■■ $>73

Bl RSIwlwll
are

CHA R OUTLINE
l&listic Thinking
i
anding Other People’s Behaviors
336 CHAPTER 9 COGNITION AND PERCEPTION

magine you’re walking through a museum and come upon some art from
different cultures. You may find terracotta figurines from Nigeria, totem poles
from the Pacific Northwest, Chulucanas pottery from Peru, marble statues
from Greece, and bronze Buddhas from Thailand. If there’s one category that
varies tremendously across cultures, and across historical time, it is artistic
traditions. Why do you think there are different traditions? Are they just a matter
of convention, where, for some reason, a certain style became popular at a par­
ticular time and place and subsequent artists responded to that style, either by
emulating it or reacting to it? Or could these differences reveal something about
the underlying psychology of the people of those cultures?
These are questions the psychologist Taka Masuda and his colleagues have
explored (Masuda, Gonzalez, Kwan, & Nisbett, 2008). As he visited art galleries
in Europe and East Asia, Masuda noticed some striking differences in the
paintings produced in the various countries. One thing he observed (and later
documented) is that the horizons in landscape scenes were painted at a much
higher level—about 15% higher on average—in East Asian pictures than in West­
ern ones. He also noticed that subjects in portraits were up to three times larger
in Western pictures than in East Asian ones. These cultural differences were
apparent across a variety of themes and styles (Figure 9.1).
What might these artistic conventions be telling us? Perhaps they simply reflect
changing fashions, and for some reason high horizons and small portrait figures
came into vogue in East Asia in ways that they didn’t in Europe. Many would con­
sider this to be the only acceptable explanation. However, Masuda, his supervisor
Richard Nisbett, and their colleagues have offered a much bolder interpretation:
East Asian art looks different from Western art because people from these cultures
are literally seeing the world differently. In other words, the aesthetic conventions
reflect some fundamental differences in basic cognitive and perceptual processes
between these two cultures (Masuda, Gonzalez et al., 2008; Nisbett, 2003).
If there is a hierarchy of disciplines within the field of psychology, the study of
cognition and perception is seen by many to occupy the top spot. Research on
cognition and perception is highly regarded because it claims to be explaining the
most elementary and essential psychological processes. Cognitive psychologists
explore how the brain processes the streams of information people encounter as
they go about their lives. These investigators are striving to isolate the building
blocks of psychological experiences—the images the eyes perceive, the brain’s
mechanisms for encoding and retrieving memories, how people categorize infor­
mation, and the means of understanding the objects and events in daily life. Since
these are such fundamental processes, it’s remarkable that cognition and per­
ception is the research area that has yielded some of the strongest evidence for
cross-cultural variation (e.g., Ueda et al., 2018). This growing body of research
results challenges some very basic assumptions about how the mind works.
 COGNITION AND PERCEPTION 337

►
I

(b) (d)

FIGURE 9.1 Cross-cultural artistic traditions. The painting by (a) the Flemish artist Berckheyde and
(b) the Japanese artist Hokusai are both landscape river scenes. Berckheyde painted the horizon at
a lower level than Hokusai did. (c) The first portrait is by French painter Blanche, and (d) the second
portrait is by a Chinese painter of the Ming dynasty. The subject of the French portrait occupies a larger
portion of the painting than the subject of the Chinese portrait These stylistic differences are commonly
found between Western and Eastern artists
 338 CHAPTER 9 COGNITION AND PERCEPTION

This chapter explores ways in which cognitive and perceptual processes vary
across cultures. Some questions we’ll investigate are: Do people from different
cultures think about and see the world differently? When they communicate
with others, do people focus on the same things or are there variations? Does
our language affect the way we think? This exploration is guided by the two
themes of this book. First, certain cognitive tools are universally available, but
they are not always used with the same frequency or for the same purpose
Second, the cultural differences that do occur in cognitive and perceptual pro­
cesses arise because of the diverse experiences people have growing up in their
respective cultures.

Analytic and Holistic Thinking

To understand the possible reasons for the variations in the artistic conv< utions of
East Asia and the West, we need to take a step back. Consider this questii Which
of these three is least like the other two: a dog, a carrot, and a rabbit? People typically
give two kinds of answers (Chiu, 1972; Ji, Zhang, 8c Nisbett, 2004). One is the ■ arrot.
Both the dog and the rabbit are animals, so they share common attributes that differ
from the features of the carrot, which is a vegetable. Recall from Chapter 5 that this
thought process reflects taxonomic categorization, the clustering of items according to
their perceived similarities. Westerners usually give this type of response. The other
typical answer is the dog. Rabbits and carrots go together because rabbits eat carrots.
Rabbits and carrots have a relationship, which dogs don’t share. This way of thinking
reflects thematic categorization, the grouping of items based on their causal, tempo­
ral, or spatial associations. Thematic categorization is especially common among East
Asians (and, as we’ll see later, people from many other cultures as well). These two
categorization strategies represent an underlying difference in the way people think
about the world. They correspond to the analytic and holistic reasoning styles intro­
duced in Chapter 5 (Nisbett, 2003; Nisbett, Peng, Choi, & Norenzayan, 2001).
Analytic thinking is characterized by a focus on objects and their attributes. Objects
are perceived as existing independently from their contexts, and they are understood
in terms of their component parts. The features that make up objects are used as a
basis for categorizing them, with fixed abstract concepts to predict and explain them.
As we’ll soon see, analytic thinking, in general, is more common in Western cultures
than elsewhere.
In contrast, holistic thinking is characterized by a focus on the context as a whole.
It is an associative way of thinking or paying attention to the relationships among
objects, as well as between the objects and their surroundings. Objects are under­
stood in terms of how they relate to the rest of the context. Holistic thinking also
emphasizes knowledge gained through experience rather than the application of fixed
abstract concepts. Holistic thinking is more common in non-Western cultures.

s
 ANALYTIC AND HOLISTIC THINKING 339

Analytic and holistic thinking may have developed from the different social
experiences people have within individualistic and collectivistic societies. As we dis­
cussed in Chapter 5, people in collectivistic cultures tend to be socialized in terms
of relationships with others (e.g., Lavin, Hall, 8c Waxman, 2006; Tamis-LeMonda,
Bornstein, 8c Cyphers, 1992; Wang 8c Conway, 2004). People in individualistic
societies, in contrast, are generally socialized to be independent, with their attention
focused on objects (e.g., Bornstein et al., 1992; Wang 8c Conway, 2004). These kinds
of cultural experiences lead people to have primarily either an independent or an
interdependent view of self.
Recall from Chapter 6 that people with an independent self-concept come to under­
stand others by focusing on their inner attributes and attending less to relationships.
People with an interdependent self-concept conceive of people more in terms of their
relationships with others. Nisbett and his colleagues argue that these cultural differ­
ences in ways of understanding people also shape thought processes about their physical
environments (Nisbett et al., 2001; also see Kiihnen, Hannover, 8c Schubert, 2001).
Researchers further suggest that a distinction between analytic and holistic think­
in:; ' a also present in Greeks and Chinese 2,500 years ago (Nisbett, 2003). Analytic
thinking is evident in Plato’s perspective that the world is a collection of discrete,
unchanging objects that can be categorized by reference to a set of universal properties.
The same type of reasoning occurs in Aristotle’s view that a stone falls through the
air because the stone possesses the property of “gravity,” and that a log floats on water
because the log possesses the property of “levity.” Analytic thinking also appears in
the Greek development of an elaborate formal logic system that searched for the truth
according to abstract rules and syllogisms existing independently of observations.
The ancient Chinese demonstrated holistic thinking in their intellectual tradi­
tions of Confucianism, Taoism, and Buddhism, which emphasized harmony, inter­
connectedness, and change. Although the Greek preference for discrete concepts
and abstract principles led to the invention of science, the Chinese tendency to view
the world as consisting of continuously interacting substances led them to the idea
of action at a distance 2,000 years before Galileo’s discovery. In addition, the Chinese
had knowledge of magnetism, acoustic resonance, and the moon’s role in the tides
long before Westerners, even though they lacked a scientific tradition at the time
(Nisbett, 2003). Holistic thinking is also evident in Chinese medical traditions and
in the culture’s emphasis on harmony between people and nature. According to
Nisbett (2003), Western and non-Western cultural differences in ways of think­
ing persist to this day because ancient Greece and Confucian China provided the
intellectual foundation from which modern Western and East Asian societies have
evolved (Nisbett, 2003).
How analytic and holistic thinking styles took root in ancient Greece and
Confucian China is not well understood (but see Nisbett, 2003, for some informed
conjectures). Some research suggests that holistic thinking is widespread through­
out the world and that analytic thinking is the relatively unusual reasoning style,
 340 CHAPTER 9 COGNITION AND PERCEPTION

because it is largely restricted to people who have had contact with Western society or
education systems (see Varnum, Grossman, Kitayama, 8c Nisbett, 2010). Some stud­
ies have found that Arabs and Brazilians show at least as much evidence of holistic
thinking as do the Chinese (de Oliviera 8c Nisbett, 2017; Norenzayan, Choi, Sc Peng,
2007) . Russians are more holistic in their reasoning than Americans (Grossman Sc
Varnum, 2011), as are Turkish farmers and fishermen (Uskul, Kitayama, 8c Nisbett,
2008). Even within cultures there are parallel differences; working-class Americans
and Russians are more holistic than their middle-class compatriots (Grossman 8c
Varnum, 2011). In addition, politically conservative Americans and Chinese are more
holistic than liberals (Talheim, 2018; Talheim et al., 2015). Tie link between analytic
thinking and political views is also supported by the fact that training in analytic
thinking can lead people to become more politically liberal in some circumstances
(Talheim, 2018; Yilmaz 8c Adil, 2017).

Attention
One of the most fundamental psychological processes is attention-. the act or state
of applying the mind to something or directing cognitive activity a certain way.
It follows that analytic thinkers, who perceive the world as consisting of discrete
components, would focus their attention on separate parts of a scene, specific items
of interest. In contrast, holistic thinkers, who perceive the world as consisting of an
integrated whole, would direct their attention more broadly, across an entire scene.
An early finding from the psychoanalytic tradition supported the idea that East
Asians and Westerners showed different patterns of attention to stimuli. Researchers
asked European Americans and Chinese Ameri­
cans to describe what they saw in some Rorschach
inkblots (Abel 8c Hsu, 1949). The Rorschach
test is a psychological measure in which peo­
ple report what they perceive in an ambiguous
visual stimulus—an inkblot on a card (Figure 9.2).
The results of the 1949 study revealed that these
two groups of Americans apparently saw things
quite differently. Tie European Americans
tended to describe what they saw based on a single
aspect of the image—like a little blotch on the
bottom that looked like a Ferrari. In contrast, the
Chinese Americans generally gave “whole-card”
responses, describing what they saw based on the
FIGURE 9.2 A Rorschach inkblot. Responding to this
entire image.
ambiguous visual stimulus is a psychological test of This cross-cultural difference was largely
perception and attention. ignored by psychologists for about 50 years,
 ANALYTIC AND HOLISTIC THINKING 341

because no one had any way of making sense of it. Perhaps it was a fluke. However,
many subsequent studies have built on this intriguing finding and have extended it in
important ways (e.g., De Vos 8c Boyer, 1989).
If East Asians do habitually perceive the world in holistic terms, they should be
especially good at certain kinds of tasks, such as detecting relationships between
elements when considering a whole scene. Researchers investigated this hypothesis by
showing American and Chinese students pairs of pictures on a computer (Ji, Peng,
Sc Nisbett, 2000). The pairings were set so that when one picture was shown (e.g., a
lightbulb), the other picture (e.g., a coin) was shown 0% of the time, 40% of the time,
60% of the time, or 100% of the time. Later the participants viewed just one of the
pictures (the lightbulb) and were asked the likelihood that the other picture (the coin)
would appear beside it. To succeed at this task, you have to be particularly attentive to
the relationships between the pictures, something holistic thinkers should do well. The
results indicated that the Chinese estimates of the likelihood of the appearance of the
correct picture were more accurate than those of the Americans. Apparently, Ameri­
cans focused more on the individual pictures than on the relationships between them.
There are other kinds of tasks on which holistic thinkers should perform poorly.
Their tendency to focus on entire scenes should mean they wouldn’t do well on tasks
that require separating a scene into its component parts. An example is the rod-and-
frame test, shown in Figure 9.3. The goal is to say whether the rod is pointing straight

FIGURE 9.3
Field independence: the rod-and-frame test.

