Visual Imagery

10
Imagery in the History of Psychology Removing Part of the Visual Cortex Decreases Image Size
Early Ideas About Imagery Problems with Perceiving Are Accompanied by Problems
with Imagery
Imagery and the Cognitive Revolution
Dissociations Between Imagery and Perception
➤➤ Method: Paired-Associate Learning Conclusions from the Imagery Debate

Imagery and Perception: Do They Share the Same Using Imagery to Improve Memory
Mechanisms?
Placing Images at Locations
Kosslyn’s Mental Scanning Experiments
➤➤ Demonstration: Method of Loci
➤➤ Method/Demonstration: Mental Scanning
Associating Images with Words
The Imagery Debate: Is Imagery Spatial or Propositional?
Comparing Imagery and Perception Something to Consider: Individual
Size in the Visual Field
Differences in Visual Imagery
Interactions of Imagery and Perception ➤➤ TEST YOURSELF 10.2
➤➤ TEST YOURSELF 10.1

Imagery and the Brain

Imagery Neurons in the Human Brain
CHAPTER SUMMARY
➤➤ Method: Recording From Single Neurons in Humans
THINK ABOUT IT
Brain Imaging
Multivoxel Pattern Analysis KEY TERMS

Transcranial Magnetic Stimulation COGLAB EXPERIMENTS

Neuropsychological Case Studies

297

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 297 4/18/18 5:09 PM

 SOME QUESTIONS WE
WILL CONSIDER A ➤➤
nswer the following questions:

How many windows are there in front of the house or apartment where you live?
◗◗ How do “pictures in your head”
created by imagining an object ➤➤ How is the furniture arranged in your bedroom?
compare to the experience you ➤➤ Are an elephant’s ears rounded or pointy?
have when you see the actual ➤➤ Is the green of grass darker or lighter than the green of a pine tree?
object? (300, 312)
How did you go about answering these questions? If you experienced visual images
◗◗ How does damage to the brain
when answering these questions, you were experiencing visual imagery—seeing in the
affect the ability to form visual
images? (310) absence of a visual stimulus (Hegarty, 2010). Mental imagery, a broader term that refers to
◗◗ How can we use visual imagery the ability to re-create the sensory world in the absence of physical stimuli, is used to include
to improve memory? (312) all of the senses. People have the ability to imagine tastes, smells, and tactile experiences.
◗◗ How do people differ in their Most people can imagine melodies of familiar songs in their head, so it is not surprising that
ability to create visual images?
musicians often report strong auditory imagery and that the ability to imagine melodies has
(314)
played an important role in musical composition. Paul McCartney says that the song “Yes-
terday” came to him as a mental image when he woke up with the tune in his head. Another
example of auditory imagery is orchestra conductors’ using a technique called the “inner
audition” to practice without their orchestras by imagining a musical score in their minds.
When they do this, they imagine not only the sounds of the various instruments but also
their locations relative to the podium.

Just as auditory imagery has played an important role in the creative process of music, visual
imagery has resulted in both scientific insights and practical applications. One of the most
famous accounts of how visual imagery led to scientific discovery is the story related by
the 19th-century German chemist Friedrich August Kekule. Kekule said that the structure
of benzene came to him in a dream in which he saw a writhing chain that formed a circle
that resembled a snake, with its head swallowing its tail. This visual image gave Kekule the
insight that the carbon atoms that make up the benzene molecule are arranged in a ring.
A more recent example of visual imagery leading to scientific discovery is Albert
Einstein’s description of how he developed the theory of relativity by imagining himself
traveling beside a beam of light (Intons-Peterson, 1993). On a more athletic level, many
competitors at the Olympics use mental imagery to visualize downhill ski runs, snowboard-
ing moves, bobsled turns, and speedskating races (Clarey, 2014).
One message of these examples is that imagery provides a way of thinking that adds
another dimension to the verbal techniques usually associated with thinking. But what is
most important about imagery is that it is associated not just with discoveries by famous
people but also with most people’s everyday experience. In this chapter, we will focus on
visual imagery, because most of the research on imagery has been on this type of imagery.
We will describe the basic characteristics of visual imagery and how visual imagery relates
to other cognitive processes such as thinking, memory, and perception. This connection
between imagery and cognition in general is an important theme in the history of
psychology, beginning in the early days of scientific psychology in the 19th century.

298

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 298 4/18/18 5:09 PM

 Imagery in the History of Psychology   299

Imagery in the History of Psychology

We can trace the history of imagery back to the first laboratory of psychology, founded by
Wilhelm Wundt (see Chapter 1, page 7).

Early Ideas About Imagery

Wundt proposed that images were one of the three basic elements of consciousness, along
with sensations and feelings. He also proposed that because images accompany thought,
studying images was a way of studying thinking. This idea of a link between imagery and
thinking gave rise to the imageless thought debate, with some psychologists taking up
Aristotle’s idea that “thought is impossible without an image” and others contending that
thinking can occur without images.
Evidence supporting the idea that imagery was not required for thinking was Francis
Galton’s (1883) observation that people who had great difficulty forming visual images were
still quite capable of thinking (also see Richardson, 1994, for more modern accounts of imagery
differences between people). Other arguments both for and against the idea that images are
necessary for thinking were proposed in the late 1800s and early 1900s, but these arguments
and counterarguments ended when behaviorism toppled imagery from its central place in
psychology (Watson, 1913; see Chapter 1, page 10). The behaviorists branded the study
of imagery as unproductive because visual images are invisible to everyone except the person
experiencing them. The founder of behaviorism, John Watson, described images as “unproven”
and “mythological” (1928) and therefore not worthy of study. The dominance of behav­ior­­ism
from the 1920s through the 1950s pushed the study of imagery out of mainstream psychology.
However, this situation changed when the study of cognition was reborn in the 1950s.

Imagery and the Cognitive Revolution

The history of cognitive psychology that we described in Chapter 1 recounts events in the
1950s and 1960s that came to be known as the cognitive revolution. One of the keys to
the success of this “revolution” was that cognitive psychologists developed ways to measure
behavior that could be used to infer cognitive processes. One example of a method that
linked behavior and cognition is Alan Paivio’s (1963) work on memory. Paivio showed that
it was easier to remember concrete nouns, like truck or tree, that can be imaged than it is to
remember abstract nouns, like truth or justice, that are difficult to image. One technique
Paivio used was paired-associate learning.

METHOD Paired-Associate Learning

In a paired-associate learning experiment, participants are presented with pairs of
words, like boat–hat or car–house, during a study period. They are then presented,
during the test period, with the first word from each pair. Their task is to recall the word
that was paired with it during the study period. Thus, if they were presented with the
word boat, the correct response would be hat.

To explain the finding that memory for pairs of concrete nouns is better than memory
for pairs of abstract nouns, Paivio (1963, 1965) proposed the conceptual peg hypothe-
sis. According to this hypothesis, concrete nouns create images that other words can “hang
onto.” For example, if presenting the pair boat–hat creates an image of a boat, then present-
ing the word boat later will bring back the boat image, which provides a number of places
on which participants can place the hat in their mind (see Paivio, 2006, for an updating of
his ideas about memory).

