Editorial: Toward understanding global networks in the brain



    Corresponding author
    1. National Institute of Information and Communications Technology, and Advanced Telecommunications Research Institute International
    • Hiroshi Imamizu, National Institute of Information and Communications Technology (NICT), and Advanced Telecommunications Research Institute International (ATR), Hikaridai, Keihanna Science City, Kyoto 619-0288, Japan. (E-mail: imamizu@gmail.com)

    Search for more papers by this author

Scope of this issue

According to the classical concept of sensorimotor control, perception is considered as the input from the external world, action as the output from the brain to the external world, and cognition as the intermediary process. However, the results of a recent spate of studies suggest that perception, cognition, and action are interrelated and continuously influence each other.

For example, it is well known that high-frequency oscillation of the eye continuously shifts the retinal image, and that if an image is artificially fixed on the retina, it disappears; therefore, eye movement is essential for visual perception (Riggs, Ratliff, Cornsweet, & Cornsweet, 1953). The discovery of mirror neurons (or systems), which are activated both by one's own action and by observation of the same action performed by others (Gallese, Fadiga, Fogassi, & Rizzolatti, 1996), indicated that cognition and action share common neural mechanisms and are interrelated. Neurophysiological studies have indicated that intensive training for the use of a tool (a rake) expands the receptive fields of parietal neurons in monkeys (Iriki, Tanaka, & Iwamura, 1996), thereby suggesting that motor learning can modify the perceptual body schema. According to a recent study, the reason why we are unable to tickle ourselves is that the prediction of sensory feedback derived from a copy of motor commands leads to the cancellation of the ticklish feeling (Blakemore, Wolpert, & Frith, 2000). As stated above, a number of empirical studies have reported phenomena suggestive of interactions between perception, cognition, and action; however, to our knowledge, only a few studies have revealed the detailed mechanisms of integration and interaction of these functions.

In this special issue, we asked for submission for papers that addressed information processing for the integration of and interaction between the mechanisms involved in perception, cognition, and action. Researchers standing in the forefront of various research areas, such as psychophysical, functional neuroimaging, developmental, computational and neuropsychological studies, contributed their original and review articles to this issue, and reported the recent advances in their works and related studies. Herein, I briefly introduce the highlights of the articles.

Highlights of articles

Murakami (2010) investigated the relationship between the characteristics of eye movements in individuals and the mechanisms of stabilizing visual image. As I have mentioned before, our eyes are incessantly in motion even if we attempt to fixate the eyes at a point. However, we do not experience the motion of the retinal image caused by such eye movements because the brain cancels a particular type of visual motion and stabilizes our perceptual world. Murakami and colleagues have developed a theory that the information necessary for visual stabilization is derived from the retinal image itself (visually based stability theory). In the current article, the author has confirmed the validity of the theory by conducting precise experiments on a large number of subjects. The most interesting result relevant to the topic of this special issue is that the eye that makes larger drifts dominates the performance of the motion detection system related to visual stability, even if the eye is not open. This result suggests that the characteristics of individual eye movement affect neural mechanisms associated with motion detection and visual stability through long-term experience in our life, and the perceptual and motor systems interact continuously at subconscious levels.

Yabe and Taga (2010) revealed that long-term experience of locomotion (walking and running) on a treadmill affects motion perception. Their results suggest that people with habitual treadmill experience perceive ambiguous apparent motion as moving in the same direction as that of the treadmill belt even if they are only standing on the treadmill (contextual effect). The authors attribute this phenomenon to habitual linkage between treadmill locomotion and the perception of optic flow caused by belt motion. It is known that a habitual pattern of leg movements in the context of stepping on an escalator causes an odd sensation on a stationary escalator (Fukui, Kimura, Kadota, Shimojo, & Gomi, 2009). The authors have speculated a similar sensory-motor linkage based on the contexts at the subconscious level. It is difficult to quantitatively investigate the experience of using escalators because they are too common in our daily life. However, the experience of locomotion on treadmills can be easily evaluated on the basis of the frequency and duration of exercise performed using treadmill. The interesting results of this study reveal a precise relationship between habitual motor experience and change of perception.

On the basis of extensive research of the literature, Shirai and Yamaguchi (2010) pointed out the uniqueness of development-based change in radial motion perception among other types of motion perception. Fundamental functions of radial motion perception emerge at the age of 3–4 months, but the functions continue to remain underdeveloped until 8 months. From 3 to 8 months, the infants' repertoire of behaviors drastically increases, such as sitting, crawling, standing, and walking, and the patterns of radial motion change according to the behaviors. The authors proposed a hypothesis that the underdeveloped functions enable the infant's brain to cope with the association between various behaviors and radial motion perception. On the basis of the measurement of brain activity of infants, the authors suggest that the posterior parietal region plays an important role in flexible linkage between action and motion perception. Their article suggested that the brain develops as a unitary system for information processing through continuous interactions between action and perception.

