus and the expected onset. To distinguish the two we will refer to the latter as the S–E interval, which was the primary independent variable in the experiments. The sequence of events in the experimental task is illustrated in Fig. 1b: it would be expected that program alterations could be made in response to the indicational stimulus only if it is presented within the Wnal stage of preparation (i.e., S–R inter-

val is greater than the visuomotor delay). However, it may be possible to make alterations even during the time (PT) during which TTC information is being processed and transmitted to the program circuits, since the program has not at this stage been triggered. In short, program modiWcation may be possible right up to the moment of triggering (Fig. 1a): the magnitude of the minimum S–E interval at which response modiWcations are possible can provide a means for assessing this possibility. Since the task and parameters are rather diVerent from those in the previous studies of Ghez and colleagues, a second experiment was run using the standard timed response protocol for comparison purposes. In this experiment, the task was similar to that used in the Wrst except that the target remained stationary (participants simply had to hit it).

Experiment 1 Participants Participants were 10 (9 men and 1 woman, age range 25–42; mean = 30 years) self-reported right-handed adults with

123

278

normal or corrected to normal vision. All participated voluntarily in the experiment and gave their informed consent prior to commencement of the testing sessions. Task and apparatus The experimental task was to hit one of two possible targets located at diVerent distances from the initial position of the bat performing a movement of 180 ms of duration for both amplitudes. The apparatus used in the Wrst experiment is shown in Fig. 2. The 1-d.f. task employed by Tresilian and Lonergan (2002) was adapted to allow the investigation of the time course of amplitude speciWcation. Originally in the 1-d.f. hitting task only one target was mounted on a carriage, which was attached to a belt system driven by computer controlled motor torque. The participants were constrained to move the handle along a straight path above and perpendicular to the target track. For this experiment, two targets (near- and far-target) were mounted on a platform attached to the carriage. The two targets (near and far) were separated by a distance of 16 cm from each other. The near target was 10 cm away from the initial position of the bat and therefore the distance to be moved to reach the far target was 26 cm. As in previous experiments, the participants were constrained to move along a straight path, however, the target to be hit was speciWed during its approach to the interception point. The target to be hit was illuminated by an LED (the indicational stimulus); simultaneously the target not to be hit was tipped over by a small solenoid motor controlled by a computer. This pair of events indicated the amplitude of the movement. Fig. 2 Diagram of the 1-d.f. hitting task showing the set up used in “Experiment 1” and “Experiment 2”. Lateral view (a) and plan view (b). The target is attached to the target track belt and moves along a straight path. The intersection between the target track and the manipulandum track determines the position in which the target must be intercepted. The participants were constrained to move the manipulandum only along the Z axis

123

Exp Brain Res (2008) 188:275–288

The targets were made of clear plastic material with embedded LEDs which illuminated the speciWed target at varying times (see below). The targets were Xat and rectangular, 9 cm tall and 6.1 cm in length. The targets were seen for 2.2 s before their arrival at the strike zone in all trials, and they moved with a constant velocity of 1.5 m/s. Ambient light was dim, so that the targets were distinctive but it did not impair vision of the surrounding objects (handle, manipulandum, and others). The manipulandum consisted of an aluminium handle mounted on a linear slide which ran with minimal friction along the manipulandum track. Mounted below the linear slide there was a steel rod, 0.5 cm diameter, which served to physically strike the moving targets (Fig. 2). Infrared emitting diodes (IREDs) were Wxed to the carriage in which the targets were housed and to the handle of the manipulandum. The positions of these IREDs were sampled at 200 Hz during experimental trials using an Optotrak (Northern Digital Inc.) optoelectronic movement recording system and stored on computer disc. Procedures and design Prior to the experimental session, the participants were trained to hit the targets with a movement time of 180 ms. They were instructed to make one single movement. During training and experimental sessions the participants were provided with knowledge of results (KR) about their performance. The KR informed the participants about their MT. The participants performed 80 trials in the training session. Within this amount of practice participants were

Exp Brain Res (2008) 188:275–288

able to produce movements whose mean MT (§10%) matched those required by precued targets. In experimental trials targets were presented as pairs as shown in Fig. 2 so that both were initially visible to the participant. During a trial the indicational stimulus (illumination of an LED) speciWed the target that was to be hit. Indicational stimuli could be presented at various times prior to the expected moment of movement onset (i.e., 180 ms prior to the target reaching the hitting location), these are the S–E intervals. Eight diVerent S–E intervals were used: 50, 100, 150, 200, 250, 300, 350, and 400 ms. As in the protocol used by Ghez and colleagues, participants performed experimental trials under two diVerent conditions, predictable (P) and unpredictable (UP). In the P condition, participants were given a precue concerning which target should be hit. This cue was an illumination of the target’s LED without a simultaneous tip of the nontarget. It was presented at the beginning of the trail (immediately after the targets started to move) and lasted for 300 ms. In this condition, the indicational stimulus provided no additional information about the target to be hit. In the UP condition, no precue was presented and participants had to wait for the indicational stimulus in order to know which target to hit. P and UP trials were performed in blocks and the diVerent S–E intervals and movement amplitudes were presented pseudorandomly within the blocks. To avoid order eVects of task presentation, half of the participants began the experimental session with the P block, and half of them began with the UP block. The participants performed 16 trials for each of the 8 S–R intervals in each of the two experimental blocks (256 trials in total). Data reduction and dependent variables All data reduction was performed using custom Labview software (version 7.1, National Instruments). Any missing data of the IREDs was interpolated by cubic spline, as long as the number of missing samples did not exceed 10% of the sampling frequency (200 Hz). The position data time series were digitally Wltered by dual pass through a second order Butterworth Wlter with a cut-oV frequency of 20 Hz. The movement onset was calculated from the tangential speed time series using the two-stage algorithm suggested by Teasdale et al. (1993). The algorithm Wrst determines the sample (S1) at which the time series Wrst exceeds 10% of its maximum value. Then working back from S1 it Wnds the Wrst sample (S2) at which speed reaches 10% of the speed value at S1. Working forward from S2 the Wnal step of the algorithm locates the onset being the sample at which speed equals the average value plus standard deviation between S1 and S2. The time at which the target was hit as well as the temporal error were estimated from the position time series of the manipulandum IRED and the target IRED. The

