[Paleopsych] Peng and Knowles: Culture, Education, and the Attribution of Physical Causality

[Premise Checker]

Culture, Education, and the Attribution of Physical Causality

Kaiping Peng and Eric D. Knowles, both University of California, Berkeley

Personality and Psychology Bulletin 29.10 (2003.10): 1272-84

Two studies investigated the impact of culturally instilled folk theories on 
the perception of physical events. In Study 1, Americans and Chinese with no 
formal physics education were found to emphasize different causes in their 
explanations for eight physical events, with Americans attributing them more to 
dispositional factors (e.g., weight) and less to contextual factors (e.g., a 
medium) than did Chinese. In Study 2, Chinese Americans' identity as Asians or 
as Americans was primed before having them explain the events used in Study 1. 
Asian-primed participants endorsed dispositional explanations to a lesser 
degree and contextual explanations to a greater degree than did American-primed 
participants, although priming effects were observed only for students with 
little physics education. Together, these studies suggest that culturally 
instilled folk theories of physics produce cultural differences in the 
perception of physical causality.

Keywords: culture; attribution; ethnic identity; physical causality

Authors' Note: Preparation of this article was supported by the Hellman Family 
Faculty Fund to Kaiping Peng. The first author and Richard Nisbett reported a 
portion of the data from Study 1 at the 7th Annual Conference on Culture and 
Science at Kentucky State University in 1997. We are grateful to Richard 
Nisbett for his generous support and sage advice concerning Study 1, to 
Fernando Lopez-Royo for assistance collecting the Study 2 data, and to members 
of the Culture and Cognition Lab at the University of California, Berkeley, for 
their helpful suggestions. Please address correspondence to Kaiping Peng or 
Eric D. Knowles, Department of Psychology, 4143 Tolman Hall, Berkeley, CA 
94720; e-mail: kppeng at socrates.berkeley.edu or eknowles at socrates.berkeley.edu.

Of all the thorny issues confronted by cultural psychologists, one foundational 
question has consistently furrowed researchers' brows: How should culture 
itself be operationalized? Different theoretical traditions have coalesced 
around different answers to this question. Two approaches in particular--the 
value tradition and the "self" tradition--have come to dominate cultural 
psychology. The value tradition (Hofstede, 1980; Schwartz, 1994; Triandis, 
1995) sees culture as a shared set of core values that regulate behavior in a 
population; Triandis's (1995) theory of individualism-collectivism is perhaps 
the preeminent example of this approach. The self tradition (Heine & Lehman, 
1997, 1999; Markus & Kitayama, 1991; Singelis, 1994) identifies culture with 
the particular type of self-conception predominant in a population; Markus and 
Kitayama's (1991) theory of interdependent versus independent self-construals 
is an important example of this tradition. Recently, a number of cultural 
psychologists have adopted a third conception of culture--that of culture as 
"knowledge structure" (Chiu, Morris, Hong, & Menon, 2000; Hong, Morris, Chiu, & 
Benet-Martinez, 2000; Peng, Ames, & Knowles, 2001; Peng & Nisbett, 1999). This 
approach portrays culture as a constellation of knowledge structures, or folk 
theories, that embody individuals' basic beliefs about the world and guide 
inferences in different domains.

Much of the research within the culture-as-theory framework (and cultural 
psychology generally) has focused on cultural differences in social perception. 
For example, cultural differences in dispositional bias (i.e., the tendency of 
lay perceivers to overattribute observed behavior to an actor's personal 
dispositions) have been traced to divergent folk theories of personal agency 
(Choi, Nisbett, & Norenzayan, 1999; Knowles, Morris, Chiu, & Hong, 2001; Menon, 
Morris, Chiu, & Hong, 1999; Morris, Menon, & Ames, 2001; Morris & Peng, 1994; 
Norenzayan & Nisbett, 2000). The influence of culture, however, may reach 
beyond social perception. In this article, we argue that different cultures 
instill their members with different folk theories of physical phenomena and 
that these theories produce cultural differences in the perception of physical 
events. Specifically, we argue that East Asians possess a contextual folk 
theory of physics emphasizing the role of external and relational factors 
(e.g., gravity) in determining an object's behavior. On the other hand, we 
argue that members of Western cultures, such as the United States, possess a 
more dispositional physical theory emphasizing the internal causes of an 
object's behavior (e.g., weight). In Study 1, we present evidence that 
cross-national differences in physical attributions--specifically, between 
those of American and Chinese individuals--reflect the application of 
dispositional or contextual physical theories. In Study 2, we investigate the 
causal impact of folk physical theories among individuals likely to possess 
both dispositional and contextual folk theories-- namely, Chinese Americans. 
Before describing the studies, however, we review evidence for and against the 
proposition that culture affects perceptions of physical events.

CULTURE AND FOLK PHYSICS

Psychologists and other scholars have disagreed as to whether culture shapes 
individuals' causal understandings of physical events. We briefly review work 
supporting and refuting the impact of culture on physical attribution and then 
propose a partial resolution that affirms the role of culture yet places limits 
on the circumstances under which cultural differences will appear.

Evidence for Cultural Variation in Folk Theories of Physics

Scholars in the humanities have long argued for the existence of fundamental 
differences between Eastern and Western folk theories of physics. Joseph 
Needham (1954), a historian of the science and civilization of ancient China, 
argued that the ancient Chinese possessed a much richer and more "advanced" 
folk understanding of physics than did ancient Westerners (e.g., the Greeks) 
and that this understanding more closely resembles modern physics. According to 
Needham (1954; see also Capra, 1975; Zukav, 1979), the core concepts of Eastern 
folk physics, Yin and Yang, are inherently relational, contextual, and 
dialectical, and thus resemble features of contemporary quantum physics. The 
contextual Eastern folk theory, it is argued, emphasizes forces that act over 
distance (e.g., gravity or magnetism) and forces exerted on objects by a medium 
(e.g., air or water).

