READING NOTES - Chapter 8
Perceiving Motion
Functions of Motion Perception
The experience of L.M. and the few other people with akinetopsia makes it clear that being unable to perceive motion is a great handicap
But looking closely at what motion perception does for us reveals a long list of functions
Detecting Things
Detection is at the head of the list because of its importance for survival
We need to detect things that might be dangerous in order to avoid them
Perceiving Objects
Movement of an observer around an object can have a similar effect:
Viewing the “horse” in Figure 8.2b from different perspectives reveals that its shape is not exactly what you may have expected based on your initial view
Thus, our own motion relative to objects is constantly adding to the information we have about those objects, and, most relevant to this chapter, we receive similar information when objects move relative to us
Observers perceive shapes more rapidly and accurately when an object is moving
Movement also serves an organizing function, which groups smaller elements into larger units
The motion of individual birds becomes perceived as the larger unit of the flock, in which the birds are flying in synchrony with each other
Perceiving Events
As we observe what’s going on around us, we typically observe ongoing behavior as a sequence of events
This description, which represents only a small fraction of what is happening in the coffee shop, is a sequence of events unfolding in time
And just as we can segment a static scene into individual objects, we can segment ongoing behavior into a sequence of events, where an event is defined as a segment of time at a particular location that is perceived by observers to have a beginning and an end
The point in time when each of these events ends and the next one begins is called an event boundary
The connection of events to motion perception becomes obvious when we consider that event boundaries are often associated with changes in the nature of motion
One pattern of motion occurs when placing the order, another when reaching out for the coffee cup, and so on
Jeffrey Zacks and coworkers (2009) measured the connection between events and motion perception by having participants watch films of common activities such as paying bills or washing dishes and asking them to press a button when they believe one unit of meaningful activity ended and another began
Social Perception
Interactions with other people involve movement at many levels
L.M.’s akinetopsia made it difficult for her to interact with people, because she couldn’t tell who was talking by seeing their lips moving
On a larger scale, we use movement cues to determine a person’s intentions
An experiment by Atesh Koul and coworkers (2019) showed that the speed and timing of the movement can help answer this type of question
Other experiments have shown that the characteristics of movement can be used to interpret emotions (Melzer et al., 2019)
Although assigning social motives to moving geometrical objects makes a great story, the real story occurs as we interact with people in social situations
Many social cues are available in person-to-person interactions, including facial expressions, language, tone of voice, eye contact, and posture, but research has shown that movement can provide social information even when these other cues aren’t available
For example, Laurie Centelles and coworkers (2013) used a way of presenting human motion called point-light walkers, which are created by placing small lights on people’s joints and then filming the patterns created by these lights when people move
The observers were able to indicate whether the two people were interacting with each other or were acting independently
Taking Action
Navigating ourselves through the environment and walking down a crowded sidewalk are examples of how our own movement depends on our perception of movement
We perceive the stationary scene moving past us as we walk down the sidewalk
We pay attention to other people’s movements to avoid colliding with them
Movement perception is also crucially involved in sports—both watching, as you follow a double-play unfold in baseball or watch the trajectory of a long pass in football, or as you participate yourself
Studying Motion Perception
When Do We Perceive Motion?
We perceive motion when something moves across our field of view, which is an example of real motion
Perceiving a car driving by, people walking, or a bug scurrying across a tabletop are all examples of the perception of real motion
However, there are three types of illusory motion—the perception of the motion of stimuli that aren’t actually moving
Apparent motion is the most famous and most studied type of illusory motion
Induced motion occurs when motion of one object (usually a large one) causes a nearby stationary object (usually smaller) to appear to move
Motion aftereffects occur when viewing a moving stimulus causes a stationary stimulus to appear to move
One example of a motion aftereffect is the waterfall illusion
Researchers studying motion perception have investigated all the types of perceived motion described above—and a number of others as well
Our purpose, however, is not to understand every type of motion perception but to understand some of the principles governing motion perception in general
Comparing Real and Apparent Motion
For many years, researchers treated the apparent motion created by flashing stationary objects or pictures and the real motion created by actual motion through space as though they were separate phenomena, governed by different mechanisms
However, there is ample evidence that these two types of motion have much in common
Because of the similarities between the neural responses to real and apparent motion, researchers study both types of motion together and concentrate on discovering general mechanisms that apply to both
The Ecological Approach to Motion Perception
Gibson’s approach (1950, 1966, 1979), involves looking for information in the environment that is useful for perception
This information for perception, according to Gibson, is located not on the retina but “out there” in the environment
He thought about information in the environment in terms of the optic array—the structure created by the surfaces, textures, and contours of the environment—and he focused on how movement of the observer causes changes in the optic array