279

sample at which the Z-position (Fig. 2) of the manipulandum reached the Z-position occupied by the target determined the time of target strike when the target is actually hit. In order to compare the accuracy of movements in the two experimental conditions (P and UP) and to detect when the S–R interval was long enough to allow complete amplitude speciWcation of the movement, both peak velocity up to 10 cm from the initial position of the bat (V10) and the temporal error to hit the center of the target were examined. Peak velocity is believed to represent a pre-programmed portion of rapid aiming movements (Elliott et al. 2001). It is known from previous research that peak velocity increases with longer target distances (Gordon and Ghez 1987; Gordon et al. 1994; Welsh and Elliott 2004). In addition, the peak velocity of rapid interceptive actions is generally close to the moment of interception (Caljouw et al. 2004; Marinovic et al. 2004; Tresilian and Lonergan 2002) and therefore indicates which target the participants are trying to aim at. A higher value of peak velocity would indicate they were aiming at the far target, whereas a lower value of peak velocity would indicate they were aiming at the near target. We employed V10 rather than the actual value of peak velocity up to the moment of interception because in the UP block the participants could specify their responses either towards the near target (at 10 cm from the initial position of the bat) or somewhere after this point. If the participants aimed at the near target the movement should last approximately 180 ms. If they aimed at a point beyond the 10 cm mark their movements ought to be faster and therefore briefer than 180 ms when they reached this mark. As a result the time available to correct their responses at this distance was limited to a maximum of 180 ms which is below the minimum latency of 200 ms for feedback-based error corrections for amplitude information (Saunders and Knill 2005; see also Brenner et al. 1998, for online corrections in interceptive actions). Thus V10 represents an index of the preprogrammed velocity of the movement which is less likely to be inXuenced by online corrections (e.g., acceleration of the moving limb to reach the far target when Wrst aiming at the near target). The standard deviation of V10 is expected to inform us when the time available to prepare the correct amplitude was adequate. In the case the participants employed a strategy to aim at an intermediate position between the two targets when the S–R interval was short (<150 ms) the standard deviation of V10 could be unmodiWed regardless of the time available to prepare the movement. Nonetheless, if they employed this strategy (known as the continuous mode of speciWcation) we should be able to determine the time necessary to specify the correct amplitude by the analysis of the variable temporal error (VTE). The VTE reXects the consistency or variability with which the participants can hit the moving target. It was

123

280

expected that this variable would indicate the temporal conditions in which the time available to specify amplitude was insuYcient. For the calculation of the standard deviation of V10 and the VTE we computed the standard deviation of eight trials in each experimental condition. These measures were expressed in terms of the actual time the participants had to prepare their responses, that is, the S–R interval (see Fig. 1b). Both hits and misses were used in the calculation of these variables.

Exp Brain Res (2008) 188:275–288

freedom. A Tukey HSD post-hoc test, P < 0.05, was conducted to determine the locus of signiWcant diVerences involving more than two means. The rationale behind this analysis is that when the participants have enough time to prepare the response they would both specify the correct movement amplitude and initiate the action in time to hit the target accurately. Here we expected a two-way interaction between predictability and S–R interval. The post-hoc analysis would indicate at which S–R intervals the performance was signiWcantly deteriorated by the lack of time to prepare the response in the UP block.

Data analysis To investigate the strategies employed to specify amplitude (continuous or discrete) the distributions of V10 of individual responses were plotted for each S–R interval. Histograms were built with bins of 135 mm/s each. The total number of values within each bin was transformed as a percentage of the total number of samples and Wtted with a cosine weight function (Chambers et al. 1983). This weight function is half the cosine function with its peak at the center value. It decreases symmetrically to zero, after which a weight of zero is applied. Therefore, data points have a smaller and smaller impact on the Wtted lines as they are further and further from the center value. This graphical method of analysis allowed us to examine the relative concentration of data points along diVerent S–R intervals and this indicated which mode of amplitude speciWcation was adopted. To determine the last moment at which modiWcations to the motor program could be implemented, the accuracy of the responses performed by the participants in each experimental condition was compared. For this purpose average values of VTE and standard deviation of V10 were submitted to separate 2 (predictability: P, UP) £ 2 (target: near, far) £ 8 (S–R interval: 50, 100, 150, 200, 250, 300, 350, and 400 ms) repeated measures analysis of variance. Departures from sphericity were veriWed through the Mauchley’s test of sphericity and when necessary the Huynh–Feldt’s method was used to correct the degrees of Fig. 3 Mean movement time (§SD) in the P (a) and UP (b) blocks. N near target, F far target. The horizontal gray line represents the MT goal

123

Results and discussion Figure 3 shows the MT results from “Experiment 1” across conditions in the P and UP blocks. As can be seen from Fig. 3a, in the P block the participants were consistent to produce an average MT which was close to 180 ms regardless of the S–R interval. Conversely, in the UP block MT tends to be more distant from its required duration in comparison with the P block, especially for short preparation intervals (·150 ms). The mode of speciWcation In the continuous mode it is expected that participants would prepare an intermediate-movement amplitude between the two targets, particularly at the shortest S–R interval. This intermediate amplitude bias should progressively diminish as the preparation interval increases. Since the movement does not stop at the speciWed distance, movement amplitude itself does not provide a suitable performance index. The maximum velocity does provide an index. Peak velocity occurs at, or at least very close, to the moment of target contact, so when peak velocity is reached, the manipulandum has moved approximately through the programmed movement distance. The value of peak velocity is determined by the programmed amplitude since MT was held constant across distances. Thus, the concentration