Ogawa, Nagai, and Inui (2010) intensively investigated brain mechanisms for copying and tracing figures, both of which are common behaviors in daily life. They have found that copying needs visuomotor transformation (transformation from allocentric to egocentric reference frames), but that tracing does not involve such transformation. On the basis of this simple but important finding, they proposed a computational framework for visuomotor transformation that consistently explains the results of previous developmental studies and those of their own neuropsychological and functional neuroimaging studies. Their framework clearly explains the individual roles of many brain structures. Needless to say, visuomotor transformation is an important function that associates perception with action. Further, they succeeded in drawing a diagram for the information flow during visuomotor transformation in the whole brain network.

Imamizu (2010) reviewed behavioral and functional neuroimaging studies on brain mechanisms for the prediction of sensory feedback from motor commands. Sensory feedback obtained after or during movements is important because it enables us to correct and learn the control of movements. However, it has been pointed out that external feedback based on vision and proprioception is significantly delayed and that this feedback is insufficient for the correction of rapid and smooth movements of humans. Recent studies have indicated that the brain learns to develop neural mechanisms that can internally and immediately predict the sensory feedback from motor commands without relying on external feedback. These mechanisms enable us to not only predict the correction of movements but to also distinguish the perceptual changes caused by our own movements from those caused by the movements of external objects. Therefore, these mechanisms contribute to both adaptive motor control and the stability of our perceived world.

Kanakogi and Itakura (2010) also investigated similar predictive mechanisms in a developmental study, but their focus was on the prediction of other's movements. The discovery of mirror neurons in the monkey brain had a great effect on studies of the interaction of action and cognition, suggesting that neural mechanisms that control one's own movements also contribute to the cognition of others' movements. However, it is unclear whether the control of one's own movements develops before the cognition (prediction) of others' movements, or vice versa. The authors intensively investigated the development of both the motor and cognitive functions in individual infants; their results revealed that infants can predict the goals of others' movements (a target of reaching movement) only after they acquire the ability to control reaching movements. It has been proposed that internal models for predictive control of one's own movements can also be used for the prediction of others' movements and the cognition of others' intentions (Wolpert, Doya, & Kawato, 2003). The article by Kanakogi and Itakura, for the first time, supported this hypothesis with appropriate evidence based on a developmental study.

Tanaka (2010) reviewed the recent advances in computational models of motor control and learning by mainly focusing on the generalization of learning (or adaptation). According to his article, the generalization effect involves learning a task based on the performance in other tasks. Such effects are closely related to cognitive mechanisms, and his review approaches cognitive functions based on the studies of motor control and learning. The issues addressed by Tanaka, such as Bayesian inference and optimal feedback controls, have recently become important concepts in studies of motor control as well as multi-sensory integration.

Features of this issue

The associations between the types of movement and the perceptual/cognitive functions that are investigated or discussed in the abovementioned articles are summarized in Figure 1. The types of movement include eye movements, locomotion (crawling, walking, and running), hand/finger movements for drawing (tracing and copying), and hand/arm movements for reaching and grasping. The perceptual/cognitive functions include various functions ranging from very early sensory processing, such as stabilizing visual image, to high-order cognitive functions, such as inference of others' goals and intentions. The brain regions that may be related to the functions investigated in the articles are roughly shown in Figure 2. The regions include not only cortical regions such as the frontal (premotor regions), parietal, and temporal regions but also the sub-cortical region (cerebellum). Of course, these two figures delineate only a part of the functions and the brain regions related to the interactions between perception, cognition, and action. However, I have tried to highlight the variety of functions and regions associated with these functions through this issue and attempted to illustrate that the interactions occur at various levels, from early sensory processing to high-order cognitive processing. Another important feature of this special issue is the combination of the various methods and approaches (Table 1). Each article is mainly based on psychophysics and behavioral experiments, and most articles combine experiments with other methods and approaches, such as measurement of brain activity, developmental studies, computational modeling, and neuropsychological studies. Hence, recent advances in these methods have enhanced the possibility of elucidating mechanisms involved in the observed phenomena. A combination of multiple methods may also be considered as a general feature of recent cognitive psychology, but it appears particularly effective for the investigation of highly complex mechanisms such as those for the integration and interaction of perception, cognition, and action. The investigation of the mechanisms underlying this integration and interaction can greatly contribute to the understanding of the global networks associated with mental and brain functions that have kept separated while investigating perception, cognition, and action. Although the theme of this issue is not entirely new, I believe that this issue introduces recent advances in the field of Japanese psychology to the world.

Figure 1.

Associations (as indicated by lines) between the types of movement (left side) and the perceptual/cognitive functions (right side) that are investigated or discussed in the articles contained in this special issue. The names above the lines indicate the authors of the articles.

Figure 2.

Brain regions (blobs) that may be related to the functions investigated or discussed in the articles. Line drawing shows the left lateral view of the brain.

Table 1. Approaches and methods addressed in the articles
Psychophysical/ behavioral experimentsMeasurement of brain activityDevelopmental studyComputational modelingNeuro- psychological study
Yabe & Taga    
Shirai & Yamaguchi  
Ogawa, Nagai, & Inui
Kanakogi & Itakura  

I would like to thank Professor Shigemasu, the ex-editor of this journal, for giving me the opportunity to organize this issue, the staff of the Japanese Psychological Association office for their continuous support with editorial work, and the authors for contributing their valuable articles to this issue.