Seeing Your Error Alters My Pointing: Observing Systematic Pointing Errors Induces Sensori-Motor After-Effects

Roberta Ronchi; Patrice Revol; Masahiro Katayama; Yves Rossetti; Alessandro Farnè

doi:10.1371/journal.pone.0021070

. 2011 Jun 23;6(6):e21070. doi: 10.1371/journal.pone.0021070

Seeing Your Error Alters My Pointing: Observing Systematic Pointing Errors Induces Sensori-Motor After-Effects

Roberta Ronchi ^1,^2,^6,^7,^*, Patrice Revol ^1,³, Masahiro Katayama ^4,⁵, Yves Rossetti ^1,^2,^3,^*, Alessandro Farnè ^1,^2,³

Editor: Paul L Gribble⁸

PMCID: PMC3121736 PMID: 21731649

Abstract

During the procedure of prism adaptation, subjects execute pointing movements to visual targets under a lateral optical displacement: As consequence of the discrepancy between visual and proprioceptive inputs, their visuo-motor activity is characterized by pointing errors. The perception of such final errors triggers error-correction processes that eventually result into sensori-motor compensation, opposite to the prismatic displacement (i.e., after-effects). Here we tested whether the mere observation of erroneous pointing movements, similar to those executed during prism adaptation, is sufficient to produce adaptation-like after-effects. Neurotypical participants observed, from a first-person perspective, the examiner's arm making incorrect pointing movements that systematically overshot visual targets location to the right, thus simulating a rightward optical deviation. Three classical after-effect measures (proprioceptive, visual and visual-proprioceptive shift) were recorded before and after first-person's perspective observation of pointing errors. Results showed that mere visual exposure to an arm that systematically points on the right-side of a target (i.e., without error correction) produces a leftward after-effect, which mostly affects the observer's proprioceptive estimation of her body midline. In addition, being exposed to such a constant visual error induced in the observer the illusion “to feel” the seen movement. These findings indicate that it is possible to elicit sensori-motor after-effects by mere observation of movement errors.

Introduction

The human brain has the capacity of quickly learning and adapting to environmental changes. Paradigmatic examples of such plasticity are provided by studies on sensori-motor learning of visuo-motor control under unnatural force-fields and adaptation to prism-induced displacement of the visual-field [1]–[8]. An interesting example of brain reversible plasticity is the one triggered by acting on visual space while wearing prismatic goggles [9], [10]. When a person reaches to visual targets under an optical displacement that induces a lateral (left- or right-ward) shift of the visual scene, her visuo-motor activity is initially characterised by systematic pointing errors in the direction of the optical displacement. If the final pointing error is visible, error correction takes place giving rise to sustained after-effects after prism removal: To their own surprise, subjects produce (compensatory) errors opposite to the direction of the prism deviation [7], [11]–[13]. Adaptation to altered force environments provides another example of sensori-motor plasticity. When a viscous force is experimentally applied to an otherwise voluntarily controlled arm movement, subjects learn to specifically counteract the induced force-field. Thus, the initially major action perturbation is progressively reduced by specific compensatory corrections, exerted in response to the rules acting in the novel environment [14], [15].

In the classical literature subjects develop adaptation or learning in conditions where they perceive their own error signals. Despite a wealth of studies have shown similarities between executing and observing an action (see, for review, [16]), little is known about the after-effects possibly induced by seeing someone else's error signals. Merely observing an action activates the same regions typically implied in the planning and actual execution of the same action, both in monkeys and in humans [17]–[21]. Thus, visually perceiving an action is thought to activate corresponding motor programs [22]. In agreement with this view, sensori-motor learning can occur by simple observation [23]–[25]. In a recent study, Mattar and Gribble [26] elegantly demonstrated that observing a person learning to adapt her reaching movements to a force-field environment facilitates the observer's motor learning when tested later in the same environment. Therefore, action observation is functional to learn new motor patterns and can provide information not only about what an action is for, but also on how to perform it. According to the notion that action observation can evoke an internal representation of the seen movement [27], interference effects have been found when subjects perform an action while observing an incongruent one being performed by somebody else [28]–[30]. Moreover, the same mechanisms are thought to be responsible for detecting one's own as other persons' motor errors (e.g., [31]). Finally, making and observing errors evoke similar error-related negativity on the same brain regions [32].

In the domain of prism adaptation, and consistent with the substantial similarity between executed and imagined actions [33], a pioneering study by Finke [34] showed that if subjects are required to imagine that their (real or imagined) pointing movement ends up producing a systematic error, they subsequently exhibit an after-effect that is opposite in direction to the imagined error. Although Finke's study demonstrated that adding an error to one's own motor imagery may produce prismatic-like after-effects, it remains presently unknown whether passive observation of somebody else's pointing errors would be sufficient to induce after-effects in the observer. Besides its own theoretical relevance as a model of plastic behaviour in the healthy human brain, another main interest for understanding the mechanisms participating (and possibly leading) to prism adaptation owes to their therapeutic implications: Prism adaptation is indeed thought of as one of the most promising techniques for the rehabilitation of left spatial neglect [10], [35]–[40].