Exp Brain Res (2008) 188:275–288

of V10 (which is an index of the preprogrammed movement speed) data points in the UP block is expected to gather around an intermediate or central value of peak velocity at the shortest S–R interval. The concentration of data points should then become progressively more distinct as the S–R interval increases. This means that histograms of V10 values should show a unimodal distribution at the shortest S–R interval (50 ms), and a progressively more bimodal distribution at longer S–R intervals. In the discrete mode of speciWcation, participants specify movement amplitudes appropriate for one target or the other. In this mode, speciWcation should also improve gradually as the preparation interval becomes longer, but this improvement should be noticed in the number of responses towards the correct target increasing and not in the amplitude of the responses. Histograms of V10 ought to show a bimodal distribution regardless of the S–R interval. To determine whether the distribution had one or more peaks we considered a peak as such when the values of the Wtted line rose to a maximum and then dropped again rather smoothly. All peaks were visually obvious. In the P block, all participants reached similar values of V10 independently of the S–R interval as would be expected. As can be seen in Fig. 4a, in the P block the responses were clustered along two distinct values (represent by the two peaks in the histograms) and no mistakes were committed by the participants in terms of preparing a slow movement towards a far target or vice-versa. This pattern was evident for all participants in the experiment.

Fig. 4 Histograms of V10 for responses made by all participants to predictable (a) and unpredictable (b) at two S–R intervals (50 and 400 ms). Distributions were Wtted with a smooth line using a cosine weight function. The histograms show responses required to both targets (near and far). Vertical gray lines indicate the most likely value of peak velocity for near and far target in the P block

281

Figure 5 shows the distribution of V10 for three representative participants in the P (A) and UP (B) blocks at the shortest S–R interval. As shown in Fig. 5b, in the UP block responses produced with about 50 ms of preparation are clustered in at least two distinct areas of each participant’s graph. At this preparation interval, however, the participants tended to prepare some of their responses towards the unspeciWed target as well as towards the mean range of targets, this result was consistent across all participants as shown in Figs. 4b (group data) and 5b (for three representative participants). The presence of responses speciWed towards the mean range of targets, more clearly seen in Fig. 5b for participant TO, could be either a switch between the two types of speciWcation (discrete and continuous) or the result of amendments attempted by the participants to strike the far target when originally aiming for the near target. As we will show in the next subsection, we found evidence that in comparison to the P block a larger percentage of responses in the UP block were modiWed during movement execution at S–R intervals ·100 ms. Thus, rather than occasionally aiming for a mean distance between targets (which could suggest continuous speciWcation), the presence of a middle peak towards the average range of targets can be explained in part by voluntary corrections to stochastically speciWed movements. Corrections to ongoing movements could explain why at the shortest S–R in Fig. 4b the most likely value of V10 for responses speciWed towards the near target slightly overshot that observed in the P block. On the other hand, when the participants chose to aim towards the far target, the value of V10 was very similar to that observed in the P block which indicates they did not have time to employ corrections up to that moment. This result, therefore, indicates that at the shortest preparation interval the participants did not aim systematically towards one single central default amplitude around the middle target range. Histograms of the distribution of V10 in the UP block of trials were bimodal for 8 out of 10 participants at the S–R interval of 50 ms (represented by participants MH and AB in Fig. 5b), whereas the other 2 participants showed more than two peaks (see participant TO in Fig. 5b). For the majority of the participants the strategy was to aim towards either the near target (participant AB in Fig. 5) or slightly ahead of the near target (participant MH in Fig. 5). At the shortest S–R interval, the group data in Fig. 4b shows that the number of responses towards a position close—t 1/3 of distance between the near and far targets—to the near target in the UP block was greater than towards the far target. The fact that the participants clustered a great proportion of their responses around either one of the two possible values of V10 (see Figs. 4b, 5b), and the observation that the histograms show at least two peaks (bimodal distribution) of this variable even at the shortest S–R interval indicates that the mode of amplitude speciWcation was predominantly discrete.

123

282

Exp Brain Res (2008) 188:275–288

Fig. 5 Histograms of V10 for responses made by three representative participants to predictable (a) and unpredictable (b) at the shortest S–R interval. Distributions were Wtted with a smooth line using a cosine weight function. The histograms show responses required to both targets (near and far). Vertical gray lines indicate the most likely value of peak velocity for near and far target in the P block

The time course of movement preparation Figure 6 shows the average VTE for P and UP blocks of trials at all S–R intervals: the mean VTE appears independent of S–R interval in the P block (Wlled squares), whereas in the UP block the VTE decreases consistently as the S–R interval increases (open circles). This pattern was conWrmed by a signiWcant two-way interaction between predictability and S–R interval (F(7,63) = 8.05; P < 0.001, p2 = 0.47). The post-hoc analysis of this interaction revealed that the VTE was signiWcantly higher in the UP conditions at S–R intervals of 50, 100, and 150 ms than at the same intervals in the P block of trials as shown in Fig. 6. These results suggest that at S–R intervals in the order of 150 ms or less the participants were not able to either prepare the appropriate response to hit the speciWed target or start their movements at the correct moment. Con-

Fig. 6 Mean variable temporal error to hit the center of the target at all S–R intervals for P and UP blocks of trials. The symbols mark the signiWcant contrasts between P and UP conditions (*P < 0.01; **P < 0.001)

