Cognitive Daily

Perceiving motion creates a fascinating problem for psychologists. Physicists for centuries have devised a whole set of rules describing how objects actually move. These rules are so precise and accurate that it’s tempting to say that the human perceptual system simply integrates them into motion we see, so that our mental representation of what we see is identical to what’s actually going on in the world.

Some research, such as this article we reported on last month, supports that notion. Since we expect objects to keep on moving (the physical principle of momentum), then our representation of an object presented onscreen that suddenly disappears will continue to move. If we’re tricked into thinking an object is moving faster, our representation takes even longer to stop.

But other explanations for this error are also possible. Scott Jordan and Günther Knoblich, for example, call this same phenomenon by a different name: “displacement of the perceived vanishing point.” The traditional name for the effect, “representational momentum,” may have a bit more of a ring to it, but Jordan and Knoblich have a point: are we really talking about people having the physical concept of momentum in their heads, or is this phenomenon more related to where we’re planning for an object to go, regardless of the physical principles involved?

To test this notion that control is more important than physics, they devised a simple experiment. Participants would use buttons to control a small dot appearing on screen. Pressing one button caused it to accelerate to the left, and the other button cause acceleration in the opposite direction. By pressing the button repeatedly, they could move it quickly from side to side. Their task was to move it back and forth from one edge of the screen to the other, as rapidly as possible. Then suddenly the dot would disappear, and the task would change: now participants had to identify the spot where the dot disappeared. As expected based on the research of Hubbard (and pioneered by Jennifer Freyd and Ronald Finke), participants made consistent errors—on average, they thought the dot had disappeared farther along its path than it really had.

But what would happen if participants were given more—or less—control over the object? If Jordan and Knoblich are right in their explanation of the phenomenon, when people have more control over the object, the size of the error should change based on what they are planning to do next. So Jordan and Knoblich set up their experiment to change the way the objects were controlled. First, they varied the amount each button-click accelerated the object: for half of the trials, button-clicks had a low impact, and for the other half, they had a high impact. Since high-impact clicks impacted the motion of the dot to a greater degree, users felt they had more control.

To give participants less control, some of them were placed in pairs. Instead of using both buttons, they were only allowed to use one—for example, one partner might only use the left key and the other only use the right key. Sometimes the dot disappeared while the participant was in control, actively pressing his or her button, but sometimes it disappeared while the partner was in control. These paired participants also had high-impact and low-impact conditions.

The key to the experiment was this: regardless of the condition: paired, individual, low-impact, or high-impact, the task was designed so that the dots disappeared only when participants were trying to slow them down, and they always disappeared while traveling at the exact same velocity. With these variables eliminated, the results reflect only the level of control each participant had. Here’s a chart summarizing the results:


Since participants were always trying to slow the dots down, we would expect their memory errors to be smaller when they have more control—and indeed, this is borne out by the results. The biggest errors occurred when the object was under the control of the partner and the button-click impact was low. Individual control led to the smallest errors (though, interestingly, the when the paired participants were in control, on the high-impact condition, their results were indistinguishable from the full-time individuals), and the paired individuals, when they were in control, were in the middle.

From a physics perspective, each of these objects had the same momentum when it disappeared, yet the phenomenon that Hubbard, Freyd, and many others call “representational momentum” varied. So clearly, whatever is going on inside of our heads isn’t exactly the same as classical physics. This makes a certain amount of sense, because to make accurate predictions about the world, we need to not only incorporate the laws of physics, but also the intentions of ourselves and others—and physics says nothing about that.

Jordan, J.S., & Knoblich, G. (2004), Spatial perception and control. Psychonomic Bulletin and Review, 11(1), 54-59.


  1. #1 Jesse
    August 23, 2005

    So this prediction that we use with moving objects is kind of like all of the other predictions we use in life. We predict that the orange we will eat tastes like the orange we did eat. And we are usually right with these kind of predictions. What happens in our minds when we are not right with a prediction. ie and orange tastes like an apple. Do we have to rethink everything in the world? or do we just say that it is a once in a lifetime thing, or what?

  2. #2 Jesse
    August 24, 2005

    Our perception of motion seems quite similar to our perception of time, where we only experience one instance at a time, but we predict the future acording to the past.

  3. #3 Dave Munger
    August 25, 2005


    There have indeed been observations of a similar phenomenon to “representational momentum” using other senses, such as sound. We’ll misremember a higher pitch when we’ve been listening to a scale going up, for example.

    Remember, however, that what’s interesting about this phenomenon is that people are being asked not to predict, but they still do. Instead of remembering where the dot really disappears, they remember it as further along its trajectory.

New comments have been disabled.