Multiple representations and mechanisms for visuomotor adaptation in young children

Abstract

In this study, we utilized transformed spatial mappings to perturb visuomotor integration in 5-yr-old children and adults. The participants were asked to perform pointing movements under five different conditions of visuomotor rotation (from 0° to 180°), which were designed to reveal explicit vs. implicit representations as well as the mechanisms underlying the visual-motor mapping. Several tests allowed us to separately evaluate sensorimotor (i.e., the dynamic dimension of movement) and cognitive (i.e., the explicit representations of target position and the strategies used by the participants) representations of visuo-proprioceptive distortion. Our results indicate that children do not establish representations in the same manner as adults and that children exhibit multiple visuomotor representations. Sensorimotor representations were relatively precise, presumably due to the recovery of proprioceptive information and efferent copy. Furthermore, a bidirectional mechanism was used to re-map visual and motor spaces. In contrast, cognitive representations were supplied with visual information and followed a unidirectional visual-motor mapping. Therefore, it appears that sensorimotor mechanisms develop before the use of explicit strategies during development, and young children showed impaired visuomotor adaptation when confronted with large distortions.

Introduction

Studies analyzing the movements of subjects in perturbed visuo-proprioceptive environments enable us to understand the mechanisms whereby the human visuomotor system adapts to new situations during life. In these studies, the mental transformation of visual information regarding hand and target position is necessary to achieve efficient levels of motor performance on the part of the participants. Paillard (2004) proposed that this transformation can occur on several different processing levels, leading to the dissociation of unconscious sensorimotor representations of movement from cognitive (or perceptual) representations.

It has been assumed that these two mechanisms of spatial-information processing (sensorimotor and cognitive) are represented by the dissociation between the dorsal and ventral streams (Milner & Goodale, 2008). The ventral stream refers to “vision for perception” (i.e., mental representations that can reach conscious awareness), whereas the dorsal stream refers to “vision for action” (i.e., processes that do not reach conscious awareness). This dissociation has also been described as the difference between knowing “where is the target” vs. knowing “how to get there”, or in other words, positional cues vs. movement execution (Paillard, 1991). However, this dissociation can lead to a dual representation of movement when subjects are confronted with visuomotor rotational paradigms (Boy, Palluel-Germain, Orliaguet, & Coello, 2005). In this study, adult participants were asked to perform pointing movements based on visual information that had been rotated 45° relative to actual arm trajectory, and two tests were administered to evaluate movement dimensions. (1) The Spatial Evaluation test (SE test), which evaluated the spatial dimensions of the response (i.e., “where is the target”). During this test participants were asked to point to their initial hand location and to trace the direction of the movements that they performed. (2) The Movement Reproduction test (MR test), on the other hand, was used to evaluate the dynamic dimensions of the response (i.e., “how to get there”). Subjects were asked to reach for the target presented during the adaptation phase. The results from this study indicated the existence of a dissociation between visuomotor representations - visual information appeared important during the evaluation of the spatial dimension of the response but not during its dynamic dimension.

When confronted with repeated visuo-proprioceptive distortions, adaptation takes place, eventually leading to a resumption of normal levels of performance (i.e., without visuomotor rotation). In addition, Krakauer (2009) showed that when adult participants were informed of a 45° counter-clockwise visual rotation and given a cognitive strategy to counter it, they could not overcome the visuomotor distortion and became progressively worse at hitting the target. Therefore, Krakauer proposed that rotational learning is implicit and is not dependent on explicit strategies. Therefore, the sensorimotor level of information processing appears to be most important in this kind of adaptation.

However, the role of cognitive representations might be more important during childhood than during adulthood. As early as 7 yrs of age, children exhibit a dissociation between cognitive and sensorimotor representations of movement (Rival et al., 2004, Rival et al., 2003). These studies showed that children confronted with Müller-Lyer or Duncker illusions made erroneous perceptual judgments but performed pointing movements toward the illusion accurately (i.e., they had precise sensorimotor representations). When a visual rotational paradigm was used, an effect of children’s cognitive level on adaptation was observed. For example, the behavior of children 4-6 yrs of age was perturbed by spatial transformations, such as a vertical display instead of an aligned display, which induced more planning variability in children than in adults (Bo, Contreras-Vidal, Kagerer, & Clark, 2006). Moreover, when confronted with a 180° rotation of a visual scene, 5-yr-old children were more highly affected than either older children or adults, and they showed difficulty integrating visual and proprioceptive information (Ferrel-Chapus, Hay, Olivier, Bard, & Fleury, 2002). This difference in adaptation may result from the fact that 5-yr-old children do not re-align map visual and motor spaces in the same way that adults do. Ferrel, Bard, and Fleury (2001) showed that 5-yr-old children increased the amplitude of errors as visual rotations were increased from 0-180°, which suggests, according to Cunningham (1989), the use of a unidirectional mechanism to realign visual and motor mapping. In contrast, when subjected to visual distortion, adults showed increasing errors from 0-90°, although this decreased from 90-180°. This finding reveals that movement direction in adults is coded via a set of bidirectional axes (i.e., a bidirectional visual-motor mapping), and these large amplitude adaptations imply that adults inverse the polarity of the axes (equal to 180° rotation) and then perform a backward shifts toward smaller angles (Abeele & Bock, 2001). Ferrel et al. (2001) proposed that 5-yr-old children re-map visual and motor spaces differently than adults because of the immaturity of their ability to perform mental rotations. Therefore, young children would be predicted to be unable to utilize efficient cognitive strategies to overcome spatial perturbations.

The present experiments were designed to determine whether the sensorimotor and cognitive levels of action representation differed between young children and adults. We tested the assumption that 5-yr-old children do not use the same sensory information when evaluating the spatial and dynamic dimensions of pointing movements. Participants were subjected to two tests that were based on the protocols reported by Boy et al. (2005) and to two new tests. The first additional test was designed to assess the influence of movement cues on spatial representations. The Spatial Evaluation test (SE test) used by Boy et al. (2005) required participants to reproduce previously performed movements, which could also rely on the “how” system. Thus, in our study, participants were asked, in a third test, to indicate the position of the target without reproducing previously performed movements, a protocol similar to that reported by Paillard (1999) and Paillard, Michel, and Stelmach (1983). A fourth test was performed to evaluate the ability of the children to use explicit strategies to encompass visuomotor perturbation. Krakauer (2009) showed that when adult participants are informed of (or have a perceived) visuo-proprioceptive distortion, they voluntarily aim away from visual targets in order to ensure that the cursor will touch the desired target. We therefore wanted to evaluate whether children have perceived visual distortions and use explicit strategies to overcome them, as adults sometimes do. We hypothesized that sensorimotor representations (assayed using the MR test, which evaluates “how” to reach a target) would be similar in adults and children because they are implicit and develop during early childhood (Rival et al., 2003, Rival et al., 2004). However, according to Ferrel et al. (2001), cognitive representations should differ between adults and children because of differences in the visual-motor mapping that is used, as well as the immaturity of children’s ability to perform mental rotations (Estes, 1998). Moreover, children have difficulties representing explicit spatial distortions, which would be predicted to impair their ability to adapt to visual-motor distortion.