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 299 4/18/18 5:09 PM

 300  CHAPTER 10 Visual Imagery

Whereas Paivio inferred cognitive processes by measuring memory,

Roger Shepard and Jacqueline Metzler (1971) inferred cognitive processes by
using mental chronometry, determining the amount of time needed to carry
out various cognitive tasks. In Shepard and Metzler’s experiment, which we
described in Chapter 5 (see page 146), participants saw pictures like the ones
in Figure 10.1. Their task was to indicate, as rapidly as possible, whether the
two pictures were of the same object or of different objects. This experiment
showed that the time it took to decide that two views were of the same object
was directly related to how different the angles were between the two views (see
➤ Figure 10.1 Stimuli for Shepard and Metzler’s Figure 5.14, page 147). This result was interpreted as showing that participants
(1971) mental rotation experiment. were mentally rotating one of the views to see whether it matched the other one.
(Source: Shepard & Metzler, 1971) What was important about this experiment was that it was one of the first to
apply quantitative methods to the study of imagery and to suggest that imagery
and perception may share the same mechanisms. (References to “mechanisms” include both
mental mechanisms, such as ways of manipulating perceptual and mental images in the
mind, and brain mechanisms, such as which structures are involved in creating perceptual
and mental images.)
We will now describe research that illustrates similarities between imagery and percep-
tion, and also the possibility that there is a basic difference between how imagery and per-
ception are represented in the mind. As we will see, these comparisons have involved a large
number of behavioral and physiological experiments, which demonstrate both similarities
and differences between imagery and perception.

I magery and Perception: Do They Share

the Same Mechanisms?
The idea that imagery and perception may share the same mechanisms is based on the obser-
vation that although mental images differ from perception in that they are not as vivid or long
lasting, imagery shares many properties with perception. Shepard and Metzler’s results showed
that mental and perceptual images both involve spatial representation of the stimulus. That
is, the spatial experience for both imagery and perception matches the layout of the actual
stimulus. This idea, that there is a spatial correspondence between imagery and perception,
is supported by a number of experiments by Stephen Kosslyn involving a task called mental
scanning, in which participants create mental images and then scan them in their minds.

Kosslyn’s Mental Scanning Experiments

Stephen Kosslyn has done enough research on imagery to fill three books (Kosslyn, 1980,
1994; Kosslyn et al., 2006), and he has proposed some influential theories of imagery based
on parallels between imagery and perception. In one of his early experiments, Kosslyn (1973)
asked participants to memorize a picture of an object, such as the boat in Figure 10.2, and
then to create an image of that object in their mind and to focus
on one part of the boat, such as the anchor. They were then
asked to look for another part of the boat, such as the motor,
and to press the “true” button when they found this part or the
“false” button when they couldn’t find it.
Kosslyn reasoned that if imagery, like perception, is
spatial, then it should take longer for participants to find parts
that are located farther from the initial point of focus because
they would be scanning across the image of the object. This is
➤ Figure 10.2 Stimulus for Kosslyn’s (1973) image-scanning actually what happened, and Kosslyn took this as evidence for
experiment. the spatial nature of imagery. But, as often happens in science,

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 300 4/18/18 5:09 PM

 Imagery and Perception: Do They Share the Same Mechanisms?   301

another researcher proposed a different explanation. Glen Lea (1975) proposed that as par-
ticipants scanned, they may have encountered other interesting parts, such as the cabin, and
this distraction may have increased their reaction time.
To answer this concern, Kosslyn and coworkers (1978) did another scanning experi-
ment, this time asking participants to scan between two places on a map. Before reading
about Kosslyn’s experiment, try the following demonstration.

M E T H O D / D E M O N S T R AT I O N Mental Scanning
Erie
Imagine a map of your state that includes three locations: the place where
you live, a town that is far away, and another town that is closer but does not
fall on a straight line connecting your location and the far town. For exam-
ple, for my state, I imagine Pittsburgh, the place where I am now; Philadel- PENNSYLVANIA
phia, all the way across the state (contrary to some people’s idea, Pittsburgh
is not a suburb of Philadelphia!); and Erie, which is closer than Philadelphia Pittsburgh
but not in the same direction (Figure 10.3). Philadelphia
Your task is to create a mental image of your state and, starting at your
location, to form an image of a black speck moving along a straight line
between your location and the closer town. Be aware of about how long it ➤ Figure 10.3 Example of a state map for the
took to arrive at this town. Then repeat the same procedure for the far town, Mental Scanning method/demonstration.
Use your own state for this method/
again noting about how long it took to arrive.
demonstration.

Kosslyn’s participants used the same procedure as you did for the demonstration but
were told to imagine an island, like the one in Figure 10.4a, that contained seven different
locations. By having participants scan between every possible pair of locations (a total of
21 trips), Kosslyn determined the relationship between reaction time and distance shown
in Figure 10.4b. Just as in the boat experiment, it took longer to scan between greater dis-
tances on the image, a result that supports the idea that visual imagery is spatial in nature.
As convincing as Kosslyn’s results were, however, Zenon Pylyshyn (1973) proposed another

2.1

1.9

1.7
Reaction time (sec)

1.5

1.3

1.1

0.9

0 2 4 6 8 10 12 14 16 18
(a) (b) Distance (cm)

➤ Figure 10.4 (a) Island used in Kosslyn et al.’s (1978) image-scanning experiment. Participants mentally traveled
between various locations on the island. (b) Results of the island experiment.
(Source: Kosslyn, Ball, & Reiser, 1978)

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 301 4/18/18 5:09 PM

 302  CHAPTER 10 Visual Imagery

explanation, which started what has been called the imagery debate—a debate about
whether imagery is based on spatial mechanisms, such as those involved in perception, or
on mechanisms related to language, called propositional mechanisms.

The Imagery Debate: Is Imagery Spatial or Propositional?

Much of the research we have described so far in this book is about determining the nature
of the mental representations that lie behind different cognitive experiences. For example,
when we considered short-term memory (STM) in Chapter 5, we presented evidence that
information in STM is often represented in auditory form, as when you rehearse a tele-
phone number you have just looked up.
Kosslyn interpreted the results of his research on imagery as
supporting the idea that the mechanism responsible for imagery
involves spatial representations—representations in which dif-
ferent parts of an image can be described as corresponding to
specific locations in space. But Pylyshyn (1973) disagreed, saying
“The cat is that just because we experience imagery as spatial, that doesn’t mean
under the table”
that the underlying representation is spatial. After all, one thing
that is clear from research in cognitive psychology is that we often
aren’t aware of what is going on in our mind. The spatial experi-
ence of mental images, argues Pylyshyn, is an epiphenomenon—
something that accompanies the real mechanism but is not
actually part of the mechanism.
Propositional Spatial, or depictive, Pylyshyn proposed that, rather than the spatial representa-
representation representation
tions suggested by Kosslyn, the mechanism underlying imagery
involves propositional representations—representations in
➤ Figure 10.5 Propositional and spatial, or depictive, which relationships can be represented by abstract symbols, such
representations of “The cat is under the table.” as an equation, or a statement, such as “The cat is under the table.”
In contrast, a spatial representation would involve a spatial layout showing the cat and the
table that could be represented in a picture (Figure 10.5). Spatial representations such as
the picture of the cat under the table in which parts of the representation correspond to
parts of the object are called depictive representations.
We can understand the propositional approach better by returning to the depictive rep-
resentation of Kosslyn’s boat in Figure 10.2. Figure 10.6 shows how the visual appearance

REAR DECK (behind) CABIN (behind) FRONT DECK

(atta
of )

(sid
f)
ar

to

ched
(re

eo
ron

f)
pf

to)
(to

MOTOR

WINDSHIELD PORTHOLE ANCHOR

of )