Kurt Lewin (1935), a founding father of modern social psychology and former 
physics student, was perhaps the first psychologist to address the 
dispositional nature of the Western folk understanding of physics:

The kind and direction of the physical vectors in Aristotelian dynamics are 
completely determined in advance by the nature of the object concerned. In 
modern physics, on the contrary, the existence of physical vectors always 
depends upon the mutual relations of several physical factors, especially upon 
the relation of the object to its environment. (p. 28)

Thus, in Western (i.e., Aristotelian) folk physics, the behavior of objects is 
understood almost exclusively in terms of the object itself: A stone sinks when 
placed in water because it is heavy, and a piece of wood floats because it 
buoyant (Lewin, 1935). On this understanding, the behavior of objects is caused 
by their discrete properties alone rather than by those properties in 
conjunction with states of the environment.

Evidence for Universality in the Perception of Physical Causality

In contrast to the work just cited, cognitive psychologists investigating the 
impact of culture on the perception of physical events have generally failed to 
uncover dramatic cultural differences (Michotte, 1963; Morris, Nisbett, & Peng, 
1995; Morris & Peng, 1994). For instance, Michotte (1963) and his students had 
participants from Europe, Africa, and various Pacific Islands explain 
mechanistic (i.e., "billiard-ball") interactions between inanimate objects, 
finding no significant cultural differences. More recent cross-cultural studies 
have identified seemingly universal rules guiding the interpretation of 
physical phenomena. In general, if the motion of an object follows, in a 
straightforward and visible way, the Newtonian law of conservation (i.e., that 
objects remain at rest or in uniform motion along a straight line until acted 
on by a outside force), then the motion is seen as externally caused (Stewart, 
1984). Only if the motion appears to deviate from the law of conservation is it 
seen as internally caused (Morris et al., 1995; Morris & Peng, 1994; Stewart, 
1984).

Research in cognitive development suggests that the perception of physical 
events is largely "hardwired" and innate and might therefore be resistant to 
the influence of culture. Even very young infants have been shown to possess 
firm and reliable expectations about objects' possible movements and 
interactions (Baillargeon, 2000; Kotovsky & Baillargeon, 2000; Spelke, 2000). 
Of particular relevance, infants younger than 3 years old expect objects to 
behave according to the forces of gravity (an external, relational factor) and 
inertia (an internal, dispositional factor) (I. Kim & Spelke, 1999). The 
emergence at an early age of such physical expectations suggests that these 
expectations could not be altered by experience in one's culture. However, the 
early emergence of expectations concerning both contextual and dispositional 
factors does not rule out the possibility that culture-specific theories 
emphasize these understandings to varying degrees and thus lead individuals to 
favor one type of factor over the other in their explanations for physical 
events.

The Role of Formal Physics Education

The research reviewed above reveals disagreement among scholars concerning 
culture's impact on causal attributions for physical events. We propose a 
partial resolution to this debate. We argue that Lewin (1935) and Needham 
(1954) are correct in their portrayal of Western folk physics as dispositional 
and of Eastern folk physics as more contextual. However, we propose that this 
difference will only be reflected in the judgments of those with little formal 
physics education. If, as we have claimed, intuitions about physical phenomena 
are guided by theory-like knowledge structures, then formal education in 
physics could supplant the Western folk theory with the more contextual 
understanding of modern physics and thus obscure any cultural differences. The 
cross-cultural similarities found in previous research might then be an 
artifact of education because many of the participants in the studies of 
Michotte (1963) and others were college students. Hence, it becomes important 
to separate the effects of culture and physics education in studying causal 
attributions in the physical domain. The studies reported here either involve 
participants with no formal physics education (Study 1) or measure physics 
education to examine its influence on physical judgments (Study 2).

ASSESSING THE CAUSAL IMPACT OF CULTURE-SPECIFIC FOLK THEORIES

In addition to documenting cultural differences in individuals' folk theories 
of physics, it is important to show that culture-specific folk theories exert a 
causal influence on individuals' perceptions of physical phenomena. In virtue 
of its portrayal of culture as a constellation of knowledge structures, the 
culture-as-theory approach suggests a way to investigate the causal impact of 
folk theories on inferences. The technique of cultural priming makes use of the 
fact that some individuals (e.g., biculturals) often possess multiple 
culture-derived theories for the same domain of phenomena and that these 
individuals can be experimentally induced to rely on a given theory in 
interpreting stimuli.

Cultural Priming

If cultures are indeed associated with divergent knowledge structures, then 
cultural knowledge should be subject to well-documented rules of knowledge 
acquisition and use (for a review, see Higgins, 1996). Most important, it 
should be possible for individuals to acquire multiple culture-derived theories 
for the same domain, even if the theories contradict one another; however, only 
one theory at a time can influence judgments (Hong et al., 2000). Which theory 
guides cognition at a given time will depend on the relative cognitive 
accessibility of the theories. According to the principle of accessibility, a 
knowledge structure will affect judgments to the extent that it is available, 
or activated, in the perceiver's mind (Higgins, 1996). Thus, the currently most 
accessible theory for a given domain will be the one that influences judgment 
in that domain.

The current accessibility of a theory can be experimentally manipulated through 
priming, in which the activation level of the construct is increased through 
the presentation of a stimulus semantically related to the construct (Higgins, 
1996). It follows that an experimenter can, by priming selected theories, 
manipulate which of two or more conflicting folk theories will influence 
judgments in a domain. In so doing, it can be shown that the primed knowledge 
structures exert a causal influence on judgments.

Researchers working within the culture-as-theory framework have used cultural 
symbols to prime culture- specific knowledge structures. Hong and colleagues 
(2000) presented individuals who had extensive experience in both East Asian 
and American culture with either East Asian cultural icons (e.g., a Chinese 
flag, the Great Wall of China, a picture of Stone Monkey) or American cultural 
icons (e.g., the American flag, the Capitol Building, a picture of Superman). 
These researchers found that priming affected attributions for social behavior, 
such that individuals exposed to East Asian primes interpreted behavior as more 
externally caused than did individuals in the American prime condition.1 In our 
research, we primed culture-specific folk theories using a cultural identity 
prime in which Asian American participants were asked to reflect on their 
identity as Asians or as Americans. These identity primes were intended to 
increase the level of activation of related networks of cultural knowledge, 
including East Asian and Western folk theories of physics.