When Jeremy walks from left to right and Maria follows him with her eyes, portions of the optic array become covered as he walks by and then are uncovered as he moves on
This result is called a local disturbance in the optic array
This local disturbance in the optic array, which occurs when Jeremy moves relative to the environment, covering and uncovering the stationary background, causes Maria to perceive Jeremy’s movement, even though his image is stationary on her retina
The fact that everything moves at once in response to movement of the observer’s eyes or body is called global optic flow; this signals that the environment is stationary and that the observer is moving, either by moving their body or by scanning with their eyes, as in this example
Thus, according to Gibson, motion is perceived when one part of the visual scene moves relative to the rest of scene, and no motion is perceived when the entire field moves, or remains stationary
The Corollary Discharge and Motion Perception
Gibson’s approach focuses on information that is “out there” in the environment
Another approach to explaining the movement situations in Figure 8.8 is to consider the neural signals that travel from the eye to the brain
This brings us back to the corollary discharge signal to explain why we don’t see the scene blur when we move our eyes from place to place when scanning a scene
As we noted in Chapter 6, corollary discharge theory distinguishes three signals:
1. the image displacement signal, which occurs when an image moves across the retina
2. the motor signal, which is sent from the motor area to the eye muscles to cause the eye to move
3. the corollary discharge signal, which is a copy of the motor signal
According to corollary discharge theory, movement will be perceived if a brain structure called the comparator (actually a number of brain structures) receives just one signal—either the image displacement signal or the corollary discharge signal
This situation has also been approached physiologically in another way, by focusing on how the moving image stimulates one retinal receptor after the other
We will describe this approach by first considering a neural circuit called the Reichardt detector
The Reichardt Detector
The Reichardt detector circuit consists of two neurons, A and B, which send their signals to an output unit that compares the signals it receives from neurons A and B
The key to the operation of this circuit is the delay unit that slows down the signals from A as they travel toward the output unit
In addition, the output unit has an important property:
It multiplies the responses from A and B to create the movement signal that results in the perception of motion
If the timing is right, the delayed signal from A (record 3) reaches the output unit just when the signal from B (record 2) arrives
Because the output unit multiplies the responses from A and B, a large movement signal results (record 4)
Thus, when Jeremy moves from left to right at the right speed, a movement signal occurs and Maria perceives Jeremy’s movement
More complicated versions of this circuit, which have been discovered in amphibians, rodents, primates, and humans (Borst & Egelhaaf, 1989), create directionally sensitive neurons, which fire only to a particular direction of motion
Single-Neuron Responses to Motion
The Reichardt detector is a neural circuit that creates a neuron that responds to movement in a specific direction
Such directionally-selective neurons were recorded from neurons in the rabbit’s retina by Horace Barlow and coworkers (1964) and from neurons in the cat’s visual cortex by David Hubel and Torsten Wiesel (1959, 1965)
We will focus on the middle temporal (MT) area, which contains many directionally selective neurons
Evidence that the MT cortex is specialized for processing information about motion comes from experiments that have used moving dot displays in which the direction of motion of individual dots can be varied
Experiments Using Moving Dot Displays
William Newsome and coworkers (1995) used the term coherence to indicate the degree to which the dots move in the same direction
Newsome and coworkers used these moving dot stimuli to determine the relationship between
1. a monkey’s ability to judge the direction in which dots were moving
2. the response of a neuron in the monkey’s MT cortex
They found that as the dots’ coherence increased, two things happened:
1. the monkey judged the direction of motion more accurately
2. the MT neuron fired more vigorously
The monkey’s behavior and the firing of the MT neurons were so closely related that the researchers could predict one from the other
Newsome’s experiment demonstrates a relationship between the monkey’s perception of motion and neural firing in its MT cortex
This relationship has also been demonstrated
1. by lesioning (destroying) or deactivating some or all of the MT cortex
2. by electrically stimulating neurons in the MT cortex
Lesioning the MT Cortex
A monkey with an intact MT cortex can begin detecting the direction dots are moving when coherence is as low as 1 or 2 percent
However, after the MT is lesioned, the coherence must be 10 to 20 percent before monkeys can begin detecting the direction of motion
Deactivating the MT Cortex
Further evidence linking neurons in MT cortex to motion perception has been determined from experiments on human participants using a method called transcranial magnetic stimulation (TMS) that temporarily disrupts the normal functioning of neurons
Stimulating the MT Cortex
The link between the MT cortex and motion perception has been studied not only by disrupting normal neural activity, but also by enhancing it using a technique called microstimulation:
Achieved by lowering a small wire electrode into the cortex and passing a weak electrical charge through the tip of the electrode
When researchers applied TMS to the MT cortex, participants had difficulty determining the direction in which a random pattern of dots was moving
Kenneth Britten and coworkers (1992) used this procedure in an experiment in which a monkey was looking at dots moving in a particular direction while indicating the direction of motion it was perceiving
In addition to the MT cortex, another area highly involved in motion perception is the nearby medial superior temporal (MST) area
