Home | Search Serendip

This laboratory is intended to illustrate some basic principles of movement control by the brain, and to give you the background and some incentive to further explore them on your own. The general question under investigation is the nature of the information processing steps which intervene between the definition of a task to be achieved and the production of a particular movement intended to achieve that task.

This laboratory, like the preceding one, is deliberately "open ended". It includes clearly specified and linked sets of questions and observations, some questions for which you need to provide the observations, and some openings for you to go off more fully on your own (the italicized sections) You should do enough observations of the first two kinds to "get a feel" for them. You are then free to focus for the remainder of the lab session (and any additional time you want to spend) on any particular set of observations and questions which interest you particularly.

Here again, you are free to choose any part or parts of this laboratory as one of your required writeups, bearing in mind the requirement for clear observations clearly described and rigorously interpreted in light of a well-defined question. Your raw data, in this case, will consist of appropriately documented descriptions of movement accuracy and variability under each of several sets of specified conditions.


In the previous laboratory, you discovered that sensory processing depends not only on input signals but also on movement. In this laboratory, we will be largely concerned with the converse question: for movement, how significant are various sensory signals and what roles do they play? We will explore the relatively simple problem of movements directed toward targets at different locations in space. Such movements involve a highly precise spatio-temporal pattern of contraction in a large number of muscles of the arm, hand, and fingers, and at least in some cases of the body as well. Underlying this is a similarly complicated and precise spatio-temporal pattern of discharge in a still larger number of motoneurons. Though we will not directly monitor the motoneuron discharges, we can (for the movements used here) draw reasonable conclusions about whether they are the same or different under varying experimental conditions simply from observing similarities and differences in the movement itself. The movements obviously depend on two input signals: the command "point to the target" and the discharge of sensory neurons associated with the appearance of the target. What we will be concerned with is whether there are any other neural signals which are important to the creation of the spatio-temporal pattern of motoneuron discharge.

A. Baseline observations

Place two targets on a board at eye level about twenty-five centimeters apart and stand 75 centimeters or so away with a pen in one hand and your arm relaxed at your side. Verify that you can accurately place the tip of the pen on either target. Now, for each target, repeat the same arm movement twenty times, making each movement as rapidly as you can. What you will probably find is that your movements are less accurate than your original ones, with the point actually touched by the tip of the pen being more variable from trial to trial. Both the reduction in accuracy and the increased variability are worth exploring further. What do you think accounts for them? What do they imply about the likely involvement in accurate pointing movements of important signals other than those coming from the task instructions and the retina? Can you develop hypotheses and further observations to test your hypotheses?

B. Focusing on the variability in rapid movements

The task instructions are not varying from movement to movement, making it unlikely this can account for the observed variability in the rapid directed movements you made. An alternate possibility is a variation in the retinal signal. There was, in your baseline observations, no requirement to hold your eyes steady and so the retinal location activated may have varied from trial to trial. Repeat your observations being certain that you are looking directly at the target on each trial. Is the variability significantly reduced? You can get a still better handle on this question (why?) by deliberately increasing the variability of the retinal signal and seeing if there corresponding increases in the movement variability. Select several spots within 2 or 3 cm of the target and look at one or another of these prior to each movement. Does this increase movement variability? Do the same for some more widely spaced spots. What can you conclude about whether variability in the retinal signal causes movement variability?

Another possible source of movement variability is the starting position of your arm and hand. Information about this is provided to the brain both by your eyes and by a large number of specialized sensory neurons (proprioceptors) providing information about joint angles as well as muscle lengths and forces. You can explore the possibility that variability in these pathways determines variability in the movements using the same logic as used to look at the role of variation in the retinal signal. Design and carry out an appropriate series of observations. What conclusions can you draw about whether variation in visual and or proprioceptive signals about arm position accounts for the observed movement variability?

How do these observations bear on the question of whether there are signals other than those associated with the task instructions and the retina which contribute to creating the motoneuron discharge patterns underlying rapid movements? Can you come up with further hypotheses as to the origin, identity, and significance of such signals? You might want, for example, to explore the question of what happens to the variability over blocks of trials repeatedly made to the same target from the same starting position.

C. The accuracy problem

As observed at the outset, slower and more deliberate movements are both more accurate and less variable than rapid movements. This suggests the possibility (why the hedge?) that something in addition to the task instructions and the retinal signal for target location is important for accurate pointing. You might want to repeat the slow, deliberate pointing several times to see whether you can develop any intutions about what this might be. Can you detect any differences in accuracy and variability between earlier and later phases of deliberate movements?

The simple difference in the time to movement completion might help to explain differences in accuracy between deliberate and rapid movements. During the former, there is more time during the movement itself for the nervous system to collect and process information about the progress and likely success of the movement. One potential source of such information is, of course, the eye, which can continuously convey during a movement information about the relative locations of hand and target. This information can be removed by the simple expedient of closing your eyes after you look at the target and before making the movement. Make twenty movements this way. Is the accuracy reduced and the variability increased? Can you draw any conclusions at this point about whether signals other than the two basic ones must be involved in the accurate movements? About what these signals must consist of and where they originate? Any hypotheses? We'll return to this shortly.

D. Directed movement: the questions of feedback dependence and additional signals

That the eye can continuously provide information about the relation between target and hand location raises the possibility that directed movements are entirely based on visual feedback, that is, that the underlying motoneuron discharge pattern is actually created by a continuous process of comparing retinal information about target and hand location and creating for each comparison motoneuron discharges to reduce the difference between the two. Its worth taking the time to prove that directed movements can occur without visual feedback, both for its own sake and because it provides a way to show the involvement in directed movements of some additional signals.