Bicultural Individuals

The cultural priming technique used here relies on the possibility that some 
individuals possess more than one culture-bound folk theory of physics. 
Bicultural individuals--individuals who identify with more than one 
culture--may possess multiple folk theories. There is good reason to believe 
that ethnic cultures within the United States possess cultural knowledge 
similar to that of their countries of origin. For instance, there is evidence 
that Japanese Americans possess social attributional tendencies consistent with 
the contextual theory of social behavior thought to be dominant within Japanese 
culture, attributing success and failure more to situational factors than do 
European Americans (Narikiyo & Kameoka, 1992; Whang & Hancock, 1994). The 
congruence between the attributional tendencies of Asian Americans and members 
of Asian national cultures may be due to the fact that Asian Americans possess 
some of the same values and cultural knowledge prevalent in the national 
cultures (U. Kim & Choi, 1994). We suggest that, as with folk theories of 
social behavior, Asian Americans may possess both Asian and Western folk 
theories of physics.

THE CURRENT RESEARCH

The current research had two goals. First, we sought to demonstrate that 
American and Chinese national cultures are associated with different folk 
theories of physics. Specifically, we hypothesized that whereas Americans have 
a dispositional physical theory that locates the causes of physical phenomena 
in the discrete dispositions of objects (e.g., weight), Chinese perceivers have 
a contextual theory that places greater emphasis on relational 
factors--specifically, forces over distance (e.g., gravity) and the influence 
of mediums (e.g., air or water). Toward this end, in Study 1, we asked Chinese 
and American nationals with no formal physics education to identify the causes 
of a variety of physical events.

Our second goal was to investigate the causal impact of culture-specific folk 
physical theories on attributions for physical events. Thus, in Study 2, we 
attempted to prime dispositional (Western) or contextual (East Asian) folk 
theories of physics in the minds of Asian Americans, who presumably possess 
both theories. Participants were subsequently asked to identify the causes of 
the same physical events as were used in Study 1. To test our hypothesis that 
folk theories will affect inferences only for individuals who have had little 
formal instruction in physics, we measured participants' level of physics 
education.

STUDY 1

Method

PARTICIPANTS

Fifteen American participants, all of them female, were drawn from the 
Psychology Department subject pool at the University of Michigan; they 
participated in return for course credit. Participants were selected who 
reported having had no formal education in physics and who had declared majors 
in the arts or humanities. Fifteen female spouses of visiting Chinese graduate 
students at the University of Michigan were recruited and paid $10 for their 
participation in this study.2 All of these participants were Chinese citizens 
and, similar to their American counterparts, were college educated, reported no 
formal physics education, and had majored in the arts and humanities. The mean 
ages of American and Chinese participants were 19.1 and 22.7 years, 
respectively. All of the Chinese participants had been in the United States for 
less than a year because their visas only permitted them to remain in the 
country for a short time.

PHYSICAL CAUSALITY DISPLAYS

Eight animated displays of physical events were created using Macromind 
Director for Macintosh by Macromedia Software, a computer animation program. 
All displays depicted a white object interacting in various ways with a black 
object or a medium (i.e., air or water) (see Figure 1 for schematic 
representations of the displays):

1. White object interacting with black object

a. "Launching" event (i.e., elastic collision). The black object collides with 
the stationary white object and stops, causing the white object to move.

b. "Launching at a distance" event. This display was identical to the launching 
interaction except that the black object stops short of the white object before 
the white object begins moving.

c. "Entraining" event (i.e., inelastic collision). The black object collides 
with the stationary white object, after which both objects move together.

d. "Balance" display depicting objects balancing on a lever. In this event, the 
black and white circles are in balance at two ends of a platform resting on a 
fulcrum.

e. "Magnetic" display depicting objects' motions in a magnetic field. In this 
event, the black and white circles appear to be magnetically attracted to one 
another, converging slowly at first, and then more quickly as the distance 
between them narrows.

2. White object interacting with a medium

a. "Hydrodynamic--floating" event. The white object bobs on the surface of a 
pool of water.

b. "Hydrodynamic--dropping" event. The white object drops into the pool, rises 
to the surface, and bobs for a moment.

c. "Aerodynamic" display depicting an object's motion in the air. In this 
event, the white object looks like a balloon, dropping gradually while buffeted 
by air currents.

The program was set to present the displays either in the order listed above or 
the reverse, as determined randomly.

Figure 1 Schematic representations of displays used in Studies 1 and 2.

PROCEDURE

Participants were run one at a time. The experimenter, who was European 
American and fluent in both English and Chinese, brought the participant into a 
test

ing room and seated her in front of a computer. All instructions were given in 
English for American participants and in Chinese for Chinese participants. The 
experimenter introduced the study as an investigation of visual perception, in 
which the participant would be shown a number of displays depicting physical 
events and asked questions about her perceptions of each. Participants were 
instructed to think of the physical events as independent episodes, such that 
objects in one display were not the same objects as in any other display. The 
experimenter then played the displays, pausing after each to ask the 
participant the following physical causality question: "Please explain in your 
own words why the white object moved in the way it did. Even if you don't have 
a strong opinion, please take a guess." Participants were given as much time as 
needed to respond to the questions. Responses were tape-recorded. The procedure 
typically lasted an hour, after which the participant was debriefed and 
dismissed.

Results

CODING OF OPEN-ENDED RESPONSES

Participants' answers to the free-response causality question were transcribed 
and content analyzed by two psychology graduate students at the University of 
Michigan. Both coders were blind to the experimental hypothesis. One coder was 
European American and the other Chinese American, and each was fluent in both 
English and Chinese. Each coder coded all of the responses. For each open-ended 
response, coders tallied dispositional explanations (e.g., weight, shape) and 
contextual explanations (e.g., gravity, liquid) for the white object's movement 
as well as the perceived nature of the object (e.g., ball, balloon). Table 1 
shows the complete coding scheme. Interjudge reliability was assessed by 
Kendall's coefficient of concordance; good reliability was achieved (W = .90, p 
< .001).

THE PERCEIVED NATURE OF THE EVENTS

Because the current research assesses the impact of culture on perceptions of 
physical causality, we intended for participants to interpret the animated 
events as physical interactions between inanimate objects. However, in light of 
research showing that moving geometrical figures can give the impression of 
animacy and personality (Heider, 1944; Michotte, 1963), it is possible that 
some participants saw the objects as representing organisms and their movements 
as social in nature. However, our coding of participants' impressions of the 
nature of the objects (refer to Table 1 for the coding scheme) indicated that 
all participants perceived the animated circles as balls or (in the aerodynamic 
display) as a balloon, not as animals or as humans. These findings are 
consistent with cross-cultural studies in judgments of animacy showing that 
across cultures, people as young as 3 years old make similar judgments in 
distinguishing animals from objects (Carey, 1991).