Here we tested whether, and to what extent, the mere passive observation of pointing errors made by another person can give rise to after-effects in the observer's sensori-motor behaviour, as well as whether these are akin to those observed following prismatic adaptation. To these aims, we ran two experiments in which healthy participants did not perform any action, but observed another person making incorrect pointing movements. The rationale for the study was that compensatory correction should be elicited by simple observation of erroneous pointing actions by virtue of being, at least partially, processed via the similar processes that are implied in actual prism adaptation [36], [41]. We submitted participants to three tasks typically used in prismatic adaptation studies to assess for the presence of prisms-induced after-effects, by comparing performance before (pre) and after (post) prismatic exposure and calculating the relative shift (S) in these variables. These were, namely, proprioceptive (PS) and visual (VS) estimation of the subjective midline and visual-proprioceptive (VPS) open-loop pointing, which are all considered of as indexes of adaptation [2].

Experiment 1

Materials and Methods

Subjects

Twelve right-handed subjects (6 males, mean age: 26.6, range 22–35) with normal or corrected to normal vision participated in this study. For this, as well as for the following experiment, all participants gave their verbal informed consent to participate in the study, which was approved by the review board of the INSERM U864 ethics committee. All subjects were naive as to the purpose of the study. Only 3 out of 12 participants were aware of the usual effects induced by prisms.

Apparatus

Two custom-made experimental set-ups served different aims: a test box (identical to Rossetti et al, 1998) was used to measure and record the three dependent variables (PS, VS, VPS) before and after exposure to pointing errors; a pointing board was used to visually expose subjects to the pointing errors. The test box was a black wooden frame (30 cm high, 80 cm wide and 80 cm deep) opened on the side facing the subject: the lower horizontal surface was covered by electro-resistive carbon paper. The distal side facing the subject was equipped with a pulley-mounted red LED that could be moved horizontally in front of the subject (see Figure 1). The speed of the LED movement was varied between trials in order to avoid counting strategies. The resistive paper and pulley were calibrated to electronically read out the final finger position on the box lower surface, as signalled by the contacting position of a metal thimble worn on the right index fingertip. The pointing board was a black wooden table (50 cm×100 cm) where two filled dots (red on the left, white on the right; 1 cm diameter) were permanently visible.

Procedure: Pre and Post Tests

Participants sat in front of the test box in a darkened, sound-attenuated booth, with their head aligned with the body's mid-sagittal axis. The head was stabilised by having their chin on a chin-rest situated on the top of the box (see Figure 1). The subjects' right arm was placed in the box, with the right hand resting on a starting location in front of their trunk and a thimble worn on the index finger. Before and after the observation phase, the following measures of egocentric reference frames were taken:

Proprioceptive (P): with eyes closed, subjects made 10 straight-ahead pointing movements on the table surface, to indicate the subjectively estimated position of their body midline;
Visual (V): with eyes open, subjects verbally stopped for 10 times the position of a red LED, moving horizontally in front of them, when it corresponded to their body midline. The LED moved 5 times from the right to the left and 5 times from the left to the right;
Visual-Proprioceptive (VP): with eyes open, subjects made 10 pointing movements on the table surface to indicate the downward projected position of the red LED, which was placed in front of their body midline while they kept their eyes close. The arm movement was occluded from view (open loop pointing) via a wooden panel horizontally positioned on the top of the box.

For each variable, a preliminary examination of the measures revealed a stable performance across the 10 trials. Therefore, the accuracy was computed by the average of the 10 trials per task, measured in degrees of visual angle with respect to the objective body midline (corresponding to 0°): negative numbers indicated leftward-, while positive numbers rightward-located estimates. The difference between post- and pre-exposure measures was also computed to express the relative shift in estimate for each task (Visual Shift: VS, Proprioceptive Shift: PS, Visual-Proprioceptive Shift: VPS).

Procedure: Exposure

Participants sat facing the pointing board where visual targets were located 20° to the left (red) and to the right (white) with respect to their mid-sagittal axis. Both the subject and the experimenter wore a white glove on their right hand: the subject's hand (unseen) was aligned with her mid-sagittal axis and placed on a support under the table with the index finger located beneath the hand of the experimenter (visible). The procedure was inspired from the Japanese traditional sketch comedy called Nininbaori, in which a person A (the ‘experimenter’) sitting behind a person B (the ‘subject’) is trying to feed B with chopsticks (http://www.english-rakugo.com/english_version/english_performance.html). The experimenter stood behind the subject with the right hand placed on the table in front of the subject's body midline, above the unseen subject's hand. The participant wore a pair of goggles fitted with neutral (non-deviating) lenses. During this phase, the subject was required to carefully observe the pointing movements performed towards the targets: the examiner called out the colour of one target (“red!” or “white!”) and right afterwards made a rapid pointing movement (entirely visible) towards it. However, the experimenter's pointing movement was “wrong” in that she voluntarily made an error on the right-side, of 2 cm on average (see Figure 2). The subject observed a total of 60 erroneous pointing movements (30 to the red and 30 to the white target, in a random order). This exposure phase lasted about 3 minutes.