(fro
m

n
tto

t of
(bo

Motor to porthole: 3 links

PROPELLER HANDLE Motor to anchor: 4 links

➤ Figure 10.6 How the visual appearance of the boat in Figure 10.2 can be represented propositionally.
Paths between motor and porthole (dashed line) and motor and anchor (dotted line) indicate the
number of nodes that would be traversed between these parts of the boat.
(Source: Kosslyn et al., 1995)

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 302 4/18/18 5:09 PM

 Imagery and Perception: Do They Share the Same Mechanisms?   303

of this boat can be represented propositionally. The words indicate parts of the boat, the
length of the lines indicate the distances between the parts, and the words in parentheses
indicate the spatial relations between the parts. A representation such as this would predict
that when starting at the motor, it should take longer to scan and find the anchor than to
find the porthole because it is necessary to travel across three links to get to the porthole
(dashed line) and four links to get to the anchor (dotted line). This kind of explanation
proposes that imagery operates in a way similar to the semantic networks we described in
Chapter 9 (see page 276).
We’ve discussed both the spatial and the propositional approaches to imagery because
these two explanations provide an excellent example of how data can be interpreted in dif-
ferent ways. (See Chapter 7, Something to Consider: Alternative Explanations in Cognitive
Psychology, page 220.) Pylyshyn’s criticisms stimulated a large number of experiments that
have taught us a great deal about the nature of visual imagery (also see Intons-Peterson,
1983). However, after many years of discussion and experimentation, the weight of the evi-
dence supports the idea that imagery is served by a spatial mechanism and that it shares
mechanisms with perception. We will now look at additional evidence that supports the
idea of spatial representation.

Comparing Imagery and Perception

We begin by describing another experiment by Kosslyn. This one looks at how imagery is
affected by the size of an object in a person’s visual field.
Size in the Visual Field If you observe an automobile from far away, it fills only a portion
of your visual field, and it is difficult to see small details such as the door handle. As you
move closer, it fills more of your visual field, and you can perceive details like the door han-
dle more easily (Figure 10.7). With these observations about perception in mind, Kosslyn
wondered whether this relationship between viewing distance and the ability to perceive
details also occurs for mental images.
To answer this question, Kosslyn (1978) asked participants to imagine two animals,
such as an elephant and a rabbit, next to each other and to imagine that they were standing
close enough so that the larger animal filled most of their visual field (Figure 10.8a). He
then posed questions such as “Does the rabbit have whiskers?” and asked his participants
to find that part of the animal in their mental image and to answer as quickly as possible.
When he repeated this procedure but told participants to imagine a rabbit and a fly next to
each other, participants created larger images of the rabbit, as shown in Figure 10.8b. The
result of these experiments, shown alongside the pictures, was that participants answered
questions about the rabbit more rapidly when it filled more of the visual field.

View from afar Move closer

➤ Figure 10.7 Moving closer to an object, such as this car, has two effects: (1) The object fills more of
your visual field, and (2) details are easier to see.

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 303 4/18/18 5:09 PM

 304  CHAPTER 10 Visual Imagery

In addition to asking participants to respond to details in

visual images, Kosslyn also asked them to do a mental walk task,
in which they were to imagine that they were walking toward
their mental image of an animal. Their task was to estimate how
RT = 2,020 ms far away they were from the animal when they began to experi-
ence “overflow”—when the image filled the visual field or when
its edges started becoming fuzzy. The result was that participants
had to move closer for small animals (less than a foot away for
a mouse) than for larger animals (about 11 feet away for an ele-
phant), just as they would have to do if they were walking toward
(a) actual animals. This result provides further evidence for the idea
that images are spatial, just like perception.
Interactions of Imagery and Perception Another way to
demonstrate connections between imagery and perception is to
RT = 1,870 ms
show that they interact with one another. The basic rationale be-
hind this approach is that if imagery affects perception, or per-
ception affects imagery, this means that imagery and perception
both have access to the same mechanisms.
The classic demonstration of interaction between percep-
tion and imagery dates back to 1910, when Cheves Perky did
(b) the experiment pictured in Figure 10.9. Perky asked her partici-
pants to “project” visual images of common objects onto a screen
➤ Figure 10.8 These pictures represent images that Kosslyn’s and then to describe these images. Unbeknownst to the partici-
(1978) participants created, which filled different portions of pants, Perky was back-projecting a very dim image of this object
their visual field. (a) Imagine elephant and rabbit, so elephant onto the screen. Thus, when participants were asked to create
fills the field. (b) Imagine rabbit and fly, so rabbit fills the an image of a banana, Perky projected a dim image of a banana
field. Reaction times indicate how long it took participants to
onto the screen. Interestingly, the participants’ descriptions of
answer questions about the rabbit.
their images matched the images that Perky was projecting. For
example, they described the banana as being oriented vertically,
just as was the projected image. Even more interesting, not one
of Perky’s 24 participants noticed that there was an actual picture
on the screen. They had apparently mistaken an actual picture for
a mental image.
Modern researchers have replicated Perky’s result (see Craver-
Lemley & Reeves, 1992; Segal & Fusella, 1970) and have demon-
strated interactions between perception and imagery in a number
of other ways. Martha Farah (1985) instructed her participants
to imagine either the letter H or the letter T on a screen (Figure
10.10a). Once they had formed a clear image on the screen, they
pressed a button that caused two squares to flash, one after the
other (Figure 10.10b). One of the squares contained a target
letter, which was either an H or a T. The participants’ task
was to indicate whether the letter was in the first square or the
second one. The results, shown in Figure 10.11c, indicate that
the target letter was detected more accurately when the partici-
➤ Figure 10.9 Participant in Perky’s (1910) experiment. pant had been imagining the same letter rather than the different
Unbeknownst to the participants, Perky was projecting dim letter. Farah interpreted this result as showing that perception
images onto the screen. and imagery share mechanisms; later experiments that have also
shown that imagery can affect perception have come to the same conclusion (Kosslyn &
Thompson, 2000; Pearson et al., 2008). In the next section, we will consider physiological
evidence for connections between imagery and perception.

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 304 4/18/18 5:09 PM

 Imagery and the Brain   305

T E ST YOUR SELF 10.1 Imagery

1. Is imagery just a “laboratory phenomenon,” or does it occur
in real life?
2. Make a list of the important events in the history of the study
of imagery in psychology, from the imageless thought debate
of the 1800s to the studies of imagery that occurred early in
the cognitive revolution in the 1960s and 1970s.
3. How did Kosslyn use the technique of mental scanning (in
(a) Create image
the boat and island experiments) to demonstrate similarities
between perception and imagery? Flash here

4. What is the imagery debate? Describe the spatial (or

depictive) and propositional explanations of the mechanism
then T
underlying imagery. How does the propositional explanation Blank Target
interpret the results of Kosslyn’s boat and island image- or
scanning experiments?
5. Describe experiments by Kosslyn, Perky, and Farah that T then

demonstrate interactions between imagery and perception. Target Blank

(b) Was target letter flashed first or second?