TABLE 1: Coding Scheme for Open-Ended Physical Causality Judgments in Study 1

Property Type Categories

Nature of the white object

Ball

Balloon

Animal

Human

No mention

Could not categorize

Dispositional causes of whole object's movement

Weight or mass

Composition

Inertia

Shape

Energy

Electricity

Magnetism

Internal dynamics (e.g., heat, engine)

Other disposition (e.g., size, color)

Could not categorize

Contextual causes of white object's movement

Other object

Gravity

Friction

Air or wind

Invisible matter

Liquid or current

Field

Outside physical conditions (e.g., smoothness of surface)

Other forces (e.g., human intervention)

Other contextual cause

Could not categorize

CAUSAL EXPLANATIONS FOR THE PHYSICAL EVENTS

Having established that all participants perceived the animated displays as 
physical in nature, we next compared American and Chinese participants' 
preferences for dispositional and contextual explanations of the events. To 
isolate the effect of nationality on participants' preferences for different 
types of explanations, we sought to control for any influence of nationality on 
the total number of causes cited. Thus, for each of the eight physical events, 
we calculated the percentage of American and Chinese explanations that referred 
to dispositional and contextual factors. For all events, we predicted that 
American participants would give proportionally more dispositional explanations 
than would Chinese but that Chinese participants would give proportionally more 
contextual explanations than would Americans. For each display, we conducted a 
z test comparing the percentage of American versus Chinese explanations that 
referred to dispositional causal factors. (Note that separate z tests comparing 
rates of contextual explanations would have results identical to the tests of 
dispositional explanations and are thus unnecessary.) Because our hypothesis 
was clearly directional, we used one-tailed tests of significance. The 
differences between American and Chinese rates of dispositional and contextual 
explanations were significant for the Launching, Magnetic, and Aerodynamic 
displays and marginally significant for the Hydrodynamic dropping display, such 
that Americans gave more dispositional explanations than did Chinese. Averaging 
across all displays, the predicted effect of nationality on percentage of 
dispositional explanations was marginally significant, p < .10. American and 
Chinese percentages of dispositional and contextual explanations, and the 
corresponding z statistics, are presented in Table 2.

Discussion

Study 1 provides partial evidence that members of different cultures possess 
divergent folk theories of physical causality. In their open-ended explanations 
for the physical events, American participants exhibited a greater preference 
for dispositional explanations than did Chinese participants on three of the 
eight displays, whereas Chinese participants emphasized contextual explanations 
more than did Americans. This suggests that, as argued by Needham (1954), Capra 
(1975), and Zukav (1979), the Eastern folk theory of physics places more 
importance on contextual factors (such as other objects, forces over distance, 
and mediums) and less importance on dispositional causes (such as shape and 
weight) than does the American folk theory of physical causality.

For all the displays, cultural differences were in the predicted 
direction--with Chinese being more contextual than Americans in their 
explanations of physical motions. Nonetheless, the cultural difference was 
statistically significant only for the Launching, Magnetic, and Aerodynamic 
displays and marginal for the Hydrodynamic dropping display. Although this 
discrepancy could be a result of the small sample size, it is interesting to 
note that cultural differences were stronger for displays that depicted salient 
energy transitions from one object to another object or a medium (Launching, 
Hydrodynamic dropping, Magnetic, and Aerodynamic) than for displays in which 
the energy transition is less salient (Balance, Launching at a distance, 
Hydrodynamic floating, and Entraining). We had no a priori reason to expect 
cultural differences to be limited to displays depicting salient energy 
transitions, and further research should examine the reliability of this 
finding.

Inspection of Table 2 suggests a possible alternative framing of our findings. 
The analyses reported above examined the effect of culture on participants' 
tendency to give dispositional versus contextual explanations for the physical 
displays. However, the data also can be analyzed by comparing differences in 
dispositional and contextual explanation within culture. It is clear that 
Chinese participants favored contextual explanations over dispositional 
explanations; however, Americans showed a relatively small preference for 
dispositional explanations. Therefore, it is possible that whereas Chinese 
possess a markedly contextual folk physical theory, Americans have a more or 
less evenhanded theory emphasizing the importance of both types of explanation. 
This alternative interpretation should be viewed with caution, however. 
Americans' unexpectedly high reliance on contextual explanations is driven 
largely by responses on only two of the eight physical displays, Hydrodynamic 
dropping and Hydrodynamic floating, in which there is a visible medium that 
exerts an obvious influence on the behavior of the white object. Thus, 
interpretation of property use within culture is vulnerable to idiosyncrasies 
of the particular displays participants were shown, some of which demanded 
mention of important contextual causes.

TABLE 2: Percentage of American and Chinese Explanations Coded as Dispositional 
or Contextual in Study 1 [dropped]

Having found evidence in Study 1 that American and Chinese cultures are 
associated with divergent folk theories of physics, we next sought to 
demonstrate a causal connection between culture-specific knowledge structures 
and patterns of physical attribution. If culture-specific lay theories are 
indeed responsible for the cultural differences observed in Study 1, then it 
should be possible to temporarily increase the cognitive accessibility of 
different theories and thus increase their influence on attributions. In Study 
2, we tapped a population likely to possess both Asian and Western lay 
theories--specifically, Chinese Americans--and attempted to influence their 
attributions by priming one or other of these theories. We primed Chinese 
Americans' identity either as Asians or as Americans before having them explain 
the same series of physical events used in Study 1. We predicted that 
participants receiving the Asian identity prime would prefer dispositional 
causes to a lesser degree, and contextual causes to a greater degree, than 
would participants receiving the American prime.

The results of Study 1 may be seen to conflict with findings in cognitive 
psychology revealing no effects of culture on perceptions of physical causality 
(e.g., Michotte, 1963). We argued earlier that formal physics education may 
sometimes supplant or obscure folk theories and thus prevent cultural 
differences from emerging. In Study 1, we were careful to choose participants 
with no formal education in physics--and thus whose inferences are likely to be 
based on their folk physical theories--allowing the observed cultural 
difference to emerge.