Statistical analysis

An ANOVA was performed on the mean shift for each measure. A separate ANOVA was conducted on the mean pointing positions, in order to compare the subjects' performance in each test before and after the observation phase. Paired t-test analysis (against zero) was additionally performed for each of the three mean shifts.

Results

Figure 3 shows the mean subjects' shift (difference Pre-Post observation) in each test. A repeated measure ANOVA with Shift (PS, VS and VPS) as a within-subject variable was conducted. The main effect of Shift was significant (F (2, 22) = 3.82, p = 0.038) revealing a stronger leftward deviation in PS (−2.29°) as compared to VPS (−0.01°; p<0.05), no other comparison being significant. The VS (−1.55°) tended to deviate towards the left, without differing from the other two measures. Paired t-tests confirmed that PS (t = −2.45, p = 0.032) and VS (t = −2.68, p = 0.022) significantly differed from zero (i.e., the value indicative of the absence of pre-post difference), whereas VPS did not (t = −0.03, p = 0.978).

The number zero indicates the lack of differences between pre- and post-observation, negative numbers the presence of a shift to the left, positive numbers to the right.

Discussion

These results show that observing pointing movements that deviate in their final landing position towards a relative rightward location alters the observer's estimation of the subjective body midline in the leftward (i.e., opposite) direction, both in the visual and proprioceptive modalities. No significant change was observed in the visual-proprioceptive test, also termed “total shift” [42]. It therefore seems that merely observing a rightward pointing bias, similar to the error normally executed during the initial phase of prism adaptation, does alter the subjectively sensed body midline, which is one of the most typical consequences of prism adaptation [43], [44]. Indeed, during prism adaptation, two sources of information are elaborated concerning hand position: proprioception and vision [45], [46]. The mismatch between proprioceptive and visual cues, which results in a final pointing error, is progressively compensated for by the brain and two components are thought to be implied in this process. Initially, the fast error reduction is thought to be a consequence of a strategic compensative behaviour, in which the subject uses the error feedback to voluntarily correct performance. Strategy is not sufficient, however [13], and during successive pointing trials a more automatic kind of compensation becomes established, leading to real adaptation [44]. In this respect, the results of Experiment 1 clearly show that seeing a systematic rightward error (i.e., constant around 2 cm from target) can generate a leftward after-effect. Even though the mere observation of error signals produced by somebody else does not enact exhaustively the consequences of adaptation to visual-field displacement (specifically because the motor commands and the proprioceptive feedback are absent), it seems to induce adaptation-like sensori-motor effects.

To better understand this phenomenon, and more clearly establishing whether adaptation-like after-effects follow mere passive exposure to somebody else's erroneous pointing movements, several aspects need to be further taken into account. A first point is to ascertain whether subjects appreciated the “wrongness” in the pointing movements they observed. In fact, debriefing revealed some of them did not notice anything “odd” or “wrong” because, e.g., “the experimenter pointed towards the correct target” as defined by the spoken colour. Therefore, considering the small amplitude of the naturally variable error, some subjects actually considered the seen movements as being correct. In spite of their wrong belief, they were nevertheless affected by the observation of such a ‘negligible’ error. Second, it is important to consider the potential contribution of the participant's cognitive interpretation of the seen movement, which was performed by another person, but could have induced the illusory feeling that the observer took part in the seen movement, or the moving arm could belong to the observer (i.e., agency and/or ownership illusions, see [47]–[49]). To address these issues, in Experiment 2 we introduced a questionnaire designed to ascertain the impression of 1) ‘wrongness’ about the observed movements and 2) feelings of agency and/or ownership of the seen moving arm.

As a third point, in Experiment 1 the error performed by the examiner was small and constant (except for the natural variability around the intended 2 cm of lateral deviation). This, however, is not what typically occurs during the whole prism exposure phase, when the pointing error is initially large and then decreases with the increasing number of trials. Our choice of a constant error was motivated by the fact that adaptation, as measured via visual, proprioceptive and visual-proprioceptive after-effect measures, is stronger in the initial phase of the exposure period when the error is not yet compensated for, then reaches a plateau when pointing errors are further reduced (e.g., Redding & Wallace [11]). We thus formulated the hypothesis that observing a stable error should maximise the ‘need’ to compensate. By contrast, observing an error reduction should decrease the pressure to modify the incorrect movement, so that an initial rise in compensation pressure should be followed by a wiping out of the nascent consequences. Finally, subjects' previous knowledge about the prisms' properties and the effects of prismatic adaptation could also influence the processes that are put into play during the observation phase and/or the subsequent test phase. To take into account both these additional issues, two observation phases were compared in Experiment 2. Participants observed pointing movements displaying large(r) final errors (6 cm on average) that were either constant or progressively reduced across trials. In addition, while in Experiment 1 subjects with different levels of prisms expertise were tested, in Experiment 2 participants were split in two subgroups: naives and non naives with respect to the effects of prism adaptation.