123

versely, when the participants had about 200 ms before movement onset to prepare their responses, the accuracy to hit the moving target was similar to that observed in the P block. Considering that the analysis of the mode of speciWcation showed that the participants speciWed amplitude discretely, the variability of V10 might be expected to show the following pattern. Since the participants were often aiming at the unspeciWed target when the preparation intervals were short, the standard deviation of V10 should remain greater than in the P block until the moment when the S–R intervals were long enough to consistently prepare the correct movement. This prediction, therefore, would be supported by an interaction between predictability and S–R interval as long as greater variability of V10 was observed in the UP block at short S–R intervals. Figure 7 shows the standard deviation of V10 in the P and UP blocks of trials for each S–R interval. As for VTE, the mean standard deviation of V10 seems independent of S–R interval in the P block (Wlled squares), whereas in the UP block the standard deviation of V10 declines steadily until the S–R interval of 200 ms (open circles). A repeated measures ANOVA conWrmed our expectation of interaction between predictability and S–R interval (F(7,63) = 8.30, P < 0.001, p2 = 0.50). Further analysis of this interaction revealed that variability in the UP conditions was greater than variability in the P conditions at S–R intervals of 50, 100, and 150 ms as shown in Fig. 7. These results show that the velocity of the movements produced by the participants was more consistent when the preparation interval was about 200 ms or greater. Since the participants were not instructed to avoid correctional movements, it is possible that such corrections were made. If so, the time required for preparation in hitting

Exp Brain Res (2008) 188:275–288

Fig. 7 Standard deviation of V10 for the diVerent experimental conditions in “Experiment 1”. The symbols mark the signiWcant contrasts between P and UP conditions (**P < 0.001)

actions may have been underestimated. An analysis of the speed proWle of the participants showed neither more than one peak nor evident inXection points in the speed proWle which could be considered as indicative of corrective movements as it is shown for a representative participant in Fig. 8. Also, we analysed the acceleration proWles of individual trials to verify the presence of sub-movements which could be considered as online corrections using the same method described in similar studies with interceptive tasks (Tresilian and Plooy 2006b). Minima in the acceleration proWle were used to deWne whether the movement had one or more components. A one-component movement (monophasic movement) was one with no minima in the acceleration proWle a two-component (biphasic) movement has one minimum and so on. A minimum in the acceleration proWle was accepted as such only if the diVerence between the

283

lowest of the two peaks and the minimum point between them was greater than 2% of the maximum overall acceleration. Figure 9 shows that in average 98% of the movements in the P block of trials had one single component (monophasic movements). Similarly, the percentage of monophasic movements in the UP block of trials was 93% across all conditions. It is important to notice here that the small diVerence in percentage is due mainly to diVerences between P and UP blocks at S–R intervals <200 ms. To further investigate these diVerences we conducted a non-parametric test of Wilcoxon on the monophasic proportions. This test showed that the proportion of monophasic movements was signiWcantly higher for the P condition only at the S–R interval of 50 ms (z = ¡2.22, P < 0.05, r = ¡0.70) and 100 ms (z = ¡2.26, P < 0.05, r = ¡0.71). These results suggest that possible corrective amendments employed by the participants were ineVective to improve performance at short S–R intervals (<150 ms) and that at longer preparation intervals (¸150 ms) there was no evidence that corrective movements were employed.

Experiment 2 Overall, the results of the Wrst experiment are consistent with those of previous research (Favilla 1997, 2002; Ghez et al. 1997, 1990) indicating that the accuracy of responses in the UP block of trials is dependent upon the S–R interval. More speciWcally, the results showed that when the participants were uncertain about the upcoming target, a minimum preparation interval of about 200 ms was required for the participants to hit the targets with both accuracy and speed similar to those demonstrated in the P block. In addition, we found

Fig. 8 Example speed proWles from an individual participant (JO) at the S–R interval of 200 ms in the P (a) and UP (b) blocks of trials in “Experiment 1”. Dark dashed lines represent movements towards the far target and thin gray lines represent movements towards the near target

123

284

Exp Brain Res (2008) 188:275–288

Procedures and design The procedures and design were identical to those used in “Experiment 1”. Once again, prior to the experimental session the participants were trained to hit the targets with a movement time of 180 ms (§10%). However, since the targets were stationary, in the P block of trials the precue (illumination of the LED of the target to be hit for 300 ms) was presented 1 s before the Wrst of four successive 50-ms tones occurred. As in the original time response method, participants were required to make a single and uncorrected movement. Fig. 9 Percentage of monophasic movements across all conditions for the group. The symbols mark the signiWcant contrasts between P and UP conditions (9P < 0.05)

evidence that most participants speciWed amplitude by switching between two values discretely rather than shifting gradually from an intermediate value. The purpose of the second experiment was twofold. Firstly, we intended to compare the protocol used in “Experiment 1” with the original timed response method introduced by Ghez et al. (1990). Secondly, we examined whether similar temporal demands of the task would yield comparable results to those in “Experiment 1” regardless the fact that the targets were stationary in “Experiment 2”.

Data reduction and dependent variables Data reduction, dependent variables, and data analysis were identical to “Experiment 1” with one exception. Since the targets were stationary in “Experiment 2”, there was no temporal error to hit the targets and therefore only V10 and its standard deviation were analysed in this experiment.

Results and discussion

Eight participants (5 men and 3 women, age range 22–32; mean = 26 years) took part in this experiment. All participated voluntarily in the experiment and gave their informed consent prior to commencement of the testing sessions. They received a small Wnancial compensation for their participation (AU$20).