Return to the situation of two targets spaced about 25 centimeters apart, and your arms at your side. Look at the first target, close your eyes, and then mark its location with the pen. Now look at the second, close your eyes, and fairly quickly mark its location with the pen. Repeat nineteen more times for each target. Are you capable of pointing with enough accuracy to distinguish between your responses to targets 25 centimeters apart? Observations like this one suggest that the motneuron discharge pattern underlying many movements are stored in the nervous system where they can be set off by the appropriate stimuli rather than being constantly produced by sensory input. This concept, termed "central pattern generation" has a number of interesting implications. What more would you have to do to prove that central pattern generation exists?

You might want to further explore the question of how close targets can be and still be discriminated by movements made in the absence of visual feedback. Another interesting question is the relation between accuracy for fast movements and movements made in the absence of visual feedback. Are they the same or different? Does this suggest any conclusions, or raise any additional questions? You might also want to see whether there is any difference in rapid as opposed to slow movements made in the absence of visual feedback.

The situation of two targets in the absence of visual feedback also makes possible some strong conclusions about what signals other than those coming from the task instruction and the retina contribute to the output pattern. What retinal region was activated when you were looking at the first spot? The second? What does this imply? Now try pointing, with your eyes closed, to each of the two targets from a variety of different inital arm positions. How does the pointing accuracy in this situation relate to that in others you have explored? Does this generate any new hypotheses, or yield any new conclusions? What additional signal does this prove is important for generating relatively accurate pointing movements?

E. Back to the accurate pointing problem

Our observations to this point suggest that relatively accurate although variable pointing can occur in the absence of visual feedback, but involves at least two signals (perhaps three) in addition to those associated with the task command and the image of the target on the retina. What is now worth exploring is the relation between accurate pointing, and the variable but relatively accurate pointing. Are they entirely distinct kinds of movements, or can the latter help in some way to better understand the former?

With your eyes open, make a series of fairly rapid movements toward the target but stopping a few centimeters away. For each, mark on the paper where the pen would have touched had the movement been continued. How does the accuracy and variability of these movements relate to those with your eyes closed? To rapid movements with your eyes open? What does this suggest about how deliberate movements become less variable and more accurate? Are they less variable and more accurate throughout the course of the movement or is this a function only of later phases of the movement?

Clearly some kind of a comparison between actual and intended hand position is going on during the late phases of movements to make them more accurate. Such corrections might be based on being able to see how close the hand is to the target, or might be based on some kind of stored representation of target location. One way to explore this is by comparing the variability and accuracy of slow and rapid second movements made with the eyes closed after an initial rapid movement of the kind just studied. Greater accuracy and less variability with slower movements might be taken as evidence that comparison with a stored target location is responsible for movement adjustments, particularly if you see some sign of corrections in the movement trajectories themselves. Another question you might want to explore is how target information is stored. One possibility is that it is represented by the direction in which your eyes are looking (even though they are closed). Is your accuracy reduced by altering your eye position? your head position? Is seeing your arm position required for the late corrective comparison? What happens if you change your arm position after closing your eyes?

F. The source and nature of additional signals

Information about eye position is clearly important for the directed movements we have been studying. Such information might reasonably be expected to come from proprioceptors in the muscles of your eyes. The following experiment is aimed at trying to verify this presumption.

The apparent location of the target shifts if the location of the image of the target is made to change on the retina. Such a change can be produced by looking at a target with one eye closed and gently pushing at the corner of the eyelid of your open eye with your finger. This alters the direction in which you are looking, and should produce an apparent downward movement of the target by several centimeters. You know from previous work that you can point accurately at a target at which you are not directly looking. Would you expect to be able to point accurately toward the target with your eye remaining open during the movement? Why or why not? Test your prediction. Now try pointing with your eye closed during the movement. Repeat this five or ten times so you have a record on the board of where you actually pointed. In this situation, proprioceptors in the eye are presumably indicating a change in eye position exactly equal and opposite to the retinal displacement of the target image from the fovea, so you should have again been pointing reasonably accurately (but with some variability since the eyes are closed) toward the target. Were you? What does this suggest about the source of the eye position signal involved in pointing?

Signals which can provide information about the location of parts of the body but do not have their origins in proprioceptive or other pathways originating in sensory neurons are called "corollary discharge" or "efference copy" signals. Can you imagine how they could come into existence? You might want to see whether you can design experiments to test for their existence in relation to arm position in visually triggered movements. Another area you might look into is the effect of introducing unexpected perturbations during movement, or creating barriers to particular movement trajectories. You might also want to explore the generality of some of your findings by extending them to studies of movements directed toward tactile or auditory rather than visual targets.


The general concern of this laboratory session was to better understand what must be going on in the brain to yield movements which achieve particular objectives. Critically review the observations you have made, and consider the extent to which they do or do not yield clear conclusions. What would you look for if you could make more direction observations on the nervous system itself? Would you expect this to verify your conclusions? alter them? extend them? What lessons from these observations would you expect to generalize to other kinds of movements? What implications do these observations have for movement control generally? For behavior generally? How could these be further explored?

| Forum | Brain and Behavior Link | Serendip Home |

Send us your comments at Serendip

© by Serendip 1994- - Last Modified: Wednesday, 04-Sep-1996 08:02:55 EDT