The rod and the surrounding frame are rotated inde­

pendently. The frame's position provides misleading
information about the angle of the rod, so it's necessary
to focus on the rod and ignore the frame, to correctly
identify the angle of the rod This is something analytic
thinkers, high in field independence, can do well.

1
342 CHAPTER 9 COGNITION AND PERCEPTION

up. What makes this task challenging is that the frame around the rod is rotated
independently, so it provides misleading information. To respond correctly, you have
to be able to ignore the misleading information of the frame’s position and focus solely
on the rod. Analytic thinkers, who generally perceive the world as separate objects,
should do well on this task. This cognitive style is known as field independence, the
tendency to separate objects from their background fields. Holistic thinkers, in con­
trast, generally show field dependence, the tendency to view objects as bound to their
backgrounds.
Research shows that the ability to judge the rod’s angle while ignoring the frame
relates to one’s general social orientation. People who pay attention to others develop
more of an orientation toward the field. Those who are extraverted are more field­
dependent than people who are more introverted (Witkin, 1969). Similarly, farmers
who live in societies where they must coordinate their activities with others are more
field-dependent than hunter-gatherers or animal herders (Witkin Sc Berry, 1975).
Likewise, religious training in Calvinism, which emphasizes the independence of
individuals, leads people to be more field-independent than atheists or people with
training in Catholicism or Judaism (Colzato et al., 2010; Colzato, v. n den Wildenberg,
Sc Hommel, 2008). More generally, people in industrialized societies tend to
be quite field-independent, except for people living in highly industrialized East
Asia, where clear evidence for field dependence is found. Overall, East Asians do
relatively poorly on tests like the rod and frame (Ji, Peng, ScNisbett, 2000; Kitayama,
Duffy, Kawamura, Sc Larsen, 2003).
Field dependence in East Asians has been investigated further. In one study,
American and Japanese participants viewed animated computer images of underwa­
ter scenes, with swimming fish, waving seaweed, and floating bubbles (Masuda Sc
Nisbett, 2001). The participants were asked to describe what they saw. The Japanese
participants made about 60% more references to background elements than the
Americans, who spoke more about the fish at the center of the images. The partici­
pants then viewed some additional scenes that included the same central fish they had
seen before. However, some of those fish were presented with the same background
as earlier, whereas other fish were in a novel background. The participants were asked
whether they had seen the fish in the picture before. Regardless of the background,
recognition of the fish by the Americans was pretty much the same (Figure 9.4). The
Japanese participants, in contrast, were much more likely to recognize the fish when
it had the original background than the novel background. They seem to have seen
the fish and the background scenery as bound together. When the background is
changed, the fish no longer look quite the same. When looking at the identical scenes,
Americans and Japanese appear to be perceiving them differently. (For an opposing
view, see Hakim, Simons, Zhao, Sc Wan, 2017.)
This cultural difference in how people perceive parts of scenes is also evident
neurologically. Researchers placed East Asian and American participants in an
 ANALYTIC AND HOLISTIC THINKING 343

76
5 FIGURE 9.4
C 74 Perceptual differences.
c 72
In this study, though American
c □ Original
6)
c. 70

ft
participants recognized fish they had
o Background
<D
seen about equally well, regardless
CC
68 ■ Novel
of the background in the images,
0) Background
O'. 66 Japanese recognized the fish with the
2■ original background much better than
§ 64 the fish with a novel background.
g
£ 62 Source: Adapted from Masuda & Nisbett, 2001.
60
Japanese American

fMRI scanner and showed them pictures of either individual objects (such as an
elephant alone), empty background scenes (such as a savannah with no animals), or
an embedded scene (such as an elephant in a savannah) (Gutchess, Welsh, Boduroglu,
He Park, 2006). The Americans showed more activation of object-processing brain
regions compared with the East Asians. In contrast, no cultural variation was found
in areas associated with processing contexts and backgrounds. In addition, other
research found that the cultural differences in object processing are more pronounced
in comparisons of elderly East Asians and elderly Westerners than in comparisons of
younger East Asians and Westerners (Goh et al., 2007). Apparently, as people age,
their neural functions continue to be shaped by cultural attention patterns. These
studies show that people from different cultures rely on different brain regions for
processing visual information (for a review, see Park He Huang, 2010).
These findings raise a big question. Are people from different cultures really see­
ing things differently or are they processing the same information differently? It’s
possible that Westerners think of objects as more important than East Asians do, and
they thus end up reflecting on them more, as demonstrated by their brain activity. It’s
also possible that because people differ cross-culturally in how they make sense of
their worlds, they would look at scenes differently, leading them to encounter differ­
ent kinds of visual information.
The best way to answer this perception question is to use an eye-tracker, a device
researchers use to determine precisely where someone is looking at any given instant.
What would happen if the eye movements of East Asians and Westerners were con­
trasted? In one study, Japanese and American participants looked at some animated
scenes on a computer while wearing an eye-tracker (Masuda, Ellsworth et al., 2008).
In one scene they saw a central figure in the foreground surrounded by other people
344 CHAPTER 9 COGNITION AND PERCEPTION

in the background. Each person had an emotional facial expression. Sometimes the
background faces had expressions inconsistent with the central figure (e.g., the target
person was smiling but the background people were frowning), and sometimes their
expressions were consistent with the central figure. Tie task for the participants was
to identify the emotion the central figure was experiencing. The Japanese judgments
of the central figure’s emotional expression were influenced by the faces of the back­
ground people. In contrast, the expressions of the background people had no impact
on the judgments of the faces for the Americans. This is another demonstration that
East Asians pay more attention to the background context than Westerners do.
The key question that can be addressed with the eye-tracker, though, is: Why do
the background people influence the East Asian participants more than the Western
ones? Were people from the two cultures looking at the same parts of the scenes?
Figure 9.5 shows the percentage of time participants gazed at the central figure
compared to the background. In the first 1,000 milliseconds, there is little cultural
difference (although this difference is significant). Both Americans and Japanese
spent more than 90% of the time looking at the central figure, with the American
percentage a little higher than the Japanese. In the next two 1,000-millisecond
intervals, people in both cultures started to look a little more at the background,
especially the Japanese. Whereas the Americans were still devoting more than 90%
of their attention to the central figure, the Japanese were giving 70-80% (also see
Chua, Boland, ScNisbett, 2005).
These differences in eye gaze are not innate, as the eye gazes of Asian Canadians
are more similar to those of European Canadians than they are to Japanese (Masuda,
Wang, Ishii, &, Ito, 2012). Furthermore, the East Asian participants made more

100 FIGURE 9.5

15
o> Time and attention.

Ibh
i o> According to this study, when looking
.!> ■ American at a picture, Americans focus more
■ Japanese on the central figure than Japanese
do, especially after the first second
*! has passed. Japanese spend rela­
tively longer focusing on parts of the
Ig background.

a. Source: Adapted from Masuda, Ellsworth et al.,

2008
-1000 msec -2000 msec -3000 msec
Viewing Time
 ANALYTIC AND HOLISTIC THINKING 345

rapid eye movements known as saccades than the Americans, indicating the shift­
ing of the gaze from one fixation point to another (Chua et al., 2005). Compared
to the Americans, the East Asians were scanning the entire scene more thoroughly.
By systematically scanning what they’re looking at, Asians are better able to detect
changes in backgrounds than Westerners are (Miyamoto, Nisbett, &. Masuda, 2006).
However, the tendency of Westerners to focus on central objects enables them to
outperform Asians on tasks that require keeping track of multiple objects (Savani Sc
Markus, 2012). Interestingly, when primed with thoughts about interdependence,
Westerners become more attentive to backgrounds as well (Choi, Connor, Wason,
Sc Kahan, 2016). This underscores the relationship between interdependent thinking
and holistic perception.
These studies suggest, quite remarkably, that people from different cultures are
not seeing the same things, even when they’re looking at identical scenes. The fact
that our eye movements occur largely outside of our voluntary control indicates the
depth of these cultural differences in attention. East Asians are socialized from such
a young age to direct their attention to relationships and associations between parts
that they do so unconsciously by scanning entire scenes. Westerners, in contrast, are
socialized to focus on central objects, so they habitually tend to direct their attention
at specifics rather than the whole.

CULTURAL DIFFERENCES IN ATTENTION AND ARTISTIC PREFERENCES.

Tliis brings us back to the paintings in Figure 9.1. Recall that Masuda and his team
found that horizons were painted at a higher level in East Asian paintings than in
Western ones (Masuda, Gonzalez et al., 2008). A higher horizon calls attention to
the depth of the setting and allows for all the various elements and places within a
scene to be seen in relation to each other, whereas a lower horizon reduces the range
of the scene that is visible. The East Asian paintings thus naturally direct a viewer’s
gaze to the relations among the component parts in a scene, whereas Western paint­
ings direct attention to particular focal points. The same kinds of landscape scenes are
represented quite differently by artists depending on their cultural background.
Similarly, Western portrait artists tend to paint larger figures than East Asians.
A larger subject, particularly a large face, serves to focus the viewer’s attention on the
portrayed individual. The figure comes to dominate the scene and stands apart from
the background. In contrast, in East Asian portraits, the subject remains firmly set
within the surrounding context. Comparing the two portraits (see Figure 9.1c and d),
you can easily tell which culture has cultivated more of a sense that individuals are
distinct and independent people.
It’s also possible, however, that the differences in the paintings reflect old artistic
conventions rather than the aesthetic preferences people currently have. To investi­
gate whether cultural differences in perception affect the art people produce today,
the same researchers had participants draw some pictures on their own (Masuda,
346 CHAPTER 9 COGNITION AND PERCEPTION

c I

zsxx
....X IfB-L
y

FIGURE 9.6 Artistic preferences: objects and backgrounds. The left-hand drawing was done by an American
participant in this study, and the right-hand drawing was done by a Japanese participant. Overall, the
American drawings had lower horizons and fewer elements to contextualize the scene. These findings
are consistent with the notion that East Asian art directs the viewer's attention more to the interrelation­
ships of the different components in a scene.
Source: Masuda, Gonzalez et al.. 2008