Figure 10 shows the MT results from “Experiment 2” across conditions in the P and UP blocks. As in “Experiment 1”, Fig. 10A shows that the participants were relatively consistent to produce an average MT which was close to 180 ms independently of the S–R interval in the P block. On the other hand, in the UP block MT tends to be more distant from its required duration. DiVerent from “Experiment 1”; however, MT at the shortest S–R interval for responses towards the near target are very close to the MT goal, whereas for the far target the error is greater than in the Wrst experiment. This result may be interpreted as evidence for a trend in “Experiment 2” to prepare a default response towards the near target that was closer to the value of V10 observed in the P block.

Task and apparatus

The mode of speciWcation

In this experiment the task required the participants to attend to two simultaneous task demands: (a) to initiate a response in synchrony with the last in a series of four successive 50-ms tones (ca. 50–60 dB, intertone interval 500 ms) presented through earphones, and (b) to hit stationary targets at two diVerent distances, presented visually at an unpredictable time prior to the last tone, performing movements of a speciWc duration. The same apparatus used in “Experiment 1” was also employed in “Experiment 2”. However, in “Experiment 2” the targets to be hit by the participants were in a Wxed position along the path of the manipulandum track (see Fig. 2).

As in “Experiment 1”, participants reached similar values of V10 independently of the S–R interval in the P block. Similarly to “Experiment 1”, it can be seen in Fig. 12b that at the shortest S–R intervals in the UP block the participants tended to prepare some of their responses towards either the unspeciWed target or the average range of targets. This result was the same for all participants except for participant MO (shown in Fig. 12) and OS (not shown) for whom the concentration of V10 data points in the UP block gathered around a central value of peak velocity at the shortest S–R interval. In “Experiment 2”, the participants did not overshoot the near target as much as the participants

Methods Participants

123

Exp Brain Res (2008) 188:275–288

285

Fig. 10 Mean movement time (§SD) in the P (a) and UP (b) blocks. N near target, F far target. The horizontal gray line represents the MT goal

in the Wrst experiment, but they markedly undershot the far target at short S–R intervals as seem in Fig. 11b. For most of the participants, the distributions of V10 at the S–R interval of 50 ms were bimodal (four out of eight participants) as for participant LM (shown in Fig. 12), or had more than two peaks (two out of eight) as for participant FB (shown in Fig. 12). Nonetheless, two participants (MO and OS) showed a unimodal distribution which could be interpreted as a manifestation of the continuous mode of speciWcation (see participant MO in Fig. 12). For these two participants the responses at short preparation intervals always overshot the near target and undershot the far target. This result suggests that for these two individuals the responses were never speciWed stochastically towards one target or the other as for the majority of the participants in “Experiment 1” and “Experiment 2”. Despite the occurrence of what can be considered as evidence for the continuous mode of spec-

iWcation, our results indicate that the mode of amplitude speciWcation was discrete or at the discrete-continuous threshold point for most participants. The time course of movement preparation Since our analysis of the mode of speciWcation indicated that the mode of speciWcation was the same for both experiments, we expected that the analysis of standard deviation of V10 to also show similar results in terms of the time required to specify the responses. Figure 13 shows the standard deviation of V10 in the P and UP blocks at all S–R intervals. The repeated measures ANOVA of this variable showed, as expected, a signiWcant interaction between predictability and S–R interval (F(7,49) = 2.94, P < 0.05, p2 = 0.30). Further analysis of this interaction revealed that variability in the UP conditions was higher than variability in the P conditions at S–R intervals of 50, 100, 150, and 200 ms as shown in Fig. 13. These results were similar to those observed in “Experiment 1” for the analysis of V10 variability. It is important to note, however, that for this experiment the mean standard variation of V10 still decreased steadily until the S–R interval of 400 ms, whereas in “Experiment 1” there was no noticeable decline for this measure after 200 ms (see Fig. 7). This observation might be explained by the simultaneous demands of the task used in “Experiment 2” (see “Task and apparatus”) which might have made the task more diYcult to be accomplished than in “Experiment 1”.

General discussion

Fig. 11 Histograms of V10 for responses made by six participants (who showed at least two distinguishable peaks of V10 in the UP block) to predictable (a) and unpredictable (b) targets at two S–R intervals (50 and 400 ms). Distributions were Wtted with a smooth line using a cosine weight function. The histograms show responses required to both targets (near and far). Vertical gray lines indicate the most likely value of peak velocity for near and far target in the P block

The main objectives of the two experiments reported here were to determine the time course of amplitude speciWcation in a rapid interceptive task and identify its mode of speciWcation (continuous or discrete). In relation to the time course of amplitude speciWcation, the results were consistent with previous research (Favilla 1996, 1997, 2002; Ghez et al. 1997, 1990). When the targets were precued the

123

286

Exp Brain Res (2008) 188:275–288

Fig. 12 Histograms of V10 for responses made by three representative participants to predictable (a) and unpredictable (b) at the shortest S–R interval. Distributions were Wtted with a smooth line using a cosine weight function. The histograms show responses required to both targets (near and far). Vertical gray lines indicate the most likely value of peak velocity for near and far target in the P block

Fig. 13 Standard deviation of V10 for the diVerent experimental conditions in “Experiment 2”. The symbols mark the signiWcant contrasts between P and UP conditions (9P < 0.05; *P < 0.01)

responses showed smaller variability in both the temporal error to hit the target and the speed of the movements regardless of the S–R interval at which the indicational stimulus was presented. This result indicates that participants took advantage of precues to prepare their responses in advance, which explains why there is no decrement in accuracy at short S–R intervals (<150 ms). In contrast, the accuracy of the responses in the UP block improved gradually as the S–R interval increased, but did not reach the level achieved in P trials until 200 ms. These Wndings were replicated by the results of the second experiment in which stationary targets were employed. Additionally, the second experiment showed that the modiWed version of the timed response paradigm used in “Experiment 1” can reproduce the temporal demands of the original method introduced by Ghez et al. (1990). The comparison between the results of “Experiment 1” and those obtained by Ghez et al. (1990) indicates that our participants achieved a high level of accuracy slightly