90

Percent correct
detection
Imagery and the Brain 80

As we look at a number of types of physiological experiments, we 70

will see that a great deal of evidence points to a connection between
imagery and perception, but the overlap is not perfect. We begin by Image Image
same as different
looking at the results of research that has measured the brain’s response target from target
(c)
to imagery and will then consider how brain damage affects the ability
to form visual images.
➤ Figure 10.10 Procedure for
Imagery Neurons in the Human Brain Farah’s (1985) letter visualization
experiment. (a) The participant
Studies in which activity is recorded from single neurons in humans are rare but have been visualizes H or T on the screen.
done in situations in which patients are being prepared for brain surgery. (b) Then two squares flash, one
after the other, on the same
screen. As shown on the right,
METHOD Recording from Single Neurons in Humans the target letter can be in the first
square or in the second one. The
The vast majority of single neuron recordings have been carried out on animals. But participants’ task is to determine
in a few experiments, single neuron responses have been recorded from humans. whether the test letter was flashed
In these experiments, the participants were patients with intractable epilepsy that in the first square or in the second
couldn’t be controlled by drugs. For these patients, a possible cure is provided by square. (c) Results showing that
surgery that removes the small area of the brain called the epileptic focus, where the accuracy was higher when the
letter in (b) was the same as the
seizures originate.
one that had been imagined in (a).
To determine the location of this focus, electrodes are implanted in these patients’
(Source: Farah, 1985)
brains and are then monitored over a period of a few days in the hope that sponta-
neous seizures will help pinpoint the location of the focus (Fried et al., 1999). Because
the electrodes are implanted, it is possible, with the patients’ consent, to record the
activity caused by cognitive actions such as perceiving, imagining, and remembering.
These experiments make it possible not only to record neural responses to stimuli, as
is routinely done in animal experiments, but also to study how these neurons respond
when the patients carry out cognitive activities such as imaging and remembering.

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 305 4/18/18 5:09 PM

 306  CHAPTER 10 Visual Imagery

Gabriel Kreiman and coworkers (2000) were able to study

patients who had electrodes implanted in various areas in their
medial temporal lobe, which includes the hippocampus and the
amygdala (see Figure 5.19, page 150). Kreiman found neurons
that responded to some objects but not to others. For example,
the records in Figure 10.11a show the response of a neuron
that responded to a picture of a baseball but did not respond to
a picture of a face. In addition, Figure 10.11b shows that this
neuron fired in the same way when the person closed his or her
eyes and imagined a baseball (good firing) or a face (no firing).
Firing rate

Firing rate
Kreiman called these neurons imagery neurons.
Kreiman’s discovery of imagery neurons is important, both
(a) Perception because it demonstrates a possible physiological mechanism for
imagery and because these neurons respond in the same way
to perceiving an object and to imagining it, thereby supporting
Firing rate

Firing rate

the idea of a close relation between perception and imagery.

However, most research on humans has involved not record-
ing from single neurons, but brain imaging that measures brain
Bruce Goldstein activity as participants are perceiving objects and were creating
(b) Imagery
visual images of these objects (see Method: Brain Imaging,
➤ Figure 10.11 Responses of single neurons in a person’s medial Chapter 2, page 41).
temporal lobe that (a) respond to perception of a baseball but
not of a face, and (b) respond to imagining a baseball but not Brain Imaging
to imagining a face.
(Source: Kreiman, Koch, & Fried, 2000) An early study of imagery using brain imaging was carried out
by Samuel Le Bihan and coworkers (1993), who demonstrated
that both perception and imagery activate the visual cortex. Figure 10.12 shows how activity
in the striate cortex increased both when a person observed presentations of actual visual
stimuli (marked “Perception”) and when the person was imagining the stimulus (“Imagery”).
In another brain imaging experiment, asking
Perception Imagery Perception participants to think about questions that
6 involved imagery—for example, “Is the green of
the trees darker than the green of the grass?”—
4 Stimulus Stimulus generated a greater response in the visual cortex
off off
Percent signal change

than asking nonimagery questions, such as “Is

the intensity of electrical current measured in
2 Stimulus
off amperes?” (Goldenberg et al., 1989).
Another imaging experiment, by Stephen
0 Kosslyn and coworkers (1995), made use of
the way the visual cortex is organized as a
–2 topographic map, in which specific locations
Stimulus Imagined Stimulus on a visual stimulus cause activity at specific
on stimulus on
locations in the visual cortex, and points next
–4
0:24 0:54 1:24 1:54 2:24 to each other on the stimulus cause activity
Time (min:sec) at locations next to each other on the cortex.
Research on the topographic map on the visual
cortex indicates that looking at a small object
➤ Figure 10.12 Results of Le Bihan et al.’s (1993) study measuring brain activity
using fMRI. Activity increases to presentation of a visual stimulus (shaded area causes activity in the back of the visual cortex,
marked “Stimulus on”) and also increases when participants are imagining the as shown by the green area in Figure 10.13a,
stimulus (area marked “Imagined stimulus”). In contrast, activity is low when and looking at larger objects causes activity to
there is no actual or imagined stimulus. spread toward the front of the visual cortex, as
(Source: Le Bihan et al., 1993) indicated by the red area.

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 306 4/18/18 5:09 PM

 Imagery and the Brain   307

➤ Figure 10.13 (a) Looking at a small

object causes activity in the back
of the visual cortex (green). Larger
objects cause activity to spread
forward (red). (b) Results of Kosslyn
et al.’s (1995) experiment. The
symbols indicate the most activated
location caused by imagery: small
image (circle); medium image
(square); large image (triangle).
(a) (b)
(Source: Kosslyn et al., 1995)

What would happen, Kosslyn wondered, if participants created mental images of

different sizes? To answer this question, participants were instructed to create small,
medium, and large visual images while they were in a brain scanner. The result, indicated
by the symbols in Figure 10.13b, is that when participants created small visual images,
activity was centered near the back of the brain (circles), but as the size of the mental
image increased, activation moved toward the front of the visual cortex (squares and tri-
angles), just as it does for perception. (Notice that one of the triangles representing large
images is near the back of the visual cortex. Kosslyn suggests that this could have been
caused by activation by internal details of the larger image.) Thus, both imagery and per-
ception result in topographically organized brain activation.
Another approach to studying imagery and the brain has been to determine whether
there is overlap between brain areas activated by perceiving an object and those activated
by creating a mental image of the object. These experiments have demonstrated an overlap
between areas activated by perception and by imagery,
but have also found differences. For example, Giorgio
Ganis and coworkers (2004) used fMRI to measure acti-
vation under two conditions, perception and imagery.
For the perception condition, participants observed a
drawing of an object, such as the tree in Figure 10.14.
For the imagery condition, participants were told to
imagine a picture that they had studied before, when
they heard a tone. For both the perception and imagery
tasks, participants had to answer a question such as “Is
the object wider than it is tall?”
Results of Ganis’s experiment are shown in Figure
10.15, which shows activation at three different loca-
tions in the brain. Figure 10.15a shows that perception
and imagery both activate the same areas in the frontal
lobe. Figure 10.15b shows the same result farther back
in the brain. However, Figure 10.15c, which shows
activation in the visual cortex in the occipital lobe at
the back of the brain, indicates that perception acti-
vates much more of this area of the brain than does ➤ Figure 10.14 Procedure for Ganis et al.’s (2004) experiment. A trial
imagery. This greater activity for perception isn’t sur- begins with the name of an object that was previously studied, in this
case “tree.” In the imagery condition, participants had their eyes closed
prising because the visual cortex is where signals from
and had to imagine the tree. In the perception condition, participants
the retina first reach the cortex. Thus, there is almost saw a faint picture of the object. Participants then heard instructions.
complete overlap of the activation caused by perception The W in this example means they were to judge whether the object was
and imagery in the front of the brain, but some differ- “wider than tall.”
ence near the back of the brain. (Source: Ganis, Thompson, & Kosslyn, 2004)

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 307 4/18/18 5:09 PM

 308  CHAPTER 10 Visual Imagery

Other experiments have also concluded that there are

G.Ganis,W.L.Thompson, & S.M.Kosslyn, Brain areas underlying visual mental imagery and visual perception:An fMRUI study,
Perception Imagery similarities but also some differences between brain acti-
vation for perception and for imagery. For example, an
fMRI experiment by Amir Amedi and coworkers (2005)
showed overlap but also found that when participants were
using visual imagery, the response of some areas associated

Cognitive Brain Research, 20,226-241.Copyright © 2004Elsevier Ltd.Reproduced by Permission

with nonvisual stimuli, such as hearing and touch, was
(a) decreased. Amedi suggests that the reason for this might be
that visual images are more fragile than real perception, and
this deactivation helps quiet down irrelevant activity that
might interfere with the mental image.