Gonzalez et al., 2008). American and East Asian college students were asked to
draw, in 5 minutes, a landscape scene containing at least a house, a tree, a river, a
person, and a horizon. They could include anything else they wanted. Two examples
are shown in Figure 9.6. The researchers analyzed horizon level and background com­
plexity. As in the landscape paintings, East Asians drew a horizon that was much
higher in the picture than in the American drawings. Also, the East Asians tended
to provide a more complex background by including 75% more contextual elements
than the Americans. Overall, East Asians were more likely than Americans to situate
their objects in context.
The same principle applies in photography. When taking photographs, East
Asians include more of the background in the frame and have smaller figures in their
portraits, compared with Americans (Masuda, Gonzalez et al., 2008). Brazilians
similarly tend to include more background details in their photographs than Ameri­
cans (de Oliveira &. Nisbett, 2017). These preferences carry over to the photos people
choose to include of themselves on their Facebook pages. One study shows that East
Asian Facebook photos have smaller figures and larger backgrounds compared to
American ones (Huang Sc Park, 2013).
Overall, East Asian art seems to contain busier scenes than Western art.
Western art is more likely to direct the viewer’s attention to the foreground, by
having either a low horizon or a large central figure or subject. Could these stylistic
differences be reflected in other aspects of daily life across cultures? Yes. Consider
how the actual landscapes are different in various cultures. Researchers in one study
compared landscape photographs in Japan and the United States. According to a
 ANALYTIC AND HOLISTIC THINKING 347

fixed set of rules, they took photographs from post offices and schools in cities of
comparable size in the two countries and then analyzed the number of boundaries
in the scenes (Miyamoto, Nisbett, 8c Masuda, 2006). The Japanese scenes actually
contained more boundaries and edges than the American ones; the physical envi­
ronment in urban areas is literally busier in Japan than in the U.S. Other research
similarly finds that living in busy surroundings leads people to pay attention to
the context, perhaps to facilitate exploration. For example, the Himba people in
Namibia show more interest in their larger surroundings if they have moved to
the city than if they have stayed in their village (Linnell, Caparos, de Fockert, 8c
Davidoff, 2013).
Life in a complex physical environment seems to foster the ability to take in a lot
of information at once. A team of researchers explored how people from different
cultures presented information to others (Wang, Masuda, Ito, 8c Rashid, 2012).
For example, they looked at government and university websites in East Asia and
North America and compared dimensions of the official portals for each site. The
East Asian websites were much longer (as indicated by the scrollbar being smaller)
than the North American sites and had many more links and words (Figure 9.7).
Simply put, the East Asian websites were busier, with more information for people
to navigate.
It appears that regular exposure to busier environments helps East Asians navi­
gate complex scenes. The same researchers conducted another study in which they
timed how long it took European Canadian and East Asian students to find a series
of images buried in a long, complex website (Wang et al., 2012). When the websites
were small, there was no cultural difference in the time it took participants to find the
images. However, when the websites were large, the East Asian students were faster
than the European Canadians at finding the hidden images. The experience East

2 FIGURE 9.7
Presenting online information.

LL
Results from this study show that main
□ East Asian government and university web portals
■ North American from East Asian websites contain a larger
number of links and words than their North
American counterparts. Regular exposure
to busier websites apparently makes East
Asians more efficient than North Americans
at navigating them.
0 ----- ------------------------------------------------------ -----
Number of Links Number of Words
Source; Adapted from Wang et al., 2012.
(Hundreds) (Thousands)
 ■

348 CHAPTER 9 COGNITION AND PERCEPTION

Asians have taking in backgrounds, and living in busy physical spaces, is associated
with the capacity to find details in a complex scene more easily.

Understanding Other People’s Behaviors

Whether a person is thinking holistically or analytically has several important con­
sequences. For example, consider how someone might go about understanding a per­
son’s behavior. Suppose you see an acquaintance, Dan, in a store arguing angrily with
a shop clerk. You might very well wonder why Dan is behaving like this. One way to
explain the behavior would be to consider Dan’s internal characteristics. Maybe he
has a short temper and his arguing demonstrates his sometimes disagreeable person­
ality. A second way would be to consider characteristics of Dan’s situation. Perhaps
he’s angry because he purchased some defective merchandise and the store clerk is
not letting him exchange it. These are two very different ways of explaining the same
behavior.
Trying to understand people’s behavior by considering their inner characteris­
tics (such as their personality traits) is an example of analytic thinking. In contrast,
explaining people’s behavior by considering how the situation is influencing them is
an example of holistic thinking. Given what we know about cultural differences in
analytic and holistic ways of thinking, we would expect Westerners to be more apt
to make dispositional attributions, explaining behavior in terms of a person’s under­
lying qualities. And we’d expect East Asians, as well as people from other cultural
backgrounds, to be more likely to make situational attributions, explaining behavior
in terms of contextual factors.

THE FUNDAMENTAL ATTRIBUTION ERROR. Imagine the following experience.

You read an essay written in the 1960s by a person who was asked to write about
Fidel Castro, who at the time was widely perceived as the number-one enemy of
the U.S. The essay makes several arguments that are clearly pro-Castro. What do
you think the essay writer’s true attitude was about Castro? You’d logically assume
the writer had a positive attitude, based on what she wrote. If the essay included
many anti-Castro arguments, you’d assume the writer had a negative attitude toward
Castro. It’s reasonable to assume that the essay writer expressed her true opinion.
Now, however, imagine that before reading the essay you’re told the writer had been
instructed to write from an angle to be used in a debate—that is, those who wrote
pro-Castro essays had to defend him, and those who wrote anti-Castro essays had to
criticize him. Your task remains the same: identifying the essay writer’s true attitude
about Castro. What do you think?
You can’t assume the writer expressed her true opinion based on instructions to
take a particular position. In fact, the essay tells you virtually nothing about the
writer’s real attitude. But as researchers found, in a classic social psychological study,
 ANALYTIC AND HOLISTIC THINKING 349

people still assumed that the person who was instructed to write a pro-Castro essay
had a positive attitude toward him and that the person who was instructed to write an
anti-Castro essay had a negative attitude (Jones 8c Harris, 1967). In other words, the
participants attributed the behavior of writing the essay to being a reflection of the
writer’s underlying personality, even though it was clear that the writer had no choice
in the angle she took in the essay. This tendency to ignore situational information
(such as the conditions under which the writer wrote) while focusing on dispositional
information (the writer’s assumed attitude) is known as the fundamental attribution
error. It is termed “fundamental” because it’s considered to be deeply ingrained in us.
When we see people acting, we assume they are doing so because of their underlying
dispositions, and we tend to ignore the situational factors that might be governing
their behavior.
However, as with so many other psychological processes, this research was con­
ducted almost exclusively with North American participants. A natural question
emerged: How fundamental is the fundamental attribution error? Is it a universal
tendency to explain behavior primarily in terms of personality while ignoring situa­
tional influences?
ihc anthropologist Clifford Geertz suggests it is not. He asserts that the Balinese
do not tend to conceive of people’s behaviors in terms of underlying dispositions but
instead see them as emerging out of the roles people have (Geertz, 1975). This idea
was developed further by researchers who contrasted Indians and Americans in terms
of the ways they described others (Schweder 8c Bourne, 1982). Participants from
Odisha in eastern India tended to describe others by saying things like, “She brings
cakes to my family on festival days.” In contrast, Americans were more likely to say
things like, “She is friendly.” Therefore, Americans were more likely to conceive of
people in terms of abstract personality traits than the Indians, who described others
in terms of their concrete behaviors.
Comparable cultural differences in the tendency to focus on traits is also evident in
several other experimental models (e.g., Maass, Karasawa, Politi, 8c Suga, 2006; Zarate,
Uleman, 8c Voils, 2001). For example, when participants memorized a series of facial
photographs that were matched with behavioral descriptions (e.g., stopped to help a tour­
ist with directions), the implied trait words (e.g., helpful) were more strongly linked to
the photos for European Americans than Asian Americans, showing that the behavior
was more likely to be interpreted by the European Americans as indicating an underlying
trait (Na 8c Kitayama, 2011). These cultural differences appear to be due to automatic
processes that occur outside a person’s awareness (Shimizu, Lee, 8c Uleman, 2017).
This issue was further explored in an attempt to learn the age at which different
ways of understanding behavior emerges across cultures. In one study, children (ages
8, 11, and 15) and university students were recruited from India and the United States
(Miller, 1984). The participants were asked to describe a situation when someone had
behaved in either an expected, socially acceptable manner or a deviant, antisocial
350 CHAPTER 9 COGNITION AND PERCEPTION

example
manner, and then to explain why the person had behaved that way. Here’s an
of antisocial behavior described by an Indian participant:

This concerns a motorcycle accident. The back wheel burst on the motorcyc'e

The passenger sitting in the rear jumped. The moment the passenger •
he struck his head on the pavement. The driver of the motorcycle—who is an
attorney—as he was on his way to court for some work, just took the passenger to
a local hospital and went on and attended to his court work. I personally feel t^e
motorcycle driver did a wrong thing. The driver left the passenger there without
consulting the doctor concerning the seriousness of the injury—the gravity of the
situation—whether the passenger should be shifted immediately—and he went on
to court. So ultimately the passenger died.

The reasons people gave for the driver’s behaviors were examined, noting, in Par”
ticular, whether they made explanations that referred to the driver’s general disposi­
tion or to the context. (Other reasons for the driver’s behavior are not included here.)
As Figure 9.8 shows, the 8-year-olds from both cultures gave quite similar responses.
However, the older the American sample was, the more likely they were to make

0.45

0.4

0.35

0.3
c
o 0.25
I 0.2
□ U.S.
g K India
0.15
0.1
0.05
0
8-year- 11-year- 15-year- Adults 8-year- 11-year- 15-year- Adults
olds olds olds olds olds olds
Dispositional Situational

FIGURE 9.8 Cultural differences in explaining behavior. In this study, while Indian and American
children showed similar attribution patterns, as they got older Americans tended to make more
dispositional attributions, and Indians made more situational attributions. This finding shows that
preferred ways of making sense of the social world are shaped by cultural experiences.
Source: Adapted from Miller, 1984
 ANALYTIC AND HOLISTIC THINKING 351

dispositional attributions, whereas their situational attributions did not differ by age.
In contrast, the older the Indian sample was, the more situational attributions they
made, whereas their dispositional attributions did not differ significantly by age. The
American university students—adults—showed clear evidence for the fundamental
attribution error by explaining behavior as largely due to personality. In contrast, the
Indian adults did not show any evidence of the fundamental attribution error; rather,
they showed evidence of a reverse fundamental attribution error because they tended
to focus more on the situation than on the disposition. The fundamental attribution
error does not look so fundamental anymore.
This cultural difference in the way people explain the actions of others has some
important consequences (e.g., Lee, Hallahan, Sc Herzog, 1996). For example, how do
we make sense of criminal behavior? Do we blame the person or the situation they were
in? This question was investigated by examining newspaper articles describing mass
murders (Morris & Peng, 1994). One instance was a Chinese graduate student living in
the U.S. who, on losing an award and failing to get a job, returned to school to kill his
supervisor and several others. Another was an American postal worker who, on losing
his job, returned to work, “went postal,” and killed his supervisor and several others.
Both of these stories were covered in the New York Times and the World Journal, the
leading English and Chinese language newspapers. The researchers noted each time
the reporters stated details about the disposition of the accused (e.g., having a “very bad
temper,” being “mentally unstable”) or about the situation that related to the killings
(e.g., had a “rivalry with a slain student,” “had been recently fired”). Regardless of the
particular story, the American articles had more references to the temperament of the
accused than to the situation. Conversely, the Chinese articles had more references to
the situation that had provoked the accused rather than the persons disposition. These
results suggest that courts in different cultures may well view criminal responsibility in
quite different ways. (For a striking contrast of criminal responsibility perspectives in
the Japanese and American legal systems, see Hamilton 8c Sanders, 1992.)
Cultural differences in making dispositional attributions extend well beyond East
Asians and Westerners. People from many other non-Western cultural contexts show a
pattern similar to East Asians (Church et al., 2006). Religious groups differ in their attribu­
tions as well. For example, American Protestants are more likely than American Catholics
to make dispositional attributions, and this difference appears to be due to Protestants hav­
ing a greater commitment to the idea that individuals have a soul (Li et al., 2012). If people
believe God is judging them on the basis of what their soul has done, it follows that they
are more likely to view the soul as being the cause of specific behaviors. In addition, socio­
economic status predicts the kinds of attributions people make. Working-class Americans
make more situational attributions, and fewer dispositional attributions, than middle-class
Americans (Kluegel Sc Smith, 1986; Kraus, Cote, Sc Keltner, 2010). The same kinds of
social class differences in explaining behavior have also been found in France (Beauvois
Sc Dubois, 1988), Russia (Grossmann Sc Varnum, 2011), and India (Mahalingam, 2003,
2007; for a review, see Kraus, Piff, Mendoza-Denton, Rheinschmidt, Sc Keltner, 2012).
352 CHAPTER 9 COGNITION AND PERCEPTION