123

sooner than the participants in their experiment. According to Ghez et al. (1990), their participants achieved complete speciWcation of amplitude at S–R intervals greater than 200 ms. One possible explanation for the diVerence between the two experiments may be due to the fact that Ghez et al.’s participants were required to reach for three possible targets, whereas in our experiment there were only two targets. Alternatively, since we employed standard deviation of peak velocity and the variable temporal error as our main measures the diVerence between our results and those obtained by Ghez et al. (1990) could be attributed to corrective modiWcations to the ongoing movement. Nonetheless, responses made by the participants in “Experiment 1” (in which there was no instruction to suppress corrections to ongoing movements) showed no visually identiWable inXection points in the speed proWle or signiWcant number of biphasic movements which suggests that online corrections did not play important role in improving accuracy when hitting moving objects. The results of “Experiment 2”, in which the participants were requested to avoid corrections to ongoing movements, showed a minimum time of 250 ms to prepare the correct response. In addition, in “Experiment 2”, the fact that the standard deviation of V10 tends to decrease until the S–R interval of 400 ms suggests that an even longer preparation interval could be beneWcial for performance in the stationary hitting task. It is possible that this improvement in performance (“Experiment 2”), as the S–R interval increases, is related to the task demands in the original timed response paradigm. Whereas in “Experiment 1” the participants were only required to execute their movements with a desired MT, in “Experiment 2” they were required to attend to two simultaneous task demands (to initiate a response in synchrony with the last in a series of four successive tones and prepare hitting movements

Exp Brain Res (2008) 188:275–288

towards targets at two diVerent distances) and this could make the task in “Experiment 2” more diYcult to be accomplished even at intermediate S–R intervals (250– 350 ms). The results of “Experiment 1” showed that the preferred mode of speciWcation adopted by the participants was predominantly discrete. That is, the participants aimed more often towards one target or the other, rather than at a central distance between the two targets at the shortest S–R interval. The discrete mode of speciWcation was also evident in “Experiment 2”, though two participants showed a very characteristic continuous mode of speciWcation. For these participants, the distribution of peak velocities was unimodal at the shortest S–R interval (50 ms) which indicates the participants were consistently aiming at a central default distance between the targets, and as the S–R interval increased this central tendency bias became progressively more distinct. This, together with the observation that some participants had signiWcant numbers of responses between the two targets (as evidenced by more than two peaks of V10 in the histograms), indicates that the distance ratio (1:2.6) employed in our experiments is near to a threshold between the discrete and the continuous modes of speciWcation. Ghez et al.’s (1997) results showed that a continuous mode of speciWcation was evident most of the time for distance ratios of 1:3 and 1:6, which corresponded to distances from the starting position of 3.2 and 9.6 cm, and 3.2 and 19.2 cm, respectively. However, for distance ratios of 1:12 and 1:24, which corresponded to 3.2 and 38.4 cm, and 1.6 and 38.4 cm, respectively, the mode of speciWcation was mostly discrete. The mode of speciWcation therefore seems to be not only dependent on the distance ratio or the magnitude of separation between targets but also on the nature and requirements of the task. The Wnding that the participants tended to aim more frequently towards the near target at shortest S–R interval (50 ms) suggests that they prepared their responses not only based on their expectations about the next target but also on the assumption that it is strategically more eYcient to prepare for the near target because the time required to employ a correctional movement requiring acceleration is shorter than the time necessary to make a correctional movement involving deceleration (Carlton and Carlton 1987; Quinn and Sherwood 1983). That is, making a correctional movement to the far target (when Wrst aiming at the near) is more likely to succeed than making a correctional movement to the near target (when Wrst aiming at the far). This tendency to direct more responses towards the near target was evident in both experiments despite the fact that the participants were instructed not to make corrections to their movements in “Experiment 2” and therefore should not be expected to use strategies that would facilitate feedback mechanisms.

287

In summary, the results of the two experiments reported here indicated that brief interceptive actions can be very precise even with preparation intervals as short as 200 ms before movement onset. The Wnding that an improvement in the accuracy of peak velocity was accompanied by higher temporal precision to hit the targets indicates that more time to prepare the oncoming movement can improve performance in the rapid interception of moving objects. This is consistent with the assumption of the program model (Tresilian 2005; Tresilian and Plooy 2006a) in the sense that the accuracy of the movement is dependent on the amount of time available to prepare the movement during the Wnal stage of motor preparation (see Fig. 1). The observation of responses towards the unspeciWed target at short S–R intervals (·100 ms) suggests that before any relevant information can be used to prepare the correct movement, the participants prepare their responses based on their expectations. If the responses could be either quickly (re)programmed within short S–R intervals and/or modiWed during execution, it would be expected that the participants would use these strategies to improve the accuracy to hit the moving targets (“Experiment 1”) and as a result no responses should be made towards the unspeciWed targets. Thus, we can conclude that not only the participants prepared their responses in advance of target information but also that feedback mechanisms were not employed eYciently to improve performance in the task. Acknowledgments This research was supported in part by a CAPES (Postgraduate Federal Agency/Brazilian Government) doctoral scholarship to Welber Marinovic and a grant from the Australian Research Council awarded to J. R. Tresilian and A. Plooy.

