I practice martial arts (Aikido), which involves a lot of rotation.

When practicing, I find that holding an abstract image in my head, instead of thinking about the actual technique, improves my performance of the technique.

For example, when rotation is required, my mental image is essentially a vertical spiral pulled tight in the middle (as if constrained by a belt), in that one comes in to the axis of rotation and must stay on it before moving off into the rest of the technique.

The question, then is:
What explains why an abstract mental image can improve adherence to an ideal set of movements?

And, can this be explained neuro-biologically and, if yes, how?

(I suspect that the two questions have similar, if not the same, answers - hence why I've included both instead of separating them)

If you prefer to not answer those questions, believe that I should be answering these myself, or believe that this is not high-enough quality for CogSci (I apologize for any improper terminology), please offer a direction in which I can start my research into this phenomena, as at the moment I am at a loss for where to begin.

  • 2
    $\begingroup$ I think this is an interesting question as is, but if you want to get a head start with it, searching on terms like "internal representation" and "intrinsic coordinates" should get you some hits for a general overview. $\endgroup$ Nov 1, 2012 at 20:09
  • $\begingroup$ Thanks! I've found some interesting material related to motion-path planning and execution, so I shall pursue that until I reach more questions I can't answer. $\endgroup$
    – BenCole
    Nov 5, 2012 at 20:00

3 Answers 3


I would hazard a guess that some type of motor system interference is taking place when you visualize the movement versus visualizing an abstract shape.

For example, Kilner et al. (2003) found that actions that are observed can interfere with incongruous executed actions. They had subjects make arm movements that were either similar or dissimilar to those of another human or a robot. Movements that were not similar caused a greater amount of error in the arm movement of the subject. Since motor visualization can recruit many of the same areas of cortex as actual observation and execution (Kosslyn et al., 2001), this could cause similar interference.

Other papers, such as Rémy et al. (2010), have shown that there is reduced motor learning when two movements were acquired at the same time. Their hypothesis was that the two movements both recruited similar areas of cortex, which caused a form of competitive interference. Visualizing a movement before execution could generate competition between similar, but not identical, movement representations.

Finally, there is evidence that varied practice, or learning a motor movement in several different ways, can cause a short-term degradation of motor learning, but can cause long term increases in retention and generalization of a motor movement. Magill and Hall (1990) had a nice review of this effect, called contextual interference. Visualizing the movement could be causing reduced performance now, but it could be argued that it may be beneficial in the long-term.

Wikipedia has some surprisingly good articles on motor learning and task interference if you are interested in learning more:
Motor learning: http://en.wikipedia.org/wiki/Motor_learning
Interference theory: http://en.wikipedia.org/wiki/Interference_theory

  • Kilner JM, Paulignan Y, Blakemore SJ. (2003) An interference effect of observed biological movement on action. Current biology, 13(6), 522–525.
  • Kosslyn SM, Ganis G, and Thomson WL. (2001) Neural foundations of imagery. Nature Reviews Neuroscience, 2, 635-642.
  • Magill RA and Hall KG (1990) A review of the contextual interference effect in motor skill acquisition. Human Movement Science, 9, 241-289.

There was an interesting study that compared practice by imagining a movement to practice by executing ot and found an increase in strength of ~35% for imagining it, compared to ~55% for actually doing it; additionally, they found an increase in cortical potentials corresponding to the increase in strength. They concluded that mental imagery alone strengthened cortical representations for that movement.

Vinoth K. Ranganathan, Vlodek Siemionow, Jing Z. Liu, Vinod Sahgal, Guang H. Yue From mental power to muscle power—gaining strength by using the mind Neuropsychologia, Volume 42, Issue 7, 2004, Pages 944–956


Maybe this paper would shed some light on your question : 5-HT and motor control: a hypothesis. The author has done some experiments on Serotonin systems within a cat's brain (Raphe Nuclei), and argues that Serotonin system is primarily involved in locomotion and (voluntary?) movement.

Here's an interesting diagram from that paper: the firing pattern of the raphe nuclei changes in the cat's brain, as the cat shifts attention from grooming to observing and trying to figure out the source of noise in the room (door opening and closing). The author explores the connection between motor and sensory processes, but you'll have to read the paper to get a better hang of what he suggests.

The cat experiment was done invasively, and I don't know if there are experiments like this on humans, where a human can be asked to shift attention, while the firing pattern of Raphe Nuclei is observed. Still, my best guess would be that by visualizing as you perform the rotation, you alter how your raphe nuclei is firing.

enter image description here

  • $\begingroup$ I don't think this has anything to do with what he's asking about, actually. Those are single-unit studies in the brainstem in a descending pathway, what the OP is referring to is more of a cortical representation problem. $\endgroup$ Nov 1, 2012 at 20:49
  • 1
    $\begingroup$ Thank you for the correction. Cortical representation is about control of voluntary movements. As a student of control theory, in mechanical controls, there's usually some controllers and a motor. The controllers are at different levels of abstraction. Are the two systems we mention (DRN and Cortical representations) just different kinds of controllers, or are they completely separate? $\endgroup$
    – Alex Stone
    Nov 2, 2012 at 5:14
  • 1
    $\begingroup$ Well, I wish I had a good answer for that. It's quite rare in the brain for anything to act in isolation, there's almost always some reciprocal connections and collaterals, so possibly thinking of it as sub-components of one big system is probably the most prudent, but leaves it as an intractable problem for the time being. It's hard to know where to draw the line. $\endgroup$ Nov 2, 2012 at 13:12
  • $\begingroup$ @ChuckSherrington - I've come to the same conclusion: my understanding is, at this point, not sufficient to evaluate any answer to the point of being able to spot shortcomings in partially-correct answers, so I'm going to leave this question open for now. $\endgroup$
    – BenCole
    Nov 5, 2012 at 19:57
  • $\begingroup$ @AlexStone - I like this answer, a lot, but it doesn't quite address what I'm looking for, which would involve and explanation integrating the motor cortex and motion-path planning vs execution, in conjunction with visualization. To my (very) limited knowledge no such study exists in its own right, and as I mention to Chuck, my understanding is currently not to the point where I could evaluate such an answer anyway. $\endgroup$
    – BenCole
    Nov 5, 2012 at 19:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.