Accepting Contradiction
In addition to the holistic view that everything is fundamentally inter­
connected, East Asians share a corresponding sense that reality is
continually in flux. This perception of the ultimate fluidity of reality
is captured in the T’ai chi symbol, which encompasses the yin and the
yang (Figure 9.9). Yin and yang represent opposites (literally, they mean
FIGURE 9.9 The Tai chi the moon and the sun), and they indicate that the universe is constantly
yin-yang symbol. This moving from one opposite pole to the other and back again. The dark­
symbol represents the
view that the universe
ness of the night will yield to the brightness of the day, which will lead
is constantly in flux, to darkness again, and the cycle will continue to repeat. This belief in a
moving back and forth fluid and cyclical reality is perhaps most clearly evident in the writings
between opposite poles.
of Lao Tzu, the legendary founder of Taoism. In the Tao Tc Ching he
wrote: “To shrink something, you need to expand it first. To weaken
something, you need to strengthen it first. To abolish something, you need to flourish
it first. To take something, you need to give it first” (Lao Tzu, 2000). This view not
only highlights that reality is in flux but also indicates that opposing truths can be
simultaneously accepted.
Around the time of Lao Tzu, a few thousand miles away Aristotle w? offering
a very different scheme for making sense of the world. He proposed the law of
non-contradiction, proposing that no statement could be both true and false, and
thus “A” could not equal “not A.” This principle is at the heart of logical reason­
ing. Opposing this Aristotelian law, ancient Chinese thought, as captured in the
I-Ching (The Book ofChanges), includes a principle ofcontradiction. Because everything
is perceived to be fundamentally connected with everything else and constantly in
flux, real contradiction ceases to exist. If “A” is connected with “not A,” and if “A”
is always changing into “not A,” then “A” is no longer in contradiction with “not
A.” With this orientation toward the world, contradiction is not something to be
rejected but should be accepted. This acceptance of contradiction has been termed
naive dialecticism (Peng & Nisbett, 1999). It reflects a profoundly different way of
making sense of the world compared to Western logical reasoning.
Consider the following two arguments:

A: A sociologist who surveyed college students from 100 univer­

sities claimed that there is a high correlation among female
students between smoking and being slender.
B: A biologist who studied nicotine addiction asserted that heavy
doses of nicotine often lead to gaining weight.

Can you see the apparent contradiction? One argument is that smoking leads to weight
loss, whereas the other is that smoking leads to weight gain. They are not strictly con­
tradictory, as it’s possible for both of them to be true, even from a logical perspective.
 ACCEPTING CONTRADICTION 353

However, the general thrusts of the arguments are in opposition to each other. Take a
moment to think about how compelling you consider each point of view to be.
Chinese and American students were presented with these arguments, as well
as several other contradictory pairs (Peng 8c Nisbett, 1999). Half the participants
received only one argument (either Argument A or Argument B) and were asked to
indicate how plausible, or compelling, they found it. As shown in the left sides of the
two graphs in Figure 9.10, both Americans and Chinese who received only Argument
A tended to view it as more compelling than those who received only Argument B.
Because these participants saw only one argument, they did not observe any potential
contradiction. The other half of the participants were asked to evaluate both of the

7 FIGURE 9.10
£
Perceivedplausibility ofarguments.
m
3 6
E This study showed that when Americans encoun­
o ter two contradictory arguments, they come
8) 5 to view the better argument as even more
c plausible than when they encounter this same
n argument by itself- In contrast, when Chinese
cc
o> 4 see two contradictory arguments, they view the
CT Argument A
a weaker argument as more plausible than when it’s
Argument B presented alone.
< 3
Only one argument Both arguments Source Adapted from Peng & Nisbett. 1999.
Americans

7
£
*5
5 6
CL
o
8, 5
C
ro
cc
0) 4
Argument A
■ct
Argument B
< 3
Only one argument Both arguments
Chinese
 354 CHAPTER 9 COGNITION AND PERCEPTION

contradictory arguments. These participants saw a potential contradiction. How did

I
seeing the contradiction affect their evaluations of the arguments?
Look at the right side of the American graph. These participants read both of
the contradictory arguments. The top line indicates that Americans who encoun'
tered the contradictory Argument B were even more convinced in Argument A’s
plausibility than those who had read only Argument A. This is a puzzling response,
although it’s consistent with some past research with Westerners (Lord, Ross, &
Lepper, 1979). It would seem that if participants did not find Argument B to be
compelling, they’d ignore it, or perhaps hedge their bets by being a little less confi­
dent in Argument A. But the Americans who saw Argument B instead responded
by being even more confident in the plausibility of Argument A. The Americans
were thus responding to the contradiction by denying its existence; they’re con­
fident that Argument A is the better argument, and thus they are denying any
potential contradiction.
Now look at the right side of the Chinese graph. The top line indicates that the
Chinese participants became less convinced in Argument A’s plausibility when they
encountered an apparently contradictory argument, even though that i gument
tended to be viewed as less compelling. They appear to have adjusted their c\ aluations
in the light of the evidence, which would seem to be sensible for reaching sound con­
clusions. The bottom line of the graph, however, indicates a puzzling pattern. When
the Chinese participants encountered Argument B by itself, they didn’t think it was
very plausible. But when they saw this unconvincing argument paired with the more
compelling Argument A that makes the opposite case, they were then more convinced
by the implausible Argument B.
The idea that smoking is associated with weight gain was compelling to the
Chinese participants only if they had also read an argument that smoking was
associated with weight loss! My helplessly undialectical mind cannot make any
sense of this kind of reasoning. Apparently, the Chinese were reacting to the argu­
ments by noticing the contradiction, which reminded them that the world is often
contradictory—so it was hard for them to say which side is right and which is wrong.
They accepted the contradiction and didn’t seem motivated to get rid of it.
This finding has broad implications. Imagine how having a more dialectical view
would influence one’s response to someone with opposite political views. Or imagine
how tolerating contradiction might affect how one makes sense of scientific findings
that appear to be in opposition.
The East Asian perception of accepting contradiction is not just evident
in how they reason about the external world. It also occurs in the way people
think about their self-concept. In one study, Chinese and American participants
described themselves by completing a questionnaire (Spencer-Rodgers, Peng,
Wang, &. Hou, 2004). The Chinese were much more likely than Americans to
provide statements that were apparent contradictions—such as suggesting they

■
 ACCEPTING CONTRADICTION 355

had both high self-esteem and low self-esteem (also see Bagozzi, Wong, 8c Yi,
1999). Other research has found that Koreans would endorse items indicating
they were both introverted and extraverted (Choi 8c Choi, 2002), and Japanese
were more likely than Canadians to hold contradictory views of their personal­
ities (Hamamura, Heine, 8c Paulhus, 2008). These more contradictory views
among East Asians are also associated with inconsistency in their self-concepts
(Boucher, 2011).
While having a contradictory self-view in the United States sometimes leads
to feelings of anxiety and depression, a study revealed that in Japan, there was
no such association (Brown, 2013). In fact, experiencing contradictory emotions
is associated with better physical health in Japan than in the U.S. (Miyamoto 8c
Ryff, 2011).
Because East Asians more than Westerners perceive life as fluid and changing,
other interesting cultural differences emerge in the ways people think. Westerners do
understand that the world changes, but their views on change are slightly different
from those of East Asians. Westerners appear to see change as occurring in more
linear ways. For example, if a stock price has risen over the past year, it will likely rise
again next year; if the birth rate has dropped over the past generation, it will probably
continue to drop.
In contrast, East Asians appear to believe that change happens in fluid and
unpredictable ways. Consider the following story, which is familiar to almost every
Chinese person:

One day an old farmer's horse ran away from him. His neighbors came by to
comfort him, but he said, "How can you know it isn’t a good thing?" And a few
days later, his horse came back, bringing a wild horse with it. His neighbors came
to congratulate the old man, who said, "How can you know it isn't a bad thing?"
A few weeks later, the old man's son was trying to ride the horse and he fell off,
breaking his leg. When the neighbors came over to express condolences, the old
man said, "How can you know it isn't a good thing?" The next month a war broke
out, and all the able-bodied young men were recruited to fight it. The old man's
son did not have to go because of his broken leg, and he survived with his father.
(Quoted in Ji, 2005)

Clearly, change can happen at any time, and often in precisely the opposite way
one is anticipating. If the story captures widely shared beliefs among East Asians, we
would expect differences in how people will predict the future.
One study investigated this topic by providing Chinese and American stu­
dents with graphs showing the past performance of certain trends, such as
global economy growth rates and worldwide cancer death rates (Ji, Nisbett, &
Su, 2001). Participants were asked to estimate what they thought would happen
356 CHAPTER 9 COGNITION AND PERCEPTION

FIGURE 9.11 Predicting the.future in nonlinear ways. Relative to Westerners, East Asian investors are more
likely to buy stocks that are falling in price and sell stocks whose prices are rising.

over the next few years. The Chinese were almost twice as likely as Americans to
predict that the trend would reverse direction in the future, whereas Americans
were more likely to predict the trend would continue in the same direction as
in the past. This nonlinear perspective of change among Chinese is associated
with their investment strategy of being more likely to go against current trends
(Figure 9.11). Relative to Canadians, Chinese are more willing to buy stocks that
are falling in value and to sell stocks that are rising in value (Ji, Zhang, 8c Guo,
2008). Likewise, people who have more dialectical beliefs are more likely to put
their money into savings when times are good, because they expect a downturn in
their future finances, whereas they put less of their money into savings when times
are bad, expecting their fortunes to rise later (Li, 2018). These cultural differences
in predicting the future are complicated by the fact that East Asians generally place
more value on things that have happened in the past, whereas the opposite pattern
is true for North Americans (Guo, Ji, Spina, 8c Zhang, 2012). These dialectical
expectations for how the future will unfold are not limited to East Asians; one
study found that Brazilians are even more dialectical than Chinese (de Oliviera
8c Nisbett, 2017).
 TALKING AND THINKING 357

diking and Thinking

Let’s think
about talking. Consider this excerpt from an article in a California
newspaper:

i A Professor . encourages his Asian students to speak up in class by making it

Part of the class grade. He makes speaking in front of the class mandatory for
some assignments. "Once they understand this is the norm you expect, they II
get used to it," he says. "But you have to make it clear." (Lubman, 1998, p A12;
quoted in Kim, 2002)

riie reporter is focusing on an issue that is commonly discussed in American

colleges and universities with a large population of students of Asian descent. Often,
students of Asian background speak up less in class than those of other cultural back­
grounds (Tweed 8c Lehman, 2002). This relative silence is often viewed as a cause for
concern; professors and other students may perceive it as shyness, or even a lack of
interest. Such assumptions imply that talking reflects thinking, leading to participat­
ing in class discussions (Figure 9.12). Cultural psychologist Heejung Kim, however,
argues that this assumption is very much grounded in Western cultural practices, and
that talking can affect people from different cultures in various ways (Kim, 2002).