References Abernethy B (1990a) Anticipation in squash: diVerences in advance cue utilization between expert and novice players. J Sports Sci 8:17–34 Abernethy B (1990b) Expertise, visual search, and information pick-up in squash. Perception 19:63–77 Bootsma RJ, Vanwieringen PCW (1990) Timing an attacking forehand drive in table tennis. J Exp Psychol Hum Percept Perform 16:21–29 Brenner E, Smeets JBJ, de Lussanet MHE (1998) Hitting moving targets—continuous control of the acceleration of the hand on the basis of the target’s velocity. Exp Brain Res 122:467–474 Caljouw SR, van der Kamp J, Savelsbergh GJP (2004) Timing of goaldirected hitting: impact requirements change the informationmovement coupling. Exp Brain Res 155:135–144 Carlton LG, Carlton MJ (1987) Response amendment latencies during discrete arm movements. J Mot Behav 19:227–239 Chambers JM, Cleveland WS, Kleiner B, Tukey PA (1983) Graphical methods for data analysis. Wadsworth, Bellmont de Lussanet MHE, Smeets JBJ, Brenner E (2001) The eVect of expectations on hitting moving targets: inXuence of the preceding target’s speed. Exp Brain Res 137:246–248 de Lussanet MHE, Smeets JBJ, Brenner E (2002) The relation between task history and movement strategy. Behav Brain Res 129:51–59

123

288 Dessing JC, Bullock D, Peper CE, Beek PJ (2002) Prospective control of manual interceptive actions: comparative simulations of extant and new model constructs. Neural Netw 15:163–179 Elliott D, Helsen WF, Chua R (2001) A century later: Woodworth’s (1899) two-component model of goal-directed aiming. Psychol Bull 127:342–357 Favilla M (1996) Reaching movements: Programming time course is independent of choice number. Neuroreport 7:2629–2634 Favilla M (1997) Reaching movements: concurrency of continuous and discrete programming. Neuroreport 8:3973–3977 Favilla M (2002) Reaching movements: mode of motor programming inXuences programming time by itself. Exp Brain Res 144:414–418 Ghez C, Favilla M, Ghilardi MF, Gordon J, Bermejo R, Pullman S (1997) Discrete and continuous planning of hand movements and isometric force trajectories. Exp Brain Res 115:217–233 Ghez C, Hening W, Favilla M (1990) Parallel interacting channels in the initiation and speciWcation of motor response features. In: Jeannerod M (ed) Attention and Performance, vol XVIII. Hillsdale, Erlbaum, pp 265–293 Gordon J, Ghez C (1987) Trajectory control in targeted force impulses.2. Pulse-height control. Exp Brain Res 67:241–252 Gordon J, Ghilardi MF, Ghez C (1994) Accuracy of planar reaching movements.1. Independence of direction and extent variability. Exp Brain Res 99:97–111 Gray R (2002a) Behavior of college baseball players in a virtual batting task. J Exp Psychol Hum Percept Perform 28:1131–1148 Gray R (2002b) “Markov at the bat”: A model of cognitive processing in baseball batters. Psychol Sci 13:542–547 Marinovic W, Iizuka CA, Freudenheim AM (2004) Striking velocity control by table tennis players. Percept Mot Skills 99:1027–1034 Morya E, Ranvaud R, Pinheiro WM (2003) Dynamics of visual feedback in a laboratory simulation of a penalty kick. J Sports Sci 21:87–95 Peper L, Bootsma RJ, Mestre DR, Bakker FC (1994) Catching balls: how to get the hand to the right place at the right time. J Exp Psychol Hum Percept Perform 20:591–612 Quinn JT, Sherwood DE (1983) Time requirements of changes in program and parameter variables in rapid ongoing movements. J Mot Behav 15:163–178

123

Exp Brain Res (2008) 188:275–288 Regan D, Gray R (2000) Visually guided collision avoidance and collision achievement. Trends Cognit Sci 4:99–107 Rosenbaum DA (1980) Human movement initiation: speciWcation of arm, direction, and extent. J Exp Psychol Gen 109:444–474 Saunders JA, Knill DC (2005) Humans use continuous visual feedback from the hand to control both the direction and distance of pointing movements. Exp Brain Res 162:458–473 Savelsbergh GJP, Williams AM, Van der Kamp J, Ward P (2002) Visual search, anticipation and expertise in soccer goalkeepers. J Sports Sci 20:279–287 Schmidt RA, Lee TD (2005) Motor control and learning: a behavioral emphasis. Human Kinetics Publishers, Champaign Schouten JF, Bekker JAM (1967) Reaction time and accuracy. Acta Psychol (Amst) 27:143–153 Teasdale N, Bard C, Fleury M, Young DE, Proteau L (1993) Determining movement onsets from temporal series. J Mot Behav 25:97–106 Tresilian JR (1994) Perceptual and motor processes in interceptive timing. Hum Mov Sci 13:335–373 Tresilian JR (1999) Visually timed action: time-out for ‘tau’? Trends Cognit Sci 3:301–310 Tresilian JR (2005) Hitting a moving target: perception and action in the timing of rapid interceptions. Percept Psychophys 67:129–149 Tresilian JR, Lonergan A (2002) Intercepting a moving target: eVects of temporal precision constraints and movement amplitude. Exp Brain Res 142:193–207 Tresilian JR, Plooy AM (2006a) EVects of acoustic startle stimuli on interceptive action. Neuroscience 142:579–594 Tresilian JR, Plooy AM (2006b) Systematic changes in the duration and precision of interception in response to variation of amplitude and eVector size. Exp Brain Res 171:421–435 Tyldesley DA, Whiting HTA (1975) Operational timing. J Hum Mov Stud 1:172–177 Watts RG, Bahill AT (1990) Keep your eye on the ball: the science and folklore of baseball. Freeman, New York Welsh TN, Elliott D (2004) Movement trajectories in the presence of a distracting stimulus: Evidence for a response activation model of selective reaching. Q J Exp Psychol A 57:1031–1057

The time course of amplitude speciWcation in brief ... - Springer Link

Apr 16, 2008 - This limits the time available to prepare the move- ment. To deal with short preparation intervals, performers are likely to prepare a motor ...