In addition to examining the causal influence of folk physical theories, Study 
2 was intended as a more direct test of the idea that formal physics education 
may supplant or obscure individuals' folk theories of physical phenomena. 
Participants in Study 2 reported the amount of physics instruction they have 
received and rated their physics expertise. We predicted that the effect of 
cultural identity priming on attributions would be qualified by an interaction 
with participants' amount of physics education such that only participants with 
little physics background would be affected by the identity prime. Participants 
high in physics education, who presumably rely on a formally inculcated theory 
of physics rather than a culture-specific folk theory, should not be affected 
by the identity prime.

STUDY 2

Method

PARTICIPANTS

Sixty-five students (44 women) at the University of California, Berkeley, 
participated in fulfillment of psychology course requirements. The mean age of 
the participants was 19.7 years. Participants were selected who had reported 
their ethnicity to be Chinese American during a mass data collection at the 
beginning of the semester.

MATERIALS

Cultural identity primes. Primes of Asian and of American identity consisted of 
a short questionnaire asking participants to reflect, in writing, on several 
aspects of their ethnic identity. First, participants were asked to "recall an 
experience you had that made your identity as an American [Asian] apparent to 
you." (Brackets indicate wording in the Asian prime condition.) Participants 
then answered the following questions about the experience: "When did you have 
this experience?" "How old were you when you had this experience?" "Briefly 
describe the experience, " and "Why do you think the experience made your 
American [Asian] identity apparent?"

Physical displays. Study 2 employed the same eight physical displays as did 
Study 1 (Launching, Launching at a distance, Entraining, Hydrodynamic floating, 
Hydrodynamic dropping, Balance, Magnetic, and Aerodynamic) (see Figure 1). 
Displays were presented using Flash by Macromedia Software, a computer 
animation program.

Rating packets. Participants made their ratings of the physical displays in a 
packet containing Likert-type questions corresponding to several causal 
factors. For each of the eight displays, participants rated the extent to which 
the white object's movement was due to five dispositional factors of the white 
object (shape, weight, composition, buoyancy, and inertia) and four contextual 
factors acting on the white object (gravity, friction, air/wind, and water). 
All ratings were made on a 5-point scale from 1 (not at all responsible) to 5 
(completely responsible).

Ratings of physics background. A short questionnaire was created to gauge 
different aspects of participants' background in physics. As a measure of 
formal instruction in physics, participants reported the total number of 
physics classes they had taken in high school and college. As a measure of 
physics expertise, participants rated their current physics expertise on a 
5-point scale from 1 (none) to 5 (expert).

PROCEDURE

Participants were run in groups of 5 to 10 in a large testing room outfitted 
with computers. As each participant entered the testing room, he or she was 
handed an Asian or American identity prime from an alternating stack, thus 
randomizing assignment of participants to the Asian and American prime 
conditions. Participants were then seated at computers and asked to spend 3 
minutes filling out the identity primes, after which the primes were collected.

Next, participants viewed each of the eight physical events in random order. 
The computer displayed each event twice, after which participants were referred 
to the appropriate page in their rating packets where they rated the degree to 
which each causal factor was responsible for the event. After completing 
ratings for all eight displays, participants completed the questionnaire 
gauging the amount of physics instruction they had received and their 
self-reported physics expertise.

The entire procedure typically lasted an hour, after which participants were 
debriefed and dismissed.

Results

DERIVATION OF AGGREGATE ATTRIBUTION SCORES

For use in the analyses reported below, we created aggregate measures of 
dispositional and contextual attribution across all physical displays. Each 
participant's aggregate dispositional attribution score was calculated by 
averaging his or her endorsement of dispositional causal factors (i.e., shape, 
weight, composition, buoyancy, and inertia); aggregate contextual attributions 
were calculated by averaging each participant's endorsement of contextual 
causal factors (i.e., black object, gravity, friction, air/wind, and water).

EFFECTS OF IDENTITY PRIMING AND PHYSICS EDUCATION

We tested two hypotheses in this study. First, we predicted that the identity 
priming manipulation would influence Chinese American participants' 
attributions for the physical events, such the participants receiving the Asian 
prime would attribute the physical events more to contextual causes, and less 
to dispositional causes, than would participants receiving the American prime. 
Second, we predicted that priming effects would occur only for participants 
with little formal education in physics. To test these hypotheses, we followed 
Aiken and West's (1991) procedure for testing Categorical × Continuous 
interactions using multiple regression. Unlike analysis of variance (ANOVA), 
the regression method has the advantage of not requiring a split (such as a 
median split) to be performed on the continuous variable, which discards useful 
variance. We began by standardizing the dummy-coded prime condition variable, 
the measure of physics education (i.e., number of physics classes taken), and 
self-reported physics expertise to create three main effect terms (see Table 3 
for the correlations between these variables and aggregate dispositional and 
contextual attribution scores). Next, we multiplied the main effect terms 
together to create interaction terms for each two-and three-way interaction 
(i.e., Prime × Physics Classes, Prime × Physics Expertise, Physics Classes × 
Physics Expertise, and Prime × Physics Classes × Physics Expertise). We then 
performed two simultaneous multiple regression analyses, one to test the 
influence of these main effect and interaction terms on dispositional 
attribution and one to test effects on the contextual attribution. Because 
dispositional and contextual attributions were highly correlated, r = .68, p < 
.01, 3 we controlled for this relationship by adding standardized contextual 
attribution score as a predictor in the analysis of dispositional attributions 
and standardized dispositional attributions scores as a predictor in the 
analysis of contextual attributions. Tables 4 and 5 summarize the regressions 
analyses.

TABLE 3: Pearson Correlations Between Variables in Study 2 (N = 65) [dropped]

Identity priming. As can be seen in the first row of Table 4, participants 
receiving the Asian prime made significantly less extreme dispositional 
attributions for the physical events than did participants receiving the 
American prime. Moreover, whereas the Asian prime decreased dispositional 
attribution among Chinese Americans, it increased contextual attribution (see 
Table 5, first row).

Physics education. As shown in the fourth row of Table 4, the effect of 
identity priming was qualified by a marginally significant interaction with the 
number of physics classes, such that the influence of priming on dispositionism 
decreased as physics instruction increased. Likewise, the effect of identity 
priming on contextualism decreased as physics instruction increased (see Table 
5, fourth row).