Multivoxel Pattern Analysis

(b) Another way brain imaging has been applied to studying
possible links between imagery and perception is multi-
voxel pattern analysis (MVPA), which we introduced in
Chapter 7 (page 213). Remember that the procedure in
MVPA is to train a classifier to associate a pattern of voxel
activation with particular stimuli, like the apple and pear in
(c)
Figure 7.17 (page 213), and then to present a stimulus and
see if the classifier can identify it, based on the pattern of
voxel activity created by the stimulus.
➤ Figure 10.15 Brain scan results from Ganis et al. (2004). The
Matthew Johnson and Marcia Johnson (2014) used this
vertical lines through the brains in the far-left column indicate
where activity was being recorded. The columns labeled
procedure to study the relation between imagery and per-
“Perception” and “Imagery” indicate responses in the perception ception by training a classifier by presenting four different
and imagery conditions. (a) Responses of areas in the frontal lobe. kinds of scenes—beach, desert, field, or house—to a person
Perception and imagery cause the same activation. (b) Responses in a scanner (Figure 10.16a). After the classifier was trained
farther back in the brain. Activation is the same in this area as on these perceptual stimuli, it was time for the test. Voxel
well. (c) Responses from the back of the brain, including the activity was recorded as a participant viewed a picture (for
primary visual area. There is much more activation in this area in example, the beach scene), and the classifier predicted, from
the perception condition.
two possibilities (say, the beach scene or the house), which
(Source: Ganis, Thompson, & Kosslyn, 2004)
picture the person was perceiving (Figure 10.16b).
The result was that the classifier predicted the correct picture on 63 percent of the trials,
which is above chance accuracy (where chance performance is 50 percent). This “train on
perception, test on perception” test showed that the classifier could use the information it
had learned during perception training to predict what the participant was seeing. But what
if the perception-trained classifier was asked to indicate which of two scenes the participant
was imagining? (Figure 10.16c).
The result of experiments in which voxel activity was measured as participants imag-
ined one of the scenes was 55 percent accuracy—not as good as predicting what the person
was perceiving, but still above chance. Clearly, a lot of work remains to be done before
classifiers can accurately predict what a person is perceiving or imagining, but identifying
with above-chance accuracy what a person is imagining based on activity collected when
the person was perceiving is impressive, and other researchers have reported similar results
(Albers et al., 2013; Cichy et al., 2012; Horikawa & Kamitani, 2017; Naselaris et al., 2015).

Transcranial Magnetic Stimulation

Another technique used to investigate connections between perception and imagery
involves transcranial magnetic stimulation (TMS), which we described in Chapter 9 (see
Method: Transcranial Magnetic Stimulation (TMS), page 292).

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 308 4/18/18 5:09 PM

 Imagery and the Brain   309

Beach Desert Field House

Measure voxel patterns to four different scenes

(a) Train Classifier on 4 scenes

Beach Beach House

Bruce Goldstein; ventdusud/Shutterstock.com; Bruce Goldstein; Breadmaker/Shutterstock.com

Present 1 Scene Classifier predicts which of the two is being perceived
(b) Perception Test

Beach Beach House

Imagine 1 Scene Classifier predicts which of the two is being imagined

(c) Imagery Test

➤ Figure 10.16 Procedure for Johnson and Johnson’s (2014) multivoxel pattern analysis
experiment. (a) The classifier is calibrated by measuring the pattern of voxel activation to
four pictures. (b) Perception test. Participant is presented with one of the pictures, and the
perception-calibrated classifier determines, from the pattern of voxel activation, which of
two possible pictures was presented. (c) Imagery test. Participant is asked to imagine one
of the pictures, and the perception-calibrated classifier determines which of two possible 1 2
pictures was being imagined.

3 4
Stephen Kosslyn and coworkers (1999) presented transcranial magnetic stimulation
to the visual cortex while participants were carrying out either a perception task or an
imagery task. For the perception task, participants briefly viewed a display like the one in
Figure 10.17 and were asked to make a judgment about the stripes in two of the quad-
➤ Figure 10.17 Bar stimuli used for
rants. For example, they might be asked to indicate whether the stripes in quadrant 3 were Kosslyn et al.’s (1999) experiment.
longer than the stripes in quadrant 2. The imagery task was the same, but instead of actually Participants created visual images of
looking at the stripes while answering the questions, the participants closed their eyes and displays such as this and answered
based their judgments on their mental image of the display. questions about the stripes.

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 309 4/18/18 5:09 PM

 310  CHAPTER 10 Visual Imagery

Kosslyn measured participants’ reaction time to make the judgment, both when
transcranial magnetic stimulation was being applied to the visual area of the brain and also
during a control condition when the stimulation was directed to another part of the brain.
The results indicated that stimulation caused participants to respond more slowly, and that
this slowing effect occurred both for perception and for imagery. Based on these results,
Kosslyn concluded that brain activity in the visual cortex plays a causal role in both percep-
tion and imagery.

Neuropsychological Case Studies

How can we use studies of people with brain damage to help us understand imagery? One
approach is to determine how brain damage affects imagery. Another approach is to deter-
mine how brain damage affects both imagery and perception, and to note whether both are
affected in the same way.
Removing Part of the Visual Cortex Decreases Image Size Patient M.G.S. was a
young woman who was about to have part of her right occipital lobe removed as treatment
for a severe case of epilepsy. Before the operation, Martha Farah and coworkers (1993) had
M.G.S. perform the mental walk task that we described earlier, in which she imagined walk-
ing toward an animal and estimated how close she was when the image began to overflow
her visual field. Figure 10.18 shows that before the operation, M.G.S. felt she was about
15 feet from an imaginary horse before its image overflowed. But when Farah had her
repeat this task after her right occipital lobe had been removed, the
distance increased to 35 feet. This occurred because removing part
of the visual cortex reduced the size of her field of view, so the horse
filled up the field when she was farther away. This result supports
the idea that the visual cortex is important for imagery.
Problems with Perceiving Are Accompanied by Problems
with Imagery A large number of cases have been studied in which
a patient with brain damage has a perceptual problem and also has
a similar problem in creating images. For example, people who have
lost the ability to see color due to brain damage are also unable
to create colors through imagery (DeRenzi & Spinnler, 1967;
DeVreese, 1991).
Damage to the parietal lobes can cause a condition called uni-
lateral neglect, in which the patient ignores objects in one half of
the visual field, even to the extent of shaving just one side of his face
or eating only the food on one side of her plate. Edoardo Bisiach and
Claudio Luzzatti (1978) tested the imagery of a patient with unilat-
eral neglect by asking him to describe things he saw when imagining
himself standing at one end of the Piazza del Duomo in Milan, a
“I can get to within 15 feet “The horse starts to overflow
of the horse in my imagination at an imagined distance of place with which he had been familiar before his brain was damaged
before it starts to overflow.” about 35 feet.” (Figure 10.19).
The patient’s responses showed that he neglected the left side
of his mental image, just as he neglected the left side of his percep-
➤ Figure 10.18 Results of the mental walk task for patient tions. Thus, when he imagined himself standing at A, he neglected
M.G.S. Left: Before her operation, she could mentally the left side and named only objects to his right (small a’s). When
“walk” to within 15 feet before the image of the horse
he imagined himself standing at B, he continued to neglect the left
overflowed her visual field. Right: After removal of the right
occipital lobe, the size of the visual field was reduced, and side, again naming only objects on his right (small b’s).
she could mentally approach only to within 35 feet of the The correspondence between the physiology of mental imagery
horse before it overflowed her visual field. and the physiology of perception, as demonstrated by brain scans in
(Source: Farah, 2000) normal participants and the effects of brain damage in participants