744KB Sizes 4 Downloads 192 Views

Recommend Documents

Time course studies of glucose-induced changes in ... - Springer Link
Time Course Studies of Glucose-Induced Changes in Glucose-6-Phosphate and. Fructose-l ... anaesthesia, in the degree of hypoxia and stress and/or.

Decadal amplitude modulation of two types of ENSO ... - Springer Link
Sep 17, 2011 - defined two climate states—strong and weak ENSO amplitude periods—and separated the characteristics of. ENSO that occurred in both periods. There are two major features in the characteristics of ENSO: the first is the asymmetric sp

Management of Diabetes in Pregnancy - Springer Link
Dec 3, 2011 - profound effects on multiple maternal organ systems. In the fetus, morbidities ... mellitus . Metformin . Glyburide . Pregnancy; diabetes management. Clinical Trial Acronyms. ACHOIS Australian Carbohydrate Intolerance Study in. Pregnant

Rapid communication A time-of-flight spectrometer for ... - Springer Link
Feb 2, 2008 - mination in single hit limit. The electron cloud e from the MCP is collected on the signal wire such that two current pulses u1 and u2 propagate.

The standard model on non-commutative space-time - Springer Link
Jan 25, 2002 - Mills theories on non-commutative space-time has recently been proposed [1–4]. Previously only U(N) gauge theories were under control, and it was thus only possible to con- sider extensions of the standard model. Recently there has b

Visceral regeneration in the crinoid - Springer Link
sic characteristic of life, although it can be lost when their costs are higher than their ... individuals with visceral regeneration in progress [7, 25–28], indicates that the ... In the following stages, the regrowth of the intestinal tract can i

A Linear Time Algorithm for the Minimum-weight ... - Springer Link
For each Bi, 1 ≤ i ≤ l, applying Algorithm II, we can compute a collection of candidate sets. FBi u. = {FBi .... W. H. Freeman and Company, New York, 1979.

A Linear Time Algorithm for the Minimum-weight ... - Springer Link
In this paper, we study the minimum-weight feedback vertex set problem in ...... ISAAC'95 Algorthms and Computations, Lecture Notes in Computer Science,.

Existence of Dyons in the Coupled Georgi–Glashow ... - Springer Link
Feb 4, 2011 - results show that, while the magnetic charge is uniquely determined by the topological monopole number, the electric charge of a solution can be arbitrarily prescribed in an open interval. 1. Introduction. Particle-like static solutions

Finding Equivalent Rewritings in the Presence of ... - Springer Link
of its applications in a wide variety of data management problems, query op- ... The original definition of conjunctive queries does not allow for comparisons.

The role of attention in illusory conjunctions - Springer Link
a major source of support for the feature-integration the- ory since, unlike other ... either within the attended subset or outside it, but not be- tween the attended and ...... Bulletin of the Psychonomic Society, 21, 247-250. POSNER, M. 1. (1980).

Climate and the evolution of serpentine endemism in ... - Springer Link
Nov 12, 2011 - that benign (e.g., high rainfall and less extreme temperatures) ... small populations with novel adaptations, and because competition with non- ...

Monitoring deterioration of vegetation cover in the ... - Springer Link
Mar 19, 2009 - a decrease in near infrared (NIR) and increase in the visible wavelengths ... the fact that soil spectra, lacking these properties, do not show a ...

Use of Patterns for Knowledge Management in the ... - Springer Link
Data Management), cPDm (collaborative Product Definition management) and PLM. (Product ... internal knowledge of the enterprise, and also to customers and stakeholders. In general ... and the patterns created in the chosen software.

Population and distribution of wolf in the world - Springer Link
In addition, 50 wolves live in the forest of the low ar- ... gulates groups (mostly red deer) to live forever. .... wolf, holding a meeting every a certain period, pub-.

The performance of rooks in a cooperative task ... - Springer Link
Received: 12 April 2009 / Revised: 30 October 2009 / Accepted: 6 December 2009 / Published online: 18 December 2009. © Springer-Verlag 2009. Abstract In ...

The Equivalence of Bayes and Causal Rationality in ... - Springer Link
revised definition of rationality given that is not subject to this criticism. .... is nonempty for every w ∈W (i.e. we assume that Bi is serial: (∀w)(∃x)wBix).

Frequency-dependence of the Switching Voltage in ... - Springer Link
Seoul National University of Science and Engineering, Seoul 01811, Korea. I-Wei Chen ... University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA.

The Equivalence of Bayes and Causal Rationality in ... - Springer Link
Page 3 .... is expressed formally by the own-choice knowledge condition: (OK) For all i and for all w,x∈W, if wBix then ... In order to introducing trembles, we would need to make ... be inter-connected: my opponents' choices given that I play si.

Automatic imitation of the arm kinematic profile in ... - Springer Link
Andrea Gaggioli2,3. Published online: 31 July 2015 ... human–robot interaction. Keywords Social neuroscience 4 Joint action 4 Automatic imitation 4 Kinematic ...

The challenge of ethical behavior in organizations - Springer Link
Ronatd R. Sims is Associate Professor in the School of Business. Administration at the ... and management training and development, and organizational transitions. His articles ... While these challenges must all be met by organiza- tions and manager

The Incredible Economics of Geoengineering - Springer Link
Dec 6, 2007 - As I shall explain in this paper, its future application seems more likely than not. ... because the incentives for countries to experiment with ...