FIGURE 9,12 Styles oftalking and thinking. There are cross-cultural variations in thought processes and
speaking habits. College professors have noticed that Western students often participate in class discus­
sions more than non-Westerners.
1

358 CHAPTER 9 COGNITION AND PERCEPTION

Talking and language have held a privileged position in much of Western intellec­
tual history. Among the ancient Greeks, Homer concluded that there was no greater
skill than to be a good debater, and Socrates believed knowledge existed within peo­
ple and could be revealed only through verbal reasoning. In Judeo-Christian beliefs
the “Word” was viewed as sacred because of its divine power to create. In the United
States the freedom to speak one’s mind is a birthright, protected by the First Amend­
ment to the Constitution. Speaking is valued in the West because it is considered an
act of self-expression and as inextricably bound to thought.
In contrast, however, in many East Asian cultures there has traditionally been
little emphasis on talking, if not outright suspicion of the spoken word. Lao I zu
wrote, “He who knows does not speak. He who speaks does not know.” Practitioners
of many Eastern religions pursue truth through silent meditation rather than spoken
prayer. And as a Korean proverb states, “An empty cart makes more noise.” In many
ways, Eastern cultural traditions have cultivated a belief that thought and speech are
not closely related.
Children are a good example of clear cultural differences in early speech habits.
Research shows that Japanese mothers speak less to their young children th; .. Ameri­
can moms (Caudill 8c Weinstein, 1969). In another study, Chinese infants as young
as 7 months vocalized less in response to laboratory events than European American
infants (Kagan, Kearsley, 8c Zelazo, 1977). Comparable differences have also been
found among older children (Minami, 1994). But it’s important to understand that
less talking does not necessarily mean less communication. In fact, the closer the rela­
tionship, the more people are likely to rely on nonverbal communication rather than
the spoken word (Azuma, 1986; Clancy, 1986), a point to which we’ll soon return.
Nonetheless, the data on the development of speech among East Asian compared to
Western children suggests talking might have different implications across cultures.
Kim (2002) reasoned that if talking really does have a different association with
thinking for East Asians and Westerners, we should see variations in performance
on cognitive tasks depending on whether participants are asked to speak. In other
words, expressing one’s thoughts out loud should interfere with the performance of
East Asians on cognitive tasks, whereas it should have little impact on the perfor­
mance of Westerners. To investigate this question, Kim had participants complete
a version of Raven’s Matrices, the nonverbal IQ_test (see Chapters 3 and 5). The
participants were all born in the U.S., and English was their native language. Half of
them were of European descent, and half of them were of East Asian descent (their
parents were all born in an East Asian country). The 20-item test was divided into
two halves of 10 items each. For the first half, the participants sat alone in a room
and completed the test without speaking at all. For the second half, they remained
by themselves but received one of two sets of instructions. Participants assigned to
a “thinking aloud” condition were asked to talk aloud into a computer microphone
as they worked through the second half of the test. They were vocalizing their
thoughts as they solved the items. In contrast, participants who were assigned to an
 TALKING AND THINKING 359

“articulatory suppression” condition were instructed to repeat the alphabet out loud
into a microphone as they completed the items. (I’ll explain the rationale for the artic­
ulatory suppression condition later.) The key variable was the number of items in the
second half answered correctly in the allotted time period compared to the number
answered correctly in the first half.
Look at the left side of Figure 9.13. The “talking effect” scores reveal the impact
of speaking aloud compared to answering the items in silence. The European Ameri­
cans performed about the same when they were talking as when they were silent. This
suggests that talking and thinking are very much related for European Americans.
In stark contrast, however, Asian Americans performed poorly on the test when they
were talking. Expressing their thoughts out loud interfered with their thinking. For
the kinds of items in Raven’s Matrices, Asian Americans perform better if they are
able to think quietly to themselves. Therefore, rather than indicating a lack of engage­
ment, silence in the classroom might indicate the presence of some good thinking.
Why does this cultural difference exist? As we learned earlier, people of Asian
descent are more likely to engage in holistic thinking, and Westerners are more
likely to engage in analytic thinking. Holistic thinking involves an attention to the
whole and the perception of how various parts are interrelated. Because speaking is
a sequential task, holistic thinkers may find it hard to express themselves out loud.
When you speak, one idea follows another, so you can’t describe multiple things at
the same time. In contrast, analytic thinkers focus on separate parts, so speaking
aloud is easier. Each part can be described separately and sequentially.
The way speech interferes with holistic thinking can be seen when people are
asked to describe faces. Even though a face consists of various parts, people generally

1 FIGURE 9.13
Cognitive tasks and speaking up.
0.5
o
o As this study demonstrated,
W 0 thinking aloud impairs the
■ European
Americans performance of Asian Ameri­
t -0.5 cans on IQ test items, although
m ■ Asian Americans
o> saying the alphabet aloud while

I -1

-1.5
thinking has little impact on
their scores. In contrast, think­
ing aloud has little impact on
the performance of European
-2 Americans, whereas reciting
Thinking Articulatory the alphabet impairs their
Aloud Suppression performance.

Source Adapted from Kim. 2002.

360 CHAPTER 9 COGNITION AND PERCEPTION

see it as an integrated whole. Participants in one study viewed photos and were asked
to describe each face they saw (Schooler 8c Engstler-Schooler, 1990; also see Alogna
et al., 2014). In describing a picture of Tom Cruise, people might say something like,
“He has thickish eyebrows, a big smile, dark hair, and is very attractive.” Tie problem
is, hundreds of faces would fit this description, but they wouldn’t all be Tom Cruise.
Every face, of course, includes many more details, and every face is unique. After
describing the faces they saw, participants looked at another set of faces and were
asked to indicate which ones they had seen before. Interestingly, people were better
able to recognize the faces they had previously seen if they had not tried to describe
them before. Apparently, their spoken descriptions interfered with their ability to
process the face as a whole, causing them to have poorer recall.
Kim (2002) reasoned that if Asian Americans were thinking holistically as they
tried to solve the items in the Raven’s IQ_test, they should be relatively unaffected
by a speaking task that was unrelated to the test. This was the purpose of including
the “articulatory suppression” condition in her study. Participants in that condition
recited the alphabet while trying to solve the Raven’s items. The researcher reasoned
that in this condition, the Asian Americans should be free to think about the items
in the test in the way that was most natural to them (i.e., considering the intercon­
nections of the parts of the items holistically), and they could then engage in the
separate, and less demanding, task of reciting the alphabet. Because these two tasks
are so different, they shouldn’t interfere with each other, and performance should
be largely unaffected. In contrast, however, if talking and thinking arc fundamen­
tally connected for European Americans, being asked to recite the alphabet while
thinking about something else should be challenging. European Americans in the
“articulatory suppression” condition must engage in two verbal tasks at once: their
unspoken verbal thoughts about how to solve the Raven’s items, and their spoken
verbal thoughts associated with the alphabet.
Tie results, as indicated in the right side of Figure 9.13, support this reasoning.
European Americans did very poorly on the IQ_test when they were reciting the
alphabet, indicating that their thoughts while solving the Raven’s items apparently
were verbal and were interfered with by reciting the alphabet. In contrast, reciting the
alphabet had little effect on the performance of the Asian Americans. This suggests
that their thoughts while solving the Raven’s items were largely nonverbal thoughts.
Tiere thus seem to be some profound cultural differences in the way people think,
verbally or nonverbally, at least for the kinds of items in the Raven’s test.

Explicit and Implicit Communication

There is more to communicating than just speaking your thoughts out loud. Much of
what is communicated in the course of a conversation depends on nonverbal cues, such
as hand gestures, facial expressions, and voice tone. You’ve undoubtedly discovered,
 TALKING AND THINKING 361

for example, how easy it can be to inadvertently upset someone in an e-mail or a text
message exchange. When the person can’t see your smile and wink, or hear the jovial
tone of your voice, it’s not always clear from what you have written whether you’re
making a joke or saying something rude. All the nonverbal cues that are so important
for effective communication are absent. That’s why people often include emoticons or
abbreviations, such as :) or LOL, to add contextual details that are lacking in elec­
tronic exchanges.
Although nonverbal communication is common everywhere, there are some clear
cultural differences in the degree to which communication relies on explicit verbal
information or more implicit nonverbal cues. In explaining these differences, the
anthropologist Edward Hall (1976) made a distinction between high-context and
low-context cultures. In a high-context culture, people are deeply involved with one
another, and they share information that guides their behavior. Appropriate behav­
ior for every situation is understood, so there’s no need for explicit communication.
Because people in a high-context culture have a great deal of information in common
they can rely on, they can be less explicit in what they say to each other. In contrast,
in a low-context culture there is less involvement among individuals, and therefore
less shared information to guide behavior. As a result, people in a low-context culture
have to communicate in more explicit detail, as others are less able to fill in the gaps
of what is not said.
East Asian societies are good examples of high-context cultures, whereas North
American, and English-speaking cultures in general, are good examples of low-
context ones. Typically, information communicated in some East Asian languages is
considerably less explicit than in English. As an English speaker who has struggled
with learning Japanese, I can testify that the difference in the explicitness of the
language can be extremely challenging. A question in Japan such as, “Is it okay if
I park my car here?” might very well be answered with a pause, a strained facial
expression, and only the words “Well, a little.” It has taken years of effort for my
helplessly explicit mind to learn that the words that are said in many situations
are sometimes less important than the way they’re spoken. A pause and a strained
look on the face send a signal that is very clear to any native Japanese speaker, yet
still remains opaque to me, that the person has information to communicate to me
that he thinks I’ll find dissatisfying—namely, that it’s not okay for me to park my
car there. The key information is conveyed nonverbally, with the content of the few
words that are actually spoken sometimes being rather empty.
Some clever lab experiments have demonstrated a cultural difference in people’s
reliance on nonverbal communication (Ishii, Reyes, 8c Kitayama, 2003; Kitayama 8c
Ishii, 2002). Japanese and American participants listened to words that were either
pleasant (e.g., grateful, refreshment) or unpleasant (e.g., bitter, complaint), accompa­
nied by either a pleasant-sounding tone or an unpleasant-sounding tone. For some
of the words the explicit meaning matched the tone, and for some the meaning was
362 CHAPTER 9 COGNITION AND PERCEPTION

opposite to the accompanying tone. Tie participants were instructed either to ignore
the tone of the word and say whether the meaning of the word was pleasant or unpleas­
ant, or to ignore the meaning of the word and comment on whether the tone sounded
pleasant or unpleasant. The key variable was how long it took participants to respond
when the tone and the meaning of the words were in conflict. Tie Americans had a
harder time making judgments about the vocal tone (while ignoring the meaning of
the words) than making judgments about the meaning of the words (while ignoring
the vocal tone). This suggests that they habitually focus on the meaning of what is said
more than they do to the tone in which it is spoken. In contrast, the Japanese par­
ticipants showed the opposite pattern of results. Japanese had a harder time focusing
on the meaning of the word while ignoring the tone than focusing on the tone while
ignoring the meaning. This suggests that Japanese are habitually paying attention to
the tone in which things are said more than they are to the precise content of what is
being said.
One alternative explanation is that there is something about the Japanese language,
more than English, that requires people to listen to tone. If this were the case, then
the variation would simply reflect the linguistic skills the languages demand, rather
than a cultural difference. To test for this alternative, the researchers ( plicated
the study with another high-context culture that used two languages. Specifically,
Filipinos belong to a high-context culture, and many are fluent in both Tagalog and
English. If Filipinos find it harder to ignore vocal tones than the meaning, regardless
of the language they are speaking, this would suggest that the findings reflect cultural
differences in attention to context, rather than features of the language. And this is
precisely what the researchers found (see Ishii et al., 2003).