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 310 4/23/18 5:00 PM

 Imagery and the Brain   311

with neglect, supports the idea that mental imagery and perception share physiolog-
ical mechanisms. However, not all physiological results support a one-to-one corre-
spondence between imagery and perception. a a

Dissociations Between Imagery and Perception In Chapter 2, we described

dissociations between different types of perception, in which some people with b
brain damage were unable to recognize faces but could recognize objects, and other b b
a
people had the opposite problem (see Method: Demonstrating a Double Dissocia-
tion, Chapter 2, page 40). Cases have also been reported of dissociations between a
b B a
imagery and perception. For example, Cecilia Guariglia and coworkers (1993) stud-
ied a patient whose brain damage had little effect on his ability to perceive but caused
b
neglect in his mental images (his mental images were limited to only one side, as in b
b a a
the case of the man imagining the piazza in Milan). 30
A a
a
Another case of normal perception but impaired imagery is the case of R.M., b
a

who had suffered damage to his occipital and parietal lobes (Farah et al., 1988). b a

R.M. was able to recognize objects and to draw accurate pictures of objects that
b
were placed before him. However, he was unable to draw objects from memory, a
task that requires imagery. He also had trouble answering questions that depend
on imagery, such as verifying whether the sentence “A grapefruit is larger than an ➤ Figure 10.19 Piazza del Duomo in Milan.
orange” is correct. When Bisiach and Luzzatti’s (1978)
Dissociations have also been reported with the opposite result, so that percep- patient imagined himself standing at A, he
tion is impaired but imagery is relatively normal. For example, Marlene Behrmann could name objects indicated by a’s. When
and coworkers (1994) studied C.K., a 33-year-old graduate student who was struck he imagined himself at B, he could name
by a car as he was jogging. C.K. suffered from visual agnosia, the inability to visu- objects indicated by b’s.
ally recognize objects. Thus, he labeled the pictures in Figure 10.20a as a “feather (Source: Bisiach & Luzzatti, 1978)

duster” (the dart), a “fencer’s mask” (the tennis racquet),

and a “rose twig with thorns” (the asparagus). These results
show that C.K. could recognize parts of objects but couldn’t
integrate them into a meaningful whole. But despite his
inability to name pictures of objects, C.K. was able to draw
objects from memory, a task that depends on imagery (Figure
10.20b). Interestingly, when he was shown his own drawings
after enough time had passed so he had forgotten the actual
drawing experience, he was unable to identify the objects he (a)
had drawn.
These neuropsychological dissociations, in which per-
ception is normal but imagery is poor (Guariglia’s patient and
R.M.), or perception is poor but imagery is normal (C.K.),
present a paradox. On one hand, evidence for a double dis-
sociation between imagery and perception (Table 10.1) is
usually interpreted to mean that the two functions are served
by different mechanisms (see page 40). However, this conclu-
sion contradicts the other evidence we have presented that
shows that imagery and perception share mechanisms.
This apparent paradox highlights the difficulty in inter-
preting neuropsychological results. For one thing, the damage
in individual cases varies greatly between individuals and
usually isn’t restricted to the borders between areas in anatom- (b)

ical diagrams. In addition, it is important to bear in mind that

much of the research that presents evidence for an overlap ➤ Figure 10.20 (a) Pictures incorrectly labeled by C.K., who had
between perception and imagery also acknowledges that the visual agnosia. (b) Drawings from memory by C.K.
overlap is only partial. (Source: Behrmann et al., 1994)

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 311 4/18/18 5:09 PM

 312  CHAPTER 10 Visual Imagery

TABLE 10.1
Dissociations Between Perception and Imagery
Case Perception Imagery
Guariglia (1993) OK Neglect (image limited to one side).

Farah et al. (1993) OK. Recognizes objects and can Poor. Can’t draw from memory or
(R.M.) draw pictures. answer questions based on imagery.

Behrmann et al. Poor. Visual agnosia so can’t OK. Can draw object from memory.
(1994) (C.K.) recognize objects.

Conclusions from the Imagery Debate

The imagery debate provides an outstanding example of a situation in which a controversy
motivated a large amount of research. Most psychologists, looking at the behavioral and
physiological evidence, have concluded that imagery and perception are closely related and
share some (but not all) mechanisms (Pearson & Kosslyn, 2015; but see Pylyshyn, 2001,
2003, who disagrees).
The idea of shared mechanisms follows from all of the parallels and interactions
between perception and imagery. The idea that not all mechanisms are shared follows
from some of the fMRI results, which show that the overlap between brain activation is
not complete; some of the neuropsychological results, which show dissociations between
imagery and perception; and also from differences between the experience of imagery
and perception. For example, perception occurs automatically when we look at some-
thing, but imagery needs to be generated with some effort. Also, perception is stable—it
continues as long as you are observing a stimulus—but imagery is fragile—it can vanish
without continued effort.
Another example of a difference between imagery and perception is that it is harder
to manipulate mental images than images that are created perceptually. This was demon-
strated by Deborah Chalmers and Daniel Reisberg (1985), who asked their participants to
create mental images of ambiguous figures such as the one in Figure 10.21, which can be
seen as a rabbit or a duck. Perceptually, it is fairly easy to “flip” between these two percep-
➤ Figure 10.21 What is this, a
rabbit (facing right) or a duck tions. However, Chalmers and Reisberg found that participants who were holding a mental
(facing left)? image of this figure were unable to flip from one perception to another.
Other research has shown that people can manipulate simpler mental images. For
example, Ronald Finke and coworkers (1989) showed that when participants followed
instructions to imagine a capital letter D, and then rotate it 90 degrees to the left and
place a capital letter J at the bottom, they reported seeing an umbrella. Also, Fred Mast
and Kosslyn (2002) showed that people who were good at imagery were able to rotate
mental images of ambiguous figures if they were provided with extra information such
as drawings of parts of the images that are partially rotated. So, the experiments on
manipulating images lead to the same conclusion as all of the other experiments we have
described: Imagery and perception have many features in common, but there are also
differences between them.

Using Imagery to Improve Memory

It is clear that imagery can play an important role in memory. But how can you harness the
power of imagery to help you remember things better? In Chapter 7, we saw that encod-
ing is aided by forming connections with other information and described an experiment
(Bower & Winzenz, 1970) in which participants who created images based on two paired

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 312 4/18/18 5:09 PM

 Using Imagery to Improve Memory   313

words (like boat and tree) remembered more than twice as many words as participants
who just repeated the words (see Figure 7.2, page 194). Another principle of memory we
described in Chapter 7 was that organization improves encoding. The mind tends to spon-
taneously organize information that is initially unorganized, and presenting information
that is organized improves memory performance. We will now describe a method based on
these principles, which involves placing images at locations.

