Journal of Research in Personality 41 (2007) 731–744

www.elsevier.com/locate/jrp

Superstitious, magical, and paranormal beliefs:

An integrative model
Marjaana Lindeman ¤, Kia Aarnio
Department of Psychology, University of Helsinki, Finland

Available online 4 August 2006

Abstract

Lack of conceptual clarity has hampered theory formation and research on superstitious, mag-
ical, and paranormal beliefs. This study oVers a conceptual framework where these concepts are
diVerentiated from other unfounded beliefs and deWned identically as a confusion of core knowl-
edge about physical, psychological, and biological phenomena. When testing this deWnition with
questionnaire items (N D 239), the results showed that superstitious individuals accepted more vio-
lations of core ontological distinctions than skeptics did and that ontological confusions discrimi-
nated believers from skeptics better than intuitive thinking, analytical thinking, or emotional
instability. The Wndings justify the present conceptualization of superstitious, magical, and para-
normal beliefs, and oVer new theoretical propositions for the familiar everyday beliefs that are yet
scientiWcally so poorly understood.
 2006 Elsevier Inc. All rights reserved.

Keywords: Superstition; Magical thinking; Paranormal beliefs; Intuitive thinking; Core knowledge

1. Introduction

More than 40% of American people believe in devils, ghosts, and spiritual healing
(National Science Foundation, 2002; Rice, 2003). Researchers, in turn, have characterized
superstitious and magical thinking as an extremely discouraging research topic (Scheibe &
Sarbin, 1965), a problem for which there is no ready solution (Campbell, 1996), and as

*
Corresponding author. Fax: +358 9 191 29 443.
E-mail address: [email protected] (M. Lindeman).

0092-6566/$ - see front matter  2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.jrp.2006.06.009
73 M. Lindeman, K. Aarnio / Journal of Research in Personality 41 (2007) 731–744

“a label for a residual category—a garbage bin Wlled with various odds and ends that we
do not otherwise know what to do with” (NemeroV & Rozin, 2000, p. 1).
Superstitious and magical beliefs have certainly been examined, and a body of research
has identiWed numerous correlates for them, ranging from personality traits, motivation,
and cognition to emotional instability, demographics, and social inXuences (reviews:
Vyse, 1997; Zusne & Jones, 1989). Despite the accumulating information, however, the
overall picture of superstition is unclear and remains to be adequately described and
explained.
We suggest that a lack of conceptual clarity has hampered theoretical progress. Because
there is no agreement about what the domain entails, theoretically important assumptions
have been diYcult to make. Most importantly, whether and how the constructs supersti-
tion, magical beliefs, and paranormal (supernatural) beliefs diVer from each other, and
above all, how they diVer from other groundless beliefs (e.g., “whales are Wsh”), is not
known. Consequently, there is a strong need for a conceptual model that clariWes the
mean- ing of magical, paranormal, and superstitious beliefs, and explains why well-
educated Western people still believe in things that seem so irrational. The present study
provides an initial step in this direction. Our aim was to oVer new theoretical propositions,
which serve to deWne the constructs and oVer criteria for their application, and to analyze
whether the deWnition can be empirically justiWed.
Very often, the deWnition of superstition, magical, or paranormal belief is either
omitted or substituted by examples of such beliefs. Nonetheless, some deWnitions can be
found in current scientiWc literature. Among the most inXuential deWnitions of magical
thinking are the laws of sympathetic magic (Frazer, 1922/1963; Tylor, 1871/1974). The
law of contagion holds that things that have once been in contact with each other continue
to act upon each other at a distance after the physical contact has been severed. The law of
similarity holds that superWcial resemblance indicates, or causes, deep resemblance.
During the last two decades, Rozin and NemeroV have conducted a series of path-breaking
studies and showed how the same laws manifest themselves among well-educated Western
adults, for example, as reluctance to wear Hitler’s sweater (NemeroV & Rozin, 2000;
Rozin, Millman, & NemeroV, 1986). However, the laws of sympathetic magic are neither
intended nor suY- cient to cover all superstitious, magical, and paranormal beliefs.
Moreover, as the research- ers themselves note, the distinction between the laws of
magical thinking and reality, for example, between magical contagion and microbial
contamination and between magical similarity and vaccination, is subtle and ambiguous
(Frazer, 1922/1963; NemeroV & Rozin, 2000).
Other authors have deWned superstitious and magical beliefs more widely as false
cogni- tions, for example, as limitations in cognitive processing (Shweder, 1977), beliefs
that are barely articulated (Campbell, 1996), tenets founded on ignorance (Padgett &
Jorgenson, 1982), and as causal beliefs that by conventional standards are invalid (Brugger
& Graves, 1997). Undeniably, deWning the beliefs as erroneous covers all superstitions
and magical beliefs as diverse they may be. But the important question still remains: how
do they diVer from other unfounded beliefs?
A concept that might stand for this diVerence is paranormality. Many researchers have
adopted Broad’s (1953) conceptualization of paranormality as a phenomenon that violates
the fundamental and scientiWcally founded principles of nature. In our view, the deWning
property, and the core that diVerentiates superstitious, magical, and paranormal beliefs
from other unfounded beliefs, can be better derived from recent studies on children’s cog-
nitive development.
1.1. Core knowledge and superstition