Linguistic Relativity
The results of the Filipino study demonstrate that specific features of the Tagalog and
English languages do not explain why Filipinos hear vocal tones more than Americans
do. However, the question raised here is an interesting one to explore more generally.
How much does the language we speak affect how we think? This question was first
formally proposed by linguist Edward Sapir and his student Benjamin Whorf, and
has become known as the Sapir-Whorf hypothesis, or more commonly, the Whorfian
hypothesis of linguistic relativity (Whorf, 1956). Tie strongest version is that language
determines how we think—that is, we are unable to do much thinking on a topic if we
don’t have the relevant words available to us. This strong version of the hypothesis has
been almost universally rejected. Much thought clearly occurs outside of language; for
example, infants and toddlers show evidence for quite complex thinking before they’re
able to speak. A weaker version of the Whorfian hypothesis is that the language we
speak influences how we think. There has been much debate, controversy, and research
involving this version.
 TALKING AND THINKING 363

file Whorfian hypothesis has a certain amount of intuitive appeal, and it has
been proposed independently by several eminent thinkers (see Hunt Sc Agnoli,
1991, for a review). For example, Herodotus claimed that the Greeks and Egyp­
tians thought differently because the Greeks wrote from left to right, whereas the
Egyptians wrote from right to left (Fishman, 1980). Some 2,000 years later, Albert
Einstein wrote:

[W]e may conclude that the mental development of the individual, and his way of
forming concepts, depends to a high degree upon language. This makes us realize
to what extent the same language means the same mentality. In this sense, think­
ing and language are linked together. (Einstein, 1954, p. 336)

George Orwell also realized the power of language on thinking in his dark, futur­
istic novel 1984, in which the secret police developed a new language, Newspeak, as a
way of controlling people’s thoughts. Today, political correctness is evidence that the
Whorfian hypothesis is still relevant: using words considered to be consistent with
desired outcomes. If, for instance, we say “physically challenged” rather than “handi­
capped” to describe people confined to wheelchairs, we’ll be more likely to think of
them as being capable and competent, which should serve to empower them.
In many ways, we assume the words we speak influence how we think. It’s not that
one language allows people to think about certain ideas, but rather that one language
obliges people to think about certain ideas (Deutscher, 2010). In German, for exam­
ple, inanimate objects are assigned a gender: a bridge is feminine. Therefore, when
Germans think about bridges, they are obliged to think about gender in ways that
English-speaking people are not (Phillips Sc Boroditsky, 2003).
Despite the intuitive appeal of the Whorfian hypothesis, it has been subject
to much intense debate and controversy (e.g., Pinker, 1994; Roberson, Davies, Sc
Davidoff, 2000). You can imagine how useful cross-cultural research could be for
testing this hypothesis. One obvious difference across cultures is the spoken lan­
guage. And it’s not just that different languages have different words for the same
objects—the English horse is called a cheval in French and a hevonen in Finnish.
Rather, numerous words and concepts simply do not exist in many languages around
the world. Tie question is thus whether people who speak different languages think
in some different ways as well.

COLOR PERCEPTION. Much of the debate over the Whorfian linguistic relativity
hypothesis has occurred in the context of color perception. Color is a perceptually
continuous variable that extends gradually through all the hues of the rainbow, yet
linguistically it is a discrete variable, as we have specific words for certain ranges of
light wavelengths within the color spectrum. More importantly, different languages
do not describe the spectrum of colors consistently.
364 CHAPTER 9 COGNITION AND PERCEPTION

An analysis of the various color lexicons of the subsistence societies around the
world that have been investigated has revealed some intriguing patterns (Berlin 8c
Kay, 1969). All known languages have a minimum of two color terms. The one iden­
tified language (Dani) with only two terms has words that roughly correspond to
“black” (which included all dark-hued colors) and “white” (which included all lighter
hues). Next in line are languages that have precisely three color terms that roughly
correspond to black, white, and red. Languages having four color terms have words
that approximate the colors black, white, red, and either green or yellow. Languages
with five color terms have the same words, including words for both green and yellow.
Languages with six color terms add blue to the list, and languages with seven terms
include a word for brown. And languages with eight color terms add cither purple,
pink, gray, or orange to the list. The focal point of each of these color categories (e.g.,
the most prototypical green in an array of different shades of green) is largely similar
across language groups. In sum, people who speak different languages carve up the
color spectrum in different ways, but the categories are not arbitrary, and there are
some strikingly consistent patterns.
Does the fact that people from different cultures describe colors differently affect
how they perceive them? Let’s use the example of the English color term blue.”
English speakers use this word to refer to the color of the sky, blueberric . and the
South Pacific Ocean. The Japanese word for blue is aot and this word is also used to
refer to the color of the sky, blueberries, and the South Pacific Ocean, but the same
word is used to refer to the color of a lawn, a freshly shaved scalp, and what English
speakers would call a green traffic signal. A Whorfian question automatically arises:
When Japanese and English speakers look at the same patch of lawn, are they actually
perceiving the same color?
The strongest test of this kind of question has been conducted by contrasting
people who speak languages that differ the most in their color lexicons. (Japanese
color terms, for the most part, correspond well to their English counterparts, mak­
ing for a weaker comparison.) A research team studied monolingual Berinmo speak­
ers from isolated villages in Papua, New Guinea (Roberson, Davies, 8c Davidoff,
2000). Their language contains five basic color terms corresponding, roughly, to
black (kel)y white (wap), red (mehi), yellow (wor), and green (nol). In addition, they
investigated monolingual Himba speakers, who are seminomadic cattle herders in
Namibia (Roberson, Davidoff, Davies, 8c Shapiro, 2005). The Himba language
also contains five basic color terms, which, likewise, correspond roughly to black
(zoozu), white (vapd), red (serandu), yellow (dumbu), and green (burou), although the
boundaries of these colors differ somewhat from those of the Berinmo (Figure 9.14).
Although there is much overlap in the color boundaries for the three languages,
there are some clear differences.
Further explorations contrasted the performance of the Berinmo and Himba
with English-speaking Britons (Roberson et al., 2000; Roberson et al., 2005).
The studies were based on the idea that different categories should affect people’s
 TALKING AND THINKING 365

^n9lish naming

Light

Dark

Berinmo naming

Light Wap Wap

I Mehi
Wor
Nol
____ . i.,

Kel

Dark

Himba naming

Light Vapa

Burou
Dumbu a
Serandu

Dark

FIGURE 9.14 Color perception and language. The different boundaries associated with color naming in
English, Himba, and Berinmo are shown here.
Source Debi Roberson, personal communication.
366 CHAPTER 9 COGNITION AND PERCEPTION

categorical perception of colors. Much research has shown that we tend to perceive
stimuli as belonging to separate and discrete categories, even though the stimuli may
gradually differ from each other along a continuum. For example, a continuum of
sounds exists between the phonemes “ba” and “pa”; however, any sound that exists
along that continuum is perceived by English speakers as either “ba” or “pa,” and not
as something in between (Macmillan, 1987).
The researchers reasoned that a good test of whether color names influence per­
ception would be whether people given different-color labels are similarly affected by
the boundaries that exist between color categories (Roberson et al., 2000; Roberson et
al., 2005). Participants saw three different colored chips and were asked which of two
chips, Chip 1 or Chip 2, was more similar to the target chip (Figure 9.15). The chips
were chosen so that the target chip was equally distant from Chips 1 and 2 in terms of
hue; however, Chips 1 and 2 fell into two different perceptual categories. For exam­
ple, Chip 1 was typically labeled green by English speakers and Chip 2 was typically
labeled blue. The target chip was usually labeled green. Most English speakers showed
some evidence for categorical perception, as they were more likely to say that the target
chip was more similar to Chip 1 than to Chip 2, because the target chip and Chip 1
shared the same category, whereas the target chip and Chip 2 did not.
The researchers also included triads of color chips based on the color categories
of the Berinmo and the Himba. English speakers, Berinmo speakers, and Himba
speakers viewed several such triads and indicated which chip was more similar in
color to the target chip. As Figure 9.16 shows, there was some evidence that people
from the different cultures made different choices. The English speakers were most
likely to make judgments in line with categorical perception for stimuli that crossed
the blue-green boundary. In contrast, the Berinmo speakers were most likely to show
evidence for categorical perception when they discriminated between two stimuli
that crossed the nol-wor boundary. Likewise, the Himba speakers were most likely
to discriminate between colors that crossed the dumbu—burou boundary. Thus, color
perception is influenced by the terms for color categories in the language of a culture.
(Also see Winawer et al., 2007; but see Heider, 1972, for earlier work that argued
against linguistic relativity for color perception.)

ODOR PERCEPTION. This research on linguistic influences on color perception is

important because it reveals a limitation on the part of psychological scientists, who
are almost exclusively from industrialized societies. With access to many color words,
it’s hard for them to imagine what color perception would be like in a culture where
such words don’t exist. The reverse situation may also be true. One area where the
English language is relatively impoverished is descriptions of odors. English speakers
may try to compare the scent of one thing to another, such as saying the bouquet of a
wine is reminiscent of coffee, banana, and cloves, with olive tapenade notes. But they
have few basic odor terms, beyond words such as “stinky” or “fragrant.” Moreover,
 TALKING AND THINKING 367

Try for Yourself: Perception of colors FIGURE 9.15

Categorical color perception.
Which chip, 1 or 2, is more similar to the color in target
chip below? This study demonstrated that people judge the
similarity of colors partly based on whether
English boundaries
the different color chips share the same
color label.

Source: Adapted from Roberson et at., 2000.

Target Chip

Chip 1 Chip 2

Again, which chip, 1 or 2, is more similar to the color in target

chip below?
Berinmo boundaries

Target Chip

Chip 1 Chip 2

Result:
Most English speakers select Chip 1 for the first one, whereas
Berinmo speakers are equally likely to select Chip 1 as they are
Chip 2. In contrast, English speakers are equally likely to select
Chip 1 as Chip 2 in the second example, whereas Berinmo
speakers tend to choose Chip 1.