To visualize the interactions between physics education and identity priming, 
we plotted the interactions according to the procedure recommended by Aiken and 
West (1991), with levels of dispositionism and contextualism predicted based on 
the regression equations. Figure 2 represents the predicted effects of the 
priming manipulation on dispositional attribution among participants one 
standard deviation above and below the mean on physics education. Similarly, 
Figure 3 represents the predicted effects of priming on contextual attributions 
for participants high and low in physics education.

Self-reported physics expertise. An unexpected finding emerged involving 
participants' self-reported physics knowledge. Specifically, we observed a 
significant Prime × Physics Knowledge interaction in our analysis of 
dispositional attributions, such that the American prime increased 
dispositional attribution among participants who self-reported a great deal of 
physics knowledge, but not among self-rated nonexperts (see Table 4, fifth 
row).

Discussion

The results of Study 2 provide evidence for our hypotheses concerning the 
causal impact of folk physical theories on attributions of physical causality. 
Chinese American participants who received the Asian identity prime, which was 
theorized to activate a contextual folk theory of physics, endorsed 
dispositional causes to a lesser extent, and contextual causes to a greater 
extent, than did participants receiving the American identity prime. The 
success of our priming manipulation supports the notion that the interpretation 
of physical phenomena is guided by knowledge structures acquired through 
experience in one's culture. Contributing further support to the 
knowledge-structure account, we found that priming only affected attributions 
for participants with little formal instruction in physics (see Figures 2 and 
3). If physical attribution is guided by learned, culture-specific knowledge 
structures, then it should be possible to supplant these folk theories with 
formal scientific theories acquired through education.

TABLE 4: Summary of Simultaneous Multiple Regression Analysis 4 of Aggregate 
Dispositional Attribution in Study 2 (N = 65) [dropped]

Figure 2 Level of dispositional attribution for physical events as a function 
of identity prime condition and physics education in Study 2.

Figure 3 Level of contextual attribution for physical events as a function of 
identity prime condition and physics education in Study 2.

An unexpected finding emerged, such that identity priming had an effect for 
individuals rating themselves as relatively expert in physics but not for 
individuals self- reporting little physics knowledge. Given that we had 
intended self-reported physics expertise, like physics education, to gauge the 
extent to which participants have internalized formal physics theories, this 
finding seems to contradict the observed Prime × Physics Classes interaction. 
However, we believe that this finding may have arisen because self-rated 
physics expertise is a less- pure measure of an individual's actual physics 
knowledge than is the number of physics classes he or she has taken. 
Specifically, we believe that the self-rated expertise measure may have been 
confounded with individuals' motivation to self-enhance (i.e., to portray 
themselves in a positive light); self-enhancement motivation, in turn, may have 
been negatively related to participants' level of identification as Asians. 
Because East Asians tend to self-enhance to a lesser degree than do European 
Americans (Heine & Lehman, 1997), it may be that self- rated physics nonexperts 
were less American-identified than were self-rated experts. If this is so, then 
it may make sense that self-rated nonexperts, being not very American 
identified--and thus less likely to possess a dispositional folk physical 
theory--would not have been susceptible to the American prime. It follows from 
priming theory (Higgins, 1996) that one cannot prime a knowledge structure that 
an individual does not possess.

GENERAL DISCUSSION

The current research employed the knowledge- structure conception of culture in 
an examination of cultural influences on perceptions of physical events. First, 
we argued that different cultures instill their members with different folk 
theories of physics. Study 1 provided evidence for this claim: American and 
Chinese individuals were found to differ in their explanations for a number of 
physical events, with Americans favoring dispositional explanations compared to 
Chinese. Second, we argued that folk theories exert a causal influence on 
physical attributions. We tested this claim in Study 2 using a procedure 
designed to temporarily increase the accessibility of dispositional or 
contextual theories in individuals presumed to possess both (i.e., Chinese 
Americans). Chinese Americans whose Asian identity was primed were found to 
endorse dispositional explanations for physical events to a lesser extent, and 
contextual explanations to a greater extent, than did Chinese Americans whose 
American identity was primed.

Our findings concerning the role of formal instruction in physics help to 
reconcile the current results with previous psychological research in which no 
cultural differences in physical attribution were found (e.g., Michotte, 1963). 
In keeping with the folk theories approach, formal education might supplant or 
obscure the operation of folk theories and thus prevent the cultural difference 
from manifesting itself. In Study 1, we argued that cultural differences in 
folk physical theories emerged in part due to the fact that participants had no 
formal physics education. In Study 2, participants' background in physics was 
measured, consistent with the idea that physics instruction blocks the 
operation of folk physical theories, priming effects were found only for 
individuals who had taken few physics classes.

Continuity of Cultural Differences in Social Attribution and Physical 
Attribution

The present research provides evidence for a cultural difference in physical 
attribution analogous to a known cultural difference in social attribution. The 
dispositional- contextual (or internal-external) distinction, used here to 
distinguish between different kinds of attributions for objects' physical 
behavior, has a long history in the study of attributions for individuals' 
social behavior. Researchers studying social explanation often distinguish 
between internal attributions, which trace behavior to personal dispositions 
(e.g., personality traits or attitudes), and external attributions, which trace 
behavior to forces in the social environment (e.g., pressure from peers or 
authorities) (Gilbert & Malone, 1995).

Researchers studying social attribution have argued for the existence of robust 
biases in social explanation. For instance, lay perceivers often have been 
observed to favor internal (dispositional) explanations for others' behavior 
over situational explanations--an inferential tendency known as the 
"correspondence bias" or the "fundamental attribution error" (Ross & Nisbett, 
1991). Although this tendency was once seen as a universal bias in social 
judgment (Heider, 1958; Ichheiser, 1949; Ross, 1977), more recent work in 
cultural psychology has recast dispositional bias as a culture-bound phenomenon 
(e.g., Miller, 1984; for a review, see Peng et al., 2001). Cross-cultural 
research suggests dispositional bias is less marked in East Asian cultures than 
in Western cultures, where most social psychological research has been 
conducted. A growing body of research using a variety of methods has 
demonstrated that East Asians are less apt to attribute behavior to an actor's 
personal dispositions, and more apt to attribute behavior to the situational 
context, than are members of Western cultures (Kitayama & Masuda, 1997; Knowles 
et al., 2001; Lee, Hallahan, & Herzog, 1996; Morris & Peng, 1994).