According to developmental psychologists, there are three major sorts of knowledge

that determine children’s understanding of the world: intuitive physics, intuitive psychol-
ogy, and with certain reservations, intuitive biology (Goswami, 2002; Wellman & Gelman,
1992, 1998). Part of this knowledge can be characterized as core knowledge in that
children learn it without instruction, it provides a basis for further development, and it is
based on domain specialized learning mechanisms (modules) about ecologically important
entities and events (Carey, 1996; Leslie, 1994; Spelke, 2000). The vast majority of core
knowledge develops by preschool age and it encompasses an intuitive comprehension of
physical, bio- logical, and psychological entities as well as diVerent forms of processes
these entities engage in.
Developmental studies show that core knowledge of physical entities includes the
notion that the world is composed of material objects, which have volume and an indepen-
dent existence in space (reviews: Carey & Spelke, 1994; Spelke, 2000; Wellman &
Gelman, 1992, 1998). The core of intuitive knowledge about psychological entities, in
turn, consists of knowledge that animate beings are intentional agents who have a mind.
By the middle of the second year children understand that animate beings can reciprocate
actions and have a capacity to move and initiate actions without external force (Carey &
Spelke, 1994; Rakison & Poulin-Dubois, 2001; Wellman & Gelman, 1992). In addition,
small children understand that the contents of mind, such as thoughts, beliefs, desires, and
symbols, are not substantial and objective but non-material and mental, and that they do
not have the properties they stand for (Johnson, 2000; Leslie, Friedman, & German, 2004;
Piaget, 1929/ 1951; Wellman, 2002; Wellman & Gelman, 1998). For example, 3- and 4-
year-old children understand that the thought of a dog does not have the material
properties of a dog (Har- ris, Brown, Marriot, Whithall, & Harmer, 1991) and that the
roads in a map do not need to be wide enough for cars (DeLoache, 2002).
As regards biological phenomena, it seems that at least notions like contamination and
healing can be characterized as core knowledge (Wellman & Gelman, 1992). They repre-
sent a species-typical adaptation to the problem of food selection and rejection as well as
illness avoidance (Rozin, 1990; Rozin & Fallon, 1987), and cultures that lack
scientiWc understanding of disease transmission understand it intuitively (Kalish, 1999).
Similarly, 4- year-olds know that abnormal behaviors are not contagious (Keil, 1994) and
they can dis- criminate between contaminated and safe substances despite lack of visible
evidence (Fal- lon, Rozin, & Pliner, 1984).
In addition, children learn early to honor the distinctions of diVerent types of processes
which physical, biological, and psychological entities are engaged in. Small children
under- stand that material objects like balls can aVect one another’s motion only if they
touch (Carey & Spelke, 1994; Wellman & Gelman, 1992), and they believe that physical
events take place because an entity gives force (energy, power) to another entity (Leslie,
1994; Wang, Kaufman, & Baillargeon, 2003). In turn, already young infants understand
people’s actions to result from intentions, and by the age of four, children have learned to
give intentional explanations for people’s actions and non-intentional explanations for
physical processes like gravitation and for biological processes such as how plants get
their coloring (Malle, 1997; Schult & Wellman, 1997; Springer & Keil, 1991; Wellman &
Gelman, 1998).
The same ideas that shape children’s intuitive knowledge about the distinctive
properties of the three ontological categories can be found in superstitions. However, in
superstitions the properties are not limited to one domain but conXated with each other
and applied across categories. As a consequence, amazing entities and processes with
extraordinary properties come into existence.
The fusion renders possible, among others, mental contents that have the attributes of
physical or animate entities, resulting in the possibility that a thought can touch objects
(psychokinesis) and move by itself (telepathy). Similarly, when contagion and healing are
attributed to psychological phenomena we Wnd that Hitler’s personality can spread into
his sweater (NemeroV, 1995; Rozin et al., 1986) and that a curer can heal someone at a
dis- tance by force of his or her thoughts. Also, in this other world, entire good and bad
minds, such as angels and devils, have independent existences and operate as animate
entities by moving and initiating actions without external force. Also symbolic
representations and the material objects they represent are confused in magical thinking, as
Piaget (1929/1951) observed. Hence, symbols are assumed to have the same attributes as
the material things they signify, leading to such phenomena as reluctance to drink from a
cyanide-labeled bot- tle known to include only sucrose–water (Rozin et al., 1986), or
beliefs that a physical object (e.g., foot of a rabbit) that someone has intended to stand for
an emotion (e.g., hap- piness) actually carries happiness in its train.
In the magical world, we also Wnd non-intentional physical and biological events turning
into intentional ones. Intention may be seen to inXuence natural or random events such as in
cases of rainmaking rites and modern craps players who roll the dice gently to coax a low
number and more vigorously to encourage a high number (a review: Vyse, 1997).
Moreover, in superstitions a force is an equally important factor as in lay physics but here
force is regarded as a living and intentional entity. For example, feng shui teaches us that
erroneous furnishings may absorbDvital (lat. life) force that leads us to crimes and divorce,
and astrol- ogers suggest that planets have living energy, which pushes and pulls on human
beings in a purposeful way thus aVecting our personality and well-being. Thus, in
superstitious thinking biological and physical processes are no longer non-intentional but
they are seen as having a purpose, that is, as directed toward certain goals and as being
intentionally caused either by a speciWc or an unspeciWed intentional agent (Bering, 2003;
Kelemen, 1999a).
The above argumentation implies that, despite diVerences in focus, the notions of
super- stition, magical thinking, and paranormal beliefs share the same ontological
confusion and can thus