Explanation:
For the first example, both the Target Chip and Chip 1 fall
within the English category of green whereas Chip 2 is
usually perceived as slightly blue. Berinmo color categories
don’t distinguish Chips 1 and 2. For the second example,
Berinmo color categories distinguish Chips 1 and 2 whereas
English color categories do not. The color categories in one’s
language influence perception at the boundaries of those
categories.

i
368 CHAPTER 9 COGNITION AND PERCEPTION

90 FIGURE 9.16
80 .—. Cross-cultural color boundaries-

lih
§ 8
P
U 3
■ Green-Blue
Boundary
According to this study, people
make more judgments consistent
O> O with categorical perception for
C co ■ Dumbu-Burou
</> - Boundary stimuli that cross the boundary
between two color labels in their
it
2 8
■ Nol-Wor
Boundary own language, compared with
stimuli that cross a color-label
£Q Q-I
boundary in other languages.

Source Adapted from Roberson et al.,

Berinmo Himba English 2000, 2005
Language

as a review of the literature on olfaction concluded, “Humans are astonishingly bad

at odor identification and naming” (Yeshurun &, Sobel, 2010; also see Cameron,
Anderson, &, Moller, 2016). However, it’s worth noting that most of the studies in this
review were about speakers of European languages, which all have few odor terms.
In contrast, the Jahai, a group of nomadic hunter-gatherers living in the Malay
Peninsula, have a much larger vocabulary of olfaction words. One study compared
native Jahai speakers with American English speakers in their ability to name both
a set of odors (e.g., cinnamon, gasoline, onion) and a set of colors (e.g., red, blue,
black; Majid <3c Burenhult, 2014). The participants were presented with the smells
and color chips and were asked to freely describe them. The researchers wanted to see
the amount of agreement. The Americans did well on the color naming, with much
agreement in the color words they used. But they did poorly in the odor naming, with
very little agreement among them (Figure 9.17). In contrast, the Jahai did about as
well in naming the odors as they did in naming the colors. They had less agreement
than the Americans in naming the colors (reflecting their smaller color vocabulary)
and had more agreement than the Americans in naming the odors. It’s possible that if
English speakers learned a larger set of olfaction words they might also become better
at recognizing different odors.

PERCEPTIONS OF AGENCY. Another example of the influence of language on

cognition is in terms of perceptions of agency. When describing an accident in
English, for instance, it’s common to attribute behavior directly to someone taking
an action—an agent. One might describe Justin Timberlake and Janet Jackson’s
famous wardrobe malfunction at the Super Bowl halftime show like this:
“Justin tore the bodice.” A less common description would be a passive construction,
 TALKING AND THINKING 369

1.0 FIGURE 9.17

c 0.9
0) Perceiving and describing colors and odors.
E 0.8
S> 0.7 In this study, Americans reached more agreement among
? 0.6 themselves when describing colors than when describing
Color
o 0.5 odors There was almost no agreement among American
Odor
o 04 participants in how they described odors. In contrast, the
■q
0.3 Jahai reached about the same amount of agreement when
o 0.2 describing colors as when they were describing odors
(L 0.1
Source: Adapted from Majid & Burenhult, 2014.
0.0
American Jahai

in which there is no actor or agent: “The bodice tore.” Interestingly, people who
heard the former description viewed Timberlake to be more at fault, and assessed a
larger penalty, than those who heard the latter description (Fausey 8c Boroditsky,
2010). fhe statement written in active voice, with Timberlake as the agent, made
him appear more responsible (Figure 9.18). Perhaps not surprisingly then, when
English speakers want to avoid blame, they might use a passive description with­
out an agent. A good example is U.S. President Reagan’s famous 1987 mea culpa:
“Mistakes were made.” Conversely, in Spanish, passive expressions such as
"Sc rompio elflorero'—“Tie vase broke itself”—are common. A Whorfian question
arises: When English and Spanish speakers observe unintentional actions, which
are more commonly described in terms of agency in English than in Spanish, will
they recall those events differently?
Researchers explored this question by having English and Spanish speakers watch
videos showing people involved in some actions, such as breaking a vase, either by
acting intentionally (with agency typically reported in both English and Spanish)
or unintentionally (with agency more likely to be reported in English than Span­
ish) (Fausey 8c Boroditsky, 2010). Later, the participants saw pictures of different
people and were asked to recall which one had broken the vase. The English and
Spanish speakers were equally accurate when identifying those who had acted inten­
tionally (the languages do not differ when describing intentional agency); however,
the English speakers were more accurate than the Spanish speakers in recalling who
had broken the vase unintentionally. Apparently, by not being obligated to describe
unintentional behaviors in terms of agency, Spanish speakers paid less attention to
who had actually broken the vase, and were less likely to recall that fact.

SPATIAL PERCEPTION. Linguistic relativity can also be considered in terms of

spatial descriptions, which can vary a great deal between languages. English speakers
370 CHAPTER 9 COGNITION AND PERCEPTION

FIGURE 9.18 Language and agency. Here are Janet Jackson and Justin Timberlake at the 2004 Super Bowl
halftime show in Houston, Texas. Was an agent involved in the wardrobe malfunction?

often identify locations based on their position relative to the speaker, using terms such
as left, right, front, and back. For example, a grocer might guide shoppers through a
supermarket freezer by letting them know the frozen peas are located in front of the
Brussels sprouts and to the left of the pyrogies. But these directions wouldn’t apply
if a shopper approached the freezer from the other side; they would be reversed. The
directions are accurate only in relation to the speaker. In contrast, the same kinds of
directions given by speakers of Guugu Yimithirr, an Australian Aboriginal language,
would not include any relative terms. The shopper would be guided by being told the
peas are located east of the Brussels sprouts and south of the pyrogies—positions
that are not influenced by the location of the speaker. In other words, speakers of
Guugu Yimithirr identify space in absolute terms (not relative terms), described by
the cardinal points on a compass.
Because Guugu Yimithirr lacks relative directional terms, it would seem that its
speakers would have to have the cardinal directions in mind constantly when inter­
acting with their world. Otherwise, they would have no way of describing spatial
locations when asked about what they had seen. In contrast, English speakers don’t
need to attend to the cardinal directions because they can conceive of space, and can
communicate it to others using terms relative to the location of their bodies.
 TALKING AND THINKING 371

Jow might people who speak different lan-

guages recall scenes they encounter? One study
^ught to test a Whorfian question of whether
uugu Yimithirr speakers and Dutch speakers
w°uld remember scenes differently (Levinson,
199n Dutch is like English in having relative
directional terms. Participants were shown a row
of figures of a cow, a pig, and a person on a table
in one room (Figure 9.19). The table was against

ife
the north wall, so participants were facing north.
Yhen they were asked to go to a different room
that had a similar table with the same figures,
except now the table was against the south wall,
JL-—
In
so participants were facing south. They were
told to arrange the figures to create the same
scene they had seen in the first room. Most of
the Dutch speakers arranged the scene based
on their own position relative to the three fig­
X
ures. Because the cow was left of the pig which "O.K., there's the sun, so that direction is ‘up.’”
was left of the person in the original room, they
maintained these relative positions in the new
room. In stark contrast, most of the Guugu Yimithirr speakers arranged the scene
using cardinal directions instead of relative terms. Because the cow had been west
of the pig, which had been west of the person in the first room facing north, when
they were in the second room facing south they preserved the positions of the figures
according to the cardinal directions. This meant that the cow was now to the right of
the pig, which was to the right of the person—the exact opposite of the arrangement
the Dutch speakers had produced!
The Whorfian explanation is that because the spatial language of Guugu
Yimithirr speakers is based solely on the cardinal directions, they conceive of the
arrangement of their world only with respect to those directions. Dutch speakers,
in contrast, describe locations and positions relative to their physical selves, so they
don’t have to focus on the cardinal directions in most situations. The tendency of
Guugu Yimithirr speakers to conceive of directions in ways that are not relative
to their own physical location appears to be more common among subsistence
populations throughout the world, and is also more similar to the way chimpan­
zees understand directions (Haun, Rapold, Call, Janzen, 8c Levinson, 2006). Paying
attention to directions relative to one’s own egocentric position is a relatively recent
development, and one that you don’t see in many non-industrialized cultures.
Following up on this work, researchers explored whether languages with different
spatial referencing would also influence the ways people represent the passage of time.
 372 CHAPTER 9 COGNITION AND PERCEPTION

Original stimulus FIGURE 9.19

(participants facing north)
Spatial perception and the use ofrelative directional terms.

After viewing a row of figures, Dutch and Guugu Yimithirr

speakers in this study were brought into another room facing
a different direction. When asked to arrange the same figures
to match their earlier order, the Dutch participants arranged
the items from right to left, and the Australian Aboriginal
Guugu Yimithirr speakers arranged them from east to west
This result demonstrates the different ways the two groups
perceive space.
Most common Dutch solution
(participants facing south) Source: Adapted from Levinson, 1997.

Most common Guugu Yimithirr solution

(participants facing south)

Studies had already shown that people’s understanding of time is grounded in their
understanding of space (Boroditsky, 2000). Building on those findings, researchers
reasoned that English speakers tend to see time as passing from the left to the right
(Boroditsky 8c Gaby, 2010). So, for example, if English speakers were to arrange
the images in Figure 9.20 (photos of Boroditsky’s grandfather at various ages) in the
!
correct chronological sequence, they would arrange them from left to right, regardless
of what direction they were facing.
The idea that left is perceived as the origin of time in English is somewhat arbi­
trary, and probably determined by the fact that English is written from left to right.
Would people from non-English-speaking cultures also show this preference? It’s
 TALKING AND THINKING 373

HGURE 9.20 Spatial referencing and the passage oftime. English speakers think of time as passing from
left to right, whereas Australian Aborigines think of time as passing from east to west. An aborigine
would chronologically arrange the pictures differently, depending on the direction he or she is facing.
Source Lera Boroditsky, personal communication

unlikely that the tendency of English speakers to see time flowing from left to right is
universal, since it’s not even universal for people to use linear sequences to represent
any kind of progression (Cooperrider, Marghetis, Sc Nunez, 2017). Another study
investigated Australian Aborigines from the village of Pormpuraaw, who, like the
Guugu ' imithirr speakers, tend to use only absolute terms when referring to spatial
direction- (Boroditsky Sc Gaby, 2010). Among those in Pormpuraaw, time is seen to
pass from east to west, tracking the sun’s movement across the sky. When they were
asked to arrange the photos in Figure 9.20 chronologically, they tended to arrange
them from east to west, regardless of which direction they were facing. This find­
ing further demonstrates how spatial perception, and time perception, are grounded
in the vocabulary of one’s language. (For more examples of cultural and linguistic
influences on time perception, see Boroditsky, Fuhran, 8c McCormick, 2011; de la
Fuente, Santiago, Roman, Dumitrache, Sc Casasanto, 2014).