Analogous to this Asian-Western cultural difference in social attribution, the 
current research suggests that Americans favor internal/dispositional 
explanations for nonsocial events more than do Chinese, whereas Chinese prefer 
external/contextual explanations more than do Americans. Whether the parallel 
between cultural differences in social and physical perception reflect the 
operation of domain-general cognitive factors-- such as dialectical versus 
linear (Peng & Nisbett, 1999) or holistic versus analytic (Nisbett, Peng, Choi, 
& Norenzayan, 2001) modes of thought--is an important question for future 
research.

Reconciling Developmental and Cultural Models of Causal Understanding

At first blush, the current studies might seem at odds with research into the 
development of physical understanding, which points to the existence of 
universal constraints guiding individuals' perceptions of physical events from 
a very early age (Carey & Spelke, 1994; Spelke, 1990). We argue, however, that 
no inherent tension exists between these developmental and cultural 
perspectives. First, the existence of cultural differences among adults in no 
way rules out the existence of universals among infants. Indeed, models of the 
development of social-causal explanation have explicitly included both early 
universals and later cultural differences. For instance, Miller (1984) argued 
that whereas early social inference may be constrained by universal cognitive 
processes, the influence of culture--as carried by folk theories--increases as 
individuals mature within their culture (Miller, 1984). The development of 
physical understanding might follow a similar pattern, in which cultural 
differences emerge only relatively late in development.

Second, as the influence of folk theories on physical perceptions increases 
over development, it need not be the case that universal perceptual and 
cognitive mechanisms stop operating. Indeed, there is no inherent contradiction 
between the types of cognitive constraints identified by developmental 
psychologists (e.g., the innate understanding, observed by Spelke, 1994, that 
two objects cannot occupy the same volume of space) and the types of divergent 
beliefs embodied in folk physical theories (i.e., that the behavior of objects 
is attributable primarily to their dispositions or to forces impinging on them 
from without). In other words, dispositional and contextual folk physics are 
equally consistent with the sorts of basic perceptual constraints identified by 
developmentalists.

Conclusion

The current research contributes to our understanding of how development within 
a particular social milieu (i.e., a culture) molds an individual's perceptions 
of his or her environment. Past research in the culture-astheory tradition 
(e.g., Hong et al., 2000; Morris & Peng, 1994) suggests that culture--both 
national and ethnic-- may profitably be construed as a constellation of folk 
theories governing one's basic understanding of the social world. The current 
research suggests that the influence of culturally instilled folk theories may 
extend further-- specifically, to one's causal understanding of nonsocial 
(i.e., physical) events. At the same time, the current studies place caveats on 
when folk theories can and cannot be expected to exert influence on causal 
attributions. When formal theories in a domain are acquired, the influence folk 
understandings may wane.

NOTES

1. It should be noted that priming techniques have not been limited to research 
in the culture-as-theory tradition. Working within the self approach, Brewer 
and Gardner (1996; Gardner, Gabriel, & Lee, 1999) used linguistic cues to prime 
personal, relational, or collective self-definitions. Value theorists, in turn, 
have primed different cultural values using value-related cues (e.g., Trafimow, 
Triandis, & Goto, 1991).

2. Female spouses of Chinese students were selected due to the difficulty of 
finding Chinese students with no formal education in physics.

3. We see two possible artifactual reasons for the strong positive correlation 
between dispositional and contextual attribution scores. First, participants 
may have differed in the degree to which they saw the physical displays as 
requiring explanation; that is, some participants may have seen many causal 
factors at work in the displays (leading to relatively high ratings for all 
causes), whereas other participants saw only a few factors at work (leading to 
relatively low ratings for all causes). Second, participants may have differed 
in terms of acquiescence bias, leading them to favor either high ratings or low 
ratings across all causal factors. Thus, the positive association between 
dispositional and contextual attribution scores does not invalidate our claim 
that these modes of explanation are distinguishable and independent.

REFERENCES

Aiken, L., & West, S. (1991). Multiple regression: Testing and interpreting 
interactions. Newbury Park, CA: Sage.

Baillargeon, R. (2000). How do infants learn about the physical world? In D. 
Muir & A. Slater (Eds.), Infant development: The essential readings (pp. 
195-212). Malden, MA: Blackwell.

Brewer, M., & Gardner, W. (1996). Who is this "We"? Levels of collective 
identity and self representations. Journal of Personality and Social 
Psychology, 71, 83-93.

Capra, F. (1975). The Tao of physics: An exploration of the parallels between 
modern physics and eastern mysticism. Berkeley, CA: Shambala.

Carey, S. (1991). Knowledge acquisition: Enrichment or conceptual change? In S. 
Carey & R. Gelman (Eds.), Epigenesis of mind: Studies in biology and cognition 
(pp. 257-291). Hillsdale, NJ: Lawrence Erlbaum.

Carey, S., & Spelke, E. (1994). Domain-specific knowledge and conceptual 
change. In L. Hirschfeld & S. Gelman (Eds.), Mapping the mind: Domain 
specificity in cognition and culture (pp. 169-200). New York: Cambridge 
University Press.

Chiu, C., Morris, M., Hong, Y., & Menon, T. (2000). Motivated cultural 
cognition: The impact of implicit cultural theories on dispositional 
attribution varies as a function of need for closure. Journal of Personality 
and Social Psychology, 78, 247-259.

Choi, I., Nisbett, R., & Norenzayan, A. (1999). Causal attribution across 
cultures: Variation and universality. Psychological Bulletin, 125, 47-63.

Gardner, W., Gabriel, S., & Lee, A. (1999). "I" value freedom, but "we" value 
relationships: Self-construal priming mirrors cultural differences in judgment. 
Psychological Science, 10, 321-326.

Gilbert, D., & Malone, P. (1995). The correspondence bias. Psychological 
Bulletin, 117, 21-38.

Heider, F. (1944). Social perception and phenomenal causality. Psychological 
Review, 51, 358-374.

Heider, F. (1958). The psychology of interpersonal relations. New York: John 
Wiley.

Heine, S., & Lehman, D. (1997). Culture, dissonance, and self- affirmation. 
Personality and Social Psychology Bulletin, 23, 389-400.