Placing Images at Locations

The power of imagery to improve memory is tied to its ability to create organized locations
at which memories for specific items can be placed. An example of the organizational func-
tion of imagery from ancient history is provided by a story about the Greek poet Simonides.
According to legend, 2,500 years ago Simonides presented an address at a banquet, and just
after he left the banquet the roof of the hall collapsed, killing most of the people inside. To
compound this tragedy, many of the bodies were so severely mutilated that they couldn’t
be identified. But Simonides realized that as he had looked out over the audience during
his address, he had created a mental picture of where each person had been seated at the
banquet table. Based on this image of people’s locations around the table, he was able to
determine who had been killed.
What is important about this rather gory example is that Simonides realized that the
technique he had used to help him remember who was at the banquet could be used to
remember other things as well. He found that he could remember things by imagining a
physical space, like the banquet table, and placing, in his mind, items to be remembered in
the seats surrounding the table. This feat of mental organization enabled him to later “read
out” the items by mentally scanning the locations around the table, just as he had done to
identify people’s bodies. Simonides had invented what is now called the method of loci—a
method in which things to be remembered are placed at different locations in a mental
image of a spatial layout. The following demonstration illustrates how to use the method of
loci to remember something from your own experience.

D E M O N S T R AT I O N Method of Loci
Pick a place with a spatial layout that is very familiar to you, such as the rooms in
your house or apartment, or the buildings on your college campus. Then pick five
to seven things that you want to remember—either events from the past or things
you need to do later today. Create an image representing each event, and place
each image at a location in the house or on campus. If you need to remember the
events in a particular order, decide on a path you would take while walking through
the house or campus, and place the images representing each event along your
walking path so they will be encountered in the correct order. After you have done
this, retrace the path in your mind, and see if encountering the images helps you
remember the events. To really test this method, try mentally “walking” this path a
few hours from now.

Placing images at locations can help with retrieving memories later. For example, to
help me remember a dentist appointment later in the day, I could visually place a huge
pair of teeth in my living room. To remind myself to go to the gym and work out, I could
imagine a treadmill on the stairs that lead from the living room to the second floor, and to
represent the This Is Us TV show that I want to watch later tonight, I could imagine one of
the characters on the show sitting on the landing at the top of the stairs.

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 313 4/18/18 5:09 PM

 314  CHAPTER 10 Visual Imagery

Associating Images with Words

The pegword technique involves imagery, as in the method of loci, but
instead of visualizing items in different locations, you associate them
with concrete words. The first step is to create a list of nouns, like the
following: one–bun; two–shoe; three–tree; four–door; five–hive;
six–sticks; seven–heaven; eight–gate; nine–mine; ten–hen. It’s easy to
remember these words in order because they were created by rhyming
them with the numbers. Also, the rhyming provides a retrieval cue (see
page 202) that helps remember each word. The next step is to pair each
of the things to be remembered with a pegword by creating a vivid image
of your item-to-be-remembered together with the object represented by
the word.
Figure 10.22 shows an image I created for the dentist appointment.
For the other items I wanted to remember, I might picture an elliptical
trainer inside a shoe, and the word US in a tree. The beauty of this system
➤ Figure 10.22 An image used by the author to is that it makes it possible to immediately identify an item based on its
remember a dentist appointment, using the order on the list. So, if I want to identify the third thing I need to do
pegword technique. today, I go straight to tree, which translates into my image of the word
US dangling in a tree, and this reminds me to watch the program This is
Us on TV.
Imagery techniques like the ones just described are often the basis behind books
that claim to provide the key to improving your memory (see Crook & Adderly,
1998; Lorayne & Lucas, 1996; Treadeau, 1997). Although these books do provide
imagery-based techniques that work, people who purchase these books in the hope
of discovering an easy way to develop “photographic memory” are often disap-
pointed. Although imagery techniques work, they do not provide easy, “magical”
improvements in memory, but rather require a great deal of practice and perseverance
(Schacter, 2001).

SOMETHING TO CONSIDER
Individual Differences in Visual Imagery
People differ in how they perceive things and how well they can maintain their attention,
remember things, and solve problems. So it isn’t surprising that there are differences between
people in terms of imagery as well. One person might remember a birthday party by seeing
a birthday cake with candles flickering on a table. But someone else might remember the
party more in terms of where people were standing and the layout of the room (Sheldon
et al., 2017).
The idea that people differ in imaging was suggested in the 19th century by Francis
Galton, who noted that there are “different degrees of vividness with which different
persons have the faculty of recalling familiar scenes under the form of mental pictures”
(Galton, 1880, p. 306). Modern researchers have confirmed Galton’s idea of differences in
people’s imaging and have added important details to the story.
Maria Kozhevnikov and coworkers (2005) did an experiment in which they first
presented a questionnaire designed to determine participants’ preference for using
imagery versus verbal-logical strategies when solving problems. This questionnaire
involved solving different kinds of problems and indicating the strategies used to solve
these problems. Kozhevnikov classified the participants as visualizers or verbalizers, so
this initial result indicated that some people use imagery to solve problems and some

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 314 4/18/18 5:09 PM

 Using Imagery to Improve Memory   315

Paper folded Hole punched Pick the result when unfolded

A B C D E

(a) (b) (c)

➤ Figure 10.23 One trial from the paper folding test. (a) A piece of paper is folded and then
pierced by a pencil to create a hole. (b) The participant’s task is to determine which of five
alternatives is what the paper will look like when unfolded.

Low VVIQ
Intermediate VVIQ

people don’t. We will describe the results of Kozhevnikov’s experiments, High VVIQ
80
focusing on the visualizers.
The visualizers were given tests designed to measure two types of 70

imagery: spatial imagery and object imagery. Spatial imagery refers

Percent of participants
60
to the ability to image spatial relations, such as the layout of a garden. 50
Object imagery refers to the ability to image visual details, features, or
40
objects, such as a rose bush with bright red roses in the garden (Sheldon
et al., 2017). 30
The paper folding test (PFT) is designed to measure spatial imagery. 20
Participants saw a piece of paper being folded and then pierced by a pencil
10
(Figure 10.23a). Their task was to pick from five choices what the paper
would look like when unfolded (Figure 10.23b). 0
Low PFT High PFT
The vividness of visual imagery questionnaire (VVIQ) was designed Spatial imagery
to measure object imagery. Participants rated, on a 5-point scale, the vivid-
ness of mental images they were asked to create. A typical item: “The sun is ➤ Figure 10.24 Results of the vividness of visual
rising above the horizon into a hazy sky.” imagery questionnaire (VVIQ) for participants in
The results of the tests, shown in Figure 10.24, demonstrate dif- Kozhevnikov et al.’s (2005) experiment who were
ferences between participants with a low score on the PFT (low spatial classified as low and high spatial imagery based on
imagery) and participants with a high score (high spatial imagery). Six- results of the paper folding test (PFT).
(Source: Kozhevnikov, Kosslyn, & Shephard, 2005)
ty-two percent of the low spatial imagers had high scores on the VVIQ,
meaning they had high object imagery, whereas 51 percent of the high
spatial imagers had low scores on the VVIQ, meaning they had low
object imagery.
In another experiment, participants were presented with the degraded
pictures task and a mental rotation task. The degraded pictures task con-
sisted of a number of degraded line drawings like the one in Figure 10.25.
Can you determine what it is? (See Figure 10.27 on page 316 for the
answer.) The mental rotation task required participants to judge whether
pictures like the ones in Figure 10.1 (page 300) were two views of the same
object or mirror-image objects. Which type of imagers do you think did
better on each task?
The answer: Spatial imagers did better in the mental rotation task,
and object imagers did better on the degraded pictures task, thus providing
more evidence distinguishing between spatial and object imagers. ➤ Figure 10.25 One of the degraded picture stimuli
In a study designed to determine how well participants with different used by Kozhevnikov et al. (2005). What does the
levels of spatial imagery performed on physics problems, Kozhevnikov and hidden line drawing depict?
coworkers (2007) presented the picture in Figure 10.26 with the following (Source: Kozhevnikov, Kosslyn, & Shephard, 2005)