NUMERICAL COGNITION. We have just reviewed linguistic relativity in the con­

text of several basic kinds of perception. But how broadly are thoughts influenced
by language? Let’s consider mathematics. In many ways, it would seem mathemat­
ics would be one area that should be independent from culture. After all, mathe­
maticians and physicists have proved the existence of mathematical principles that
occur outside of our cultural worlds, such as Einstein’s famous equation, E = me2.
This equation is as true in America as it is in the Amazon or in the constellation of
Andromeda. Mathematical principles exist outside of culture, regardless of our ability
to understand them.
However, mathematics is also an area of pronounced cultural variation in people’s
comprehension of its concepts and details. You probably learned a base-10 number
system at school, but that system is a cultural invention that has not always been
around. The Mayas and Aztecs had a base-20 number system, and the ancient
Babylonians had a base-60 number system. The ancient Romans, despite all their
technological achievements, did not have an understanding of the concept of zero,
374 CHAPTER 9 COGNITION AND PERCEPTION

which was invented in India in the 3rd century. Likewise, there was no real concept
of probability in the West before Pascal invented one in the 17th century. People
in many cultures today learn to do calculations on an abacus, and they conceive of
numbers differently, as is evident in the kinds of errors they make (Miller Sc Paredes,
1996). Our ability to reason with numbers very much reflects our cultural experi­
ences, and most modern industrial societies have developed to the point that they
require their citizens to have mathematics education.
The most compelling evidence that numerical cognition is largely cultural, and
not innate, comes from looking at the numerical systems of subsistence societies. Sev­
eral cultures have relatively impoverished number systems. For example, the Piraha, a
tribe from the Lowland Amazon region of Brazil, has a number system that contains
only 1, 2, and “many.” They have no terms for any specific numbers greater than 2.
The Whorfian question that begs to be asked here is whether people who do not have
number terms in their language can understand the concepts of those numbers.
The Piraha have been studied to investigate the kinds of mathematical computa­
tions they are able to do. In one study, the researcher placed a few nuts in a can, and
the participants were allowed to inspect them (Gordon, 2004). The experimenter
then removed the can from view and took out one nut at a time. After each nut was
removed, the participants were asked whether there were any nuts remaining in the
can. The participants did well on this task when there were initially up to two nuts in
the can; however, their performance steadily dropped in trials that contained increas­
ing numbers of nuts. Those participants who were shown a can containing 6 nuts
rarely were able to correctly determine when the last nut was removed. They did not
seem to be able to perform this task well with numbers larger than 2.
In another task, participants were shown an array of objects—for example, 5 bat­
teries (Gordon, 2004). They were then given some nuts and asked to place out the
same number of nuts as there were batteries (Figure 9.21). Similar to the nuts in the
can task, participants could do this task well with small numbers, but their perfor­
mance deteriorated as the numbers got larger. (For similar findings with another
culture, the Mundurucu of the Amazon, see Pica, Lerner, Izard, Sc Dehaene, 2004.)
In all the tasks completed, the amount of errors the Piraha made increased
with the magnitude of the numbers they were asked to estimate. This suggests that
although the Piraha do not have concepts for specific numbers greater than 2, they
do seem to have rough quantity estimation skills. They know, for example, that
12 batteries is a larger quantity than 8 batteries, even if they can’t distinguish between
more similar quantities, such as between 8 and 9 batteries. This suggests that rough
quantity estimation skills might be innate, whereas numerical skills beyond 2
are acquired through cultural experience. Parallel findings for the development of
numerical cognition in children is also consistent with this notion (Dehaene, 1997).
Studies with other cultures having a limited numerical vocabulary, such as the
Warlpiri in Australia’s Northern Territory, have found that people may use spatial
 TALKING AND THINKING 375

FIG i 1. resting numerical computation. This Piraha man from the Lowland Amazon region in Brazil
is trying to match the number of nuts with the number of batteries shown to him.
Source Gordon, 2004

strategies of trying to match the pattern of a set of objects to estimate quantities in

the absence of numbers (Butterworth, Reeve, 8c Reynolds, 2011).
In sum, without words for specific numbers, people from some cultural groups
seem unable to understand the associated concepts. This may be interpreted as evi­
dence that numerical terms are necessary for calculations and many other aspects
of mathematical thinking (Gordon, 2004). But others argue that these findings
reveal that numerical cognition depends heavily on cultural learning (Everett, 2005;
Gelman 8c Gallistel, 2004).
Cultures that have few number terms also seem to think about numbers differ­
ently. The way people in industrialized societies learn about numbers is linear: The
distance between 17 and 18 is understood to be the same as the distance between
1 and 2. Each unit is of the same magnitude in a linear scale, and counting means
counting in linear units. However, without cultural input, people do not have an
innate linear sense of numbers (Nunez, 2011). Research with American kindergarten
children, who have not yet mastered numerical concepts, reveals that rather than
having a linear sense of numbers, they have what seems to be a logarithmic sense
 376 CHAPTER 9 COGNITION AND PERCEPTION

(Siegler 8c Opfer, 2003). They see larger amounts of space between small numbers
than they do between large numbers. For example, they identify the number 10 as
existing approximately halfway between the numbers 1 and 100, consistent with
its logarithmic value. As they get older, they come to have an increasingly linear
understanding of numbers. Likewise, research with Mundurucu adults, an Amazo­
nian indigenous group with few number words, also finds that they appear to think
about numbers logarithmically (Dehaene, Izard, Spelke, 8cPica, 2008). This suggests
that humans’ default understanding of numbers may be logarithmic, and that we
only come to have a linear understanding of numbers by cultural learning, including
exposure to linear numerical terms. (For another interpretation, see Cantion, Cordes,
Libertus, 8c Brannon, 2009.)
To summarize, the literature on linguistic relativity has been controversial ever
since the Whorfian hypothesis was first proposed. However, many recent studies
have provided compelling evidence for Whorfian effects in several areas. The debate
is by no means over, and with the Whorfian door propped open again, wc might
see researchers attempt to explore linguistic relativity more broadly. If Whorf was
I correct, there should be much variation in psychological processes acros cultures
because of the vast diversity of languages. The Whorfian hypothesis provides a useful
tool with which to make predictions for the kinds of cultural differences that arc
likely to emerge. It is one way of making sense of the different ways people think and
perceive across cultures.

Chapter Review

SUMMARY
• Cognition and perception is the research area that has yielded some of the clearest
evidence for cultural variation, especially in styles of reasoning Westerners are
largely analytic thinkers, focusing on objects separate from their context. Non­
Westerners, especially East Asians, tend to be holistic thinkers, focusing on the
context as a whole.
• In their surroundings, East Asians generally see foreground objects as part of the
background context, whereas Westerners focus on foreground objects, disregarding the
background. This cultural difference in perception appears to be driven by where the
viewer is looking.
 CHAPTER REVIEW 377

Artistic style preferences reveal cultural differences in perceptual habits. East Asian paint­
ings emphasize the context by incorporating small figures and scenes with high horizons,
m contrast to Western paintings, in which figures are relatively large and horizons are low.
• Cross-cultural variations in analytic and holistic thinking affect how people understand
others. Westerners generally explain behavior as arising from dispositional attributions,
or personality traits; East Asians are more likely than Westerners to view behavior as a
function of particular situations.
’ The tendency for analytic thinkers to view objects as separate and internally consistent
means it's hard to tolerate contradiction. In contrast, the tendency for holistic thinkers
to view the world as consisting of fluid and interrelated parts leads to their belief that
contradiction is natural, and they accept it in themselves and their world.
• For East Asians, as holistic thinkers, it's harder to focus on a silent cognitive task and
talk at the same time. Westerners, as analytic thinkers, find it relatively easy to think
and speak simultaneously.
• People in high-context cultures, where shared information guides behavior, observe
nonverbal cues and rely less on explicit language In low-context cultures, explicit
communication guides behavior. East Asians pay attention to vocal tone rather than
the specific content of spoken words; Westerners find it easier to ignore the tone in order
to focus on the meaning.
• The Whorfian hypothesis of linguistic relativity proposes that language influences
thought processes.
• Color perception is affected by the words for colors in one's language, and the same
is true for describing odors. Language also influences perceptions of agency, spatial
perception, and numerical cognition.

KEY TERMS
analytic thinking, p. 338 situational attributions, low-context culture, p. 361
holistic thinking, p 338 p. 348 Whorfian hypothesis of
field independence, p. 342 fundamental attribution linguistic relativity, p. 362
field dependence, p. 342 error, p. 349 categorical perception,
saccades, p. 345 naive dialecticism, p. 352 p. 366
dispositional attributions, high-context culture,
p. 348 p. 361

REVIEW QUESTIONS
i. Analytic thinking is best characterized by
a. taxonomic categorization.
b. thematic categorization.
c. multiple level categorization.
d. categorical analysis.
e. synthetically integrating theses and antitheses.
378 CHAPTER 9 COGNITION AND PERCEPTION

2. Comparisons of Asian Americans and European Americans in their performance on

Raven's Matrices IQ test reveals that
a. Asian Americans tend to outperform European Americans
b. European Americans do poorly on a task if they are reciting the alphabet at the same time
European Americans consistently do better on a task if they are saying something than
if they are silent.
d. Asian Americans do better on a task if they are reciting the alphabet than if they are silent
e. Thinking and talking are largely unrelated for European Americans.

3. People who are especially field-dependent tend to be

a. introverted
b. from Western cultures.
c. hunters or herders.
d. from large, industrialized cities.
e. holistic thinkers. I
4. You are walking with your friend when both of you see someone on the street trip and
fall. You think the person is clumsy, but your friend thinks the person was distracted by
something across the street. In this situation, which of the following most accurately
captures what you are engaging in?
an argument
b. dispositional attribution
inferential judgment
d. conclusion making
e. field dependence

5. Jade, a European Canadian real estate agent, and Jing, a Chinese real estate agent, are trying
to figure out what to tell their clients about the market's outlook. Compared to Jade, Jing will
a. have predictions that are consistently more pessimistic.
b. have predictions that are consistently more optimistic
c. have predictions that follow smoothly from past patterns.
d have predictions that are less linear given past patterns.
e. not be able to have predictions because multiple alternatives are possible for East Asians

6. Which statement is true for people from cultures that do not have words for numbers
beyond 2 or 3?
a. They can detect changes in the number of objects the same way as those who have
words for a complete number set.
b. They can't discriminate between different quantities.
c. They are able to discriminate between rough proportions, but not between precise
numbers.
d. They can do basic counting, but not basic addition or subtraction.
e. They have no concept of what amount is visually greater or smaller than another

•0 - 9 -p - g -q - -a - £ ’q - 3 "P - l
SJdMSUV MaiAajj
 CHAPTER REVIEW 379

THINK ABOUT IT
1. What are the differences between analytic and holistic thinking?
2. Why would East Asians be more comfortable with busier, more complex websites
than Westerners?
3. How might cultural differences in attribution styles affect how people assign penalties to
lawbreakers?
4. Why are Chinese more likely to be contrarian stock investors than Canadians?
5. If talking interferes with thinking more for East Asian students than Western students,
should class participation be rewarded to the same degree in different cultures? What do
you think should be done in multicultural classes?
6. Do people who speak languages with different color terms see colors differently? What
evidence suggests that they do, and what suggests that they don’t?
7. How do people who speak languages that have only a few words for numbers think
about quantities?

s STED READINGS
Dehaene, S. (2011). The number sense: How the mind creates mathematics. Oxford, UK: Oxford
University Press.
Deutscher, G. (2011). Through the language glass. Why the world looks different in other
languages. New York: Picador.
Hall, E. T. (1976). Beyond culture. New York: Anchor Books.
Nisbett, R. E. (2004). The geography of thought. New York: Free Press.