Heine, S., & Lehman, D. (1999). Culture, self-discrepancies, and self- 
satisfaction. Personality and Social Psychology Bulletin, 25, 915-925.

Higgins, E. (1996). Knowledge activation: Accessibility, applicability, and 
salience. In E. Higgins & A. Kruglanski (Eds.), Social psychology: Handbook of 
basic principles (pp. 133-168). New York: Guilford.

Hofstede, G. (1980). Culture's consequences: International differences in 
work-related values. Beverly Hills, CA: Sage.

Hong, Y., Morris, M., Chiu, C., & Benet-Martinez, V. (2000). Multicultural 
minds: A dynamic constructivist approach to culture and cognition, American 
Psychologist, 55, 709-720.

Ichheiser, G. (1949). Misunderstandings in human relations. Chicago: University 
of Chicago Press.

Kim, I., & Spelke, E. (1999). Perception and understanding of effects of 
gravity and inertia on object motion. Developmental Science, 2, 339-362.

Kim, U., & Choi, S. (1994). Individualism, collectivism, and child development: 
A Korean perspective. In P. M. Greenfield & R. R. Cocking (Eds.), 
Cross-cultural roots of minority child development (pp. 227-257). Hillsdale, 
NJ: Lawrence Erlbaum.

Kitayama, S., & Masuda, T. (1997). Cultural psychology of social inference: The 
correspondence bias largely vanishes in Japan. Unpublished manuscript, Kyoto 
University.

Knowles, E. D., Morris, M., Chiu, C., & Hong, Y. (2001). Culture and the 
process of person perception: Evidence for automaticity among East Asians in 
correcting for situational influences on behavior. Personality and Social 
Psychology Bulletin, 27, 1344-1356.

Kotovsky, L., & Baillargeon, R. (2000). Reasoning about collisions involving 
inert objects in 7.5-month-old infants. Developmental Science, 3, 344-359.

Lee, F., Hallahan, M., & Herzog, T. (1996). Explaining real-life events: How 
culture and domain shape attributions. Personality and Social Psychology 
Bulletin, 22, 732-741.

Lewin, K. (1935). Adynamic theory of personality: Selected papers. New York: 
McGraw-Hill.

Markus, H., & Kitayama, S. (1991). Culture and the self: Implications for 
cognition, emotion, and motivation. Psychological Review, 98, 224-253.

Menon, T., Morris, M., Chiu, C., & Hong, Y. (1999). Culture and the construal 
of agency: Attribution to individual versus group dispositions. Journal of 
Personality and Social Psychology, 76, 701-717.

Michotte, A. (1963). The perception of causality. New York: Basic Books.

Miller, J. (1984). Culture and the development of everyday social explanation. 
Journal of Personality and Social Psychology, 46, 961-978.

Morris, M., Menon, T., & Ames, D. (2001). Culturally conferred conceptions of 
agency: A key to social perception of persons, groups, and other actors. 
Personality and Social Psychology Review, 5, 169-182.

Morris, M., Nisbett, R., & Peng, K. (1995). Causal attribution across domains 
and cultures. In D. Sperber & D. Premack (Eds.), Causal cognition: 
Amultidisciplinary debate (pp. 577-614). New York: Clarendon.

Morris, M., & Peng, K. (1994). Culture and cause: American and Chinese 
attributions for social and physical events. Journal of Personality and Social 
Psychology, 67, 949-971.

Narikiyo, T., & Kameoka, V. A. (1992). Attributions of mental illness and 
judgments about help seeking among Japanese-American and White American 
students. Journal of Counseling Psychology, 39, 363-369.

Needham, J. (1954). Science and civilisation in China (Vol. 4). Cambridge, UK: 
Cambridge University Press.

Nisbett, R., Peng, K., Choi, I., & Norenzayan, A. (2001). Culture and system of 
thoughts: Holistic versus analytic cognition. Psychological Review, 108, 
291-310.

Norenzayan, A., & Nisbett, R. (2000). Culture and causal cognition. Current 
Directions in Psychological Science, 9, 132-135.

Peng, K., Ames, D., & Knowles, E. D. (2001). Culture and human inference: 
Perspectives from three traditions. In D. R. Matsumoto (Ed.), Handbook of 
culture and psychology (pp. 245-264). New York: Oxford University Press.

Peng, K., & Nisbett, R. (1999). Culture, dialectics, and reasoning about 
contradiction. American Psychologist, 54, 741-754.

Ross, L. (1977). The intuitive psychologist and his shortcomings: Distortions 
in the attribution process. In L. Berkowitz (Ed.), Advances in experimental 
social psychology (Vol. 10, pp. 174-221). New York: Academic Press.

Ross, L., & Nisbett, R. (1991). The person and the situation: Perspectives of 
social psychology. New York: McGraw-Hill.

Schwartz, S. (1994). Beyond individualism/collectivism: New cultural dimensions 
of values. In U. Kim & H. Triandis (Eds.), Individualism and collectivism: 
Theory, method, and applications (pp. 85-119). Thousand Oaks, CA: Sage.

Singelis, T. (1994). The measurement of independent and interdependent 
self-construals. Personality and Social Psychology Bulletin, 20, 580-591.

Spelke, E. (1990). Principles of object perception. Cognitive Science, 14, 
29-56.

Spelke, E. (1994). Initial knowledge: Six suggestions. Cognition, 50, 431-445.

Spelke, E. (2000). Core knowledge. American Psychologist, 55, 12331243.

Stewart, J. (1984). Object motion and the perception of animacy. Paper 
presented at the meeting of the Psychonomic Society, San Antonio, TX.

Trafimow, D., Triandis, H., & Goto, S. (1991). Some tests of the distinction 
between the private self and the collective self. Journal of Personality and 
Social Psychology, 60, 649-655.

Triandis, H. (1995). Individualism and collectivism. Boulder, CO: Westview.

Whang, P., & Hancock, G. (1994). Motivation and mathematics achievement: 
Comparisons between Asian-American and non- Asian students. Contemporary 
Educational Psychology, 19, 302-322.

Zukav, G. (1979). The dancing wu li masters: An overview of the new physics. 
New York: Morrow.

Received March 1, 2002 Revision accepted October 15, 2002