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 315 4/18/18 5:09 PM

 316  CHAPTER 10 Visual Imagery

➤ Figure 10.26 Picture for the frame of reference problem presented in the Kozhevnikov et al.
(2007) experiment.
(Source: Kozhevnikov, Motes, & Hegarty, 2007)

text and questions to a group of students who had not taken any
physics courses in high school or college.
Frame of Reference Problem
A small metal ball is being held by a magnet attached to a post
on a cart. A cup is on the cart directly below the ball. The cart is
moving at a constant speed as shown by the arrow in the figure.
Suppose the ball falls off the magnet while the cart is in motion.
Observer A stands in the cart, and observer B stands on the
road, directly opposite the post of the cart at the moment of
ball releasing.
Which of the reports described below corresponds to observer
A’s view of the falling ball:
(a) The falling ball moves straight down.
(b) The falling ball moves forward.
➤ Figure 10.27 Answer to the degraded picture problem in
Figure 10.25. (c) The falling ball moves backward.
(Source: Kozhevnikov, Kosslyn, & Shephard, 2005) Which of the reports described below corresponds to observer
B’s view of the falling ball:
(a) The falling ball moves straight down.
(b) The falling ball moves forward.
(c) The falling ball moves backward.
Spoiler alert! Answer before reading further.
The answer: Observer A is in the cart, moving along with the ball, so he will see the ball
as moving straight down into the cup. Because observer B is standing outside the cart, he
will see the falling ball move down and forward before falling into the cup.

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 316 4/18/18 5:09 PM

 Chapter Summary   317

Half of the students correctly answered that observer A would see the ball move straight
down into the cup. Focusing on the students who answered correctly for observer A,
Kozhevnikov and coworkers (2007) found that 70 percent of students who were high
spatial imagers correctly answered that observer B would see the ball moving down and
forward, but only 18 percent of the low spatial imagers came up with the correct answer.
From these results, and the results for additional physics problems, Kozhevnikov concluded
that spatial ability is related to solving many types of physics problems.
The results of experiments like the ones above confirm Galton’s idea that people vary in
their experience of visual imagery, and we now know that people who are good at imagery
are often good at one kind of imagery—spatial or object—but not as good at the other kind.
And, yes, there are also some people who are good at both!

T E ST YOUR SELF 10.2

1. Describe how experiments using the following physiological techniques have
provided evidence of parallels between imagery and perception: (a) recording
from single neurons in the human brain; (b) brain imaging; (c) multivoxel pattern
analysis; (d) deactivation of part of the brain; and (e) neuropsychology
2. What are some differences between imagery and perception? What have
most psychologists concluded about the connection between imagery and
perception?
3. Under what conditions does imagery improve memory? Describe techniques
that use imagery as a tool to improve memory. What is the basic principle that
underlies these techniques?
4. What is the evidence for individual differences in imagery? What is the difference
between spatial visualizers and object visualizers?

CHAPTER SUMMARY
1. Mental imagery is experiencing a sensory impression in 4. The following experiments demonstrated parallels between
the absence of sensory input. Visual imagery is “seeing” imagery and perception: (a) size in the visual field (visual
in the absence of a visual stimulus. Imagery has played walk task); (b) interaction between perception and imagery
an important role in the creative process and as a way of (Perky’s 1910 experiment; Farah’s experiment in which
thinking in addition to purely verbal techniques. participants imagined H or T); and (c) physiological
2. Early ideas about imagery included the imageless thought experiments.
debate and Galton’s work with visual images, but imagery 5. Parallels between perception and imagery have been
research stopped during the behaviorist era. Imagery research demonstrated physiologically by the following methods:
began again in the 1960s with the advent of the cognitive (a) recording from single neurons (imagery neurons);
revolution. (b) brain imaging (demonstrating overlapping activation in
3. Kosslyn’s mental scanning experiments suggested that the brain); (c) multivoxel pattern analysis; (d) transcranial
imagery shares the same mechanisms as perception (that magnetic stimulation experiments (comparing the effect of
is, creates a depictive representation in the person’s mind), brain inactivation on perception and imagery); and
but these results and others were challenged by Pylyshyn, (e) neuropsychological case studies (removal of visual cortex
who stated that imagery is based on a mechanism related to affects image size; unilateral neglect).
language (that is, it creates a propositional representation in 6. There is also physiological evidence for differences
a person’s mind). between imagery and perception. This evidence includes

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 317 4/18/18 5:09 PM

 318  CHAPTER 10 Visual Imagery

(a) differences in areas of the brain activated and (b) brain using the method of loci; and (c) associating items with
damage causing dissociations between perception and nouns using the pegword technique.
imagery. 9. There is variability in people’s ability to use imagery and
7. Most psychologists, taking all of the above evidence into what they experience when they create images. Some people
account, have concluded that imagery is closely related to prefer using verbal-logical reasoning to solve problems, and
perception and shares some (but not all) mechanisms. others are more comfortable using imagery. Among people
8. The use of imagery can improve memory in a number of who are “imagers,” there are spatial imagers and object
ways: (a) visualizing interacting images; (b) organization imagers. Kozhevnikov found that students with high spatial
imagery tend to perform better on physics problems.

THINK ABOUT IT
1. Look at an object for a minute; then look away, create a words”? How does your comparison of written and visual
mental image of it, and draw a sketch of the object based on representations relate to the discussion of propositional
your mental image. Then draw a sketch of the same object versus depictive representations in this chapter?
while you are looking at it. What kinds of information about 3. Try using one of the techniques described at the end of this
the object in the imagery drawing were omitted, compared chapter to create images that represent things you have to
to the sketch you made while looking at the object? do later today or during the coming week. Then, after some
2. Write a description of an object as you are looking at it. time passes (anywhere from an hour to a few days), check to
Then compare the written description with the information see whether you can retrieve the memories for these images
you can obtain by looking at the object or at a picture of and if you can remember what they stand for.
the object. Is it true that “a picture is worth a thousand

KEY TERMS
Conceptual peg hypothesis, 299 Mental rotation task, 315 Spatial imagery, 315
Degraded pictures task, 315 Mental scanning, 300 Spatial representation, 302
Depictive representation, 302 Mental walk task, 304 Topographic map, 306
Epiphenomenon, 302 Method of loci, 313 Unilateral neglect, 310
Imageless thought debate, 299 Object imagery, 315 Visual imagery, 298
Imagery debate, 302 Paired-associate learning, 299 Vividness of visual imagery
Imagery neuron, 306 Paper folding task (PFT), 315 questionnaire (VVIQ), 315
Mental chronometry, 300 Pegword technique, 314
Mental imagery, 298 Propositional representation, 302

COGLAB EXPERIMENTS Numbers in parentheses refer to the experiment number in CogLab.

Link Word (37) Mental Rotation (38)

Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 318 4/18/18 5:09 PM

 Copyright 2019 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. WCN 02-200-202

08271_ch10_ptg01.indd 319 4/18/18 5:09 PM