The MST area is involved in eye movements, so it is particularly important in localizing a moving object in space
Beyond Single-Neuron Responses to Motion
As important as these studies are, just showing that a particular neuron responds to motion does not explain how we perceive motion in real life
We can appreciate why this is so by considering how motion signaled by single neurons is ambiguous and can differ from what we perceive
We are going to focus on the pole, which is essentially a vertical bar
The ellipse represents the area of the receptive field of a neuron in the cortex that responds when a vertical bar moves to the right across the neuron’s receptive field
We know this because we can see the woman and the flag moving up
But the neuron, which only sees movement through the narrow view of its receptive field, only receives information about the rightward movement (red arrows)
This is called the aperture problem, because the neuron’s receptive field is functioning like an aperture, which reveals only a small portion of the scene
The Aperture Problem
If you were able to focus only on what was happening inside the aperture, you probably noticed that the direction that the front edge of the pencil was moving appeared the same whether the pencil was moving
a. horizontally to the right
b. up and to the right
In both cases, the front edge of the pencil moves across the aperture horizontally, as indicated by the red arrow
Solutions to the Aperture Problem
There are at least two solutions to the aperture problem (Bruno & Bertamini, 2015)
The first was highlighted by one of my students who tried the pencil demonstration in Figure 8.16
He noticed that when he followed the directions for the demonstration, the edge of the pencil did appear to be moving horizontally across the aperture, whether the pencil was moving horizontally or up at an angle
The second solution is to pool, or combine, responses from a number of neurons
Evidence for pooling comes from studies in which the activity of neurons in the monkey’s MT cortex is recorded while the monkey looks at moving oriented lines like the pole or our pencil
What all of this means is that the “simple” situation of an object moving across the visual field as an observer looks straight ahead is not so simple because of the aperture problem
The visual system apparently solves this problem
1. by using information from neurons in the striate cortex that respond to the movement of the ends of objects
2. by using information from neurons in the MT cortex that pool the responses of a number of directionally selective neurons
Motion and the Human Body
Apparent Motion of the Body
Even though these stimuli are stationary, movement is perceived back and forth between them if they are alternated with the correct timing
Generally, this movement follows a principle called the shortest path constraint—apparent movement tends to occur along the shortest path between two stimuli
When the pictures were alternated less than five times per second, observers began perceiving the motion shown in Figure 8.18c:
The hand appeared to move around the woman’s head
These results are interesting for two reasons:
1. They show that the visual system needs time to process information in order to perceive the movement of complex meaningful stimuli
2. They suggest that there may be something special about the meaning of the stimulus—in this case, the human body—that influences the way movement is perceived
To test the idea that the human body is special, Shiffrar and coworkers showed that when objects such as boards are used as stimuli, the likelihood of perceiving movement along the longer path does not increase at lower rates of alternation, as it does for pictures of humans
To find out, Jennifer Stevens and coworkers (2000) measured brain activation using brain imaging
They found that both movement through the head and movement around the head activated areas in the parietal cortex associated with movement
Biological Motion Studied by Point-Light Walkers
Research using point-light walkers shows that motion of the body creates perceptual organization by causing the movements of the individual dots to become organized into “a person moving.”
When the person wearing the lights is stationary, the lights look like a meaningless pattern
However, as soon as the person starts walking, with arms and legs swinging back and forth and feet moving in flattened arcs, first one leaving the ground and touching down, and then the other, the motion of the lights is immediately perceived as being caused by a walking person
This self-produced motion of a person or other living organism is called biological motion
One reason we are particularly good at perceptually organizing the complex motion of an array of moving dots into the perception of a walking person is that we see biological motion all the time
Emily Grossman and Randolph Blake (2001) provided evidence supporting the idea of a specialized area in the brain for biological motion by measuring observers’ brain activity as they viewed the moving dots created by a point-light walker and as they viewed dots that moved similarly to the point-light walker dots, but were scrambled so they did not result in the impression of a person walking
Motion Responses to Still Pictures
It is not hard to imagine the person moving to a different location immediately after this picture was taken
A situation such as this, in which a still picture depicts an action involving motion, is called implied motion
Jennifer Freyd (1983) conducted an experiment involving implied motion by briefly showing observers pictures that depicted a situation involving motion, such as a person jumping off a low wall
The idea that the motion depicted in a picture tends to continue in the observer’s mind is called representational momentum
If implied motion causes an object to continue moving in a person’s mind, then it would seem reasonable that this continued motion might be reflected by activity in the brain
To determine whether implied motion stimuli would have the same effect, Winawer had his participants observe a series of pictures showing implied motion
The key result of this experiment was that before observing the implied-motion stimuli, participants were equally likely to perceive dot stimuli with zero coherence (all the dots moving in random directions) as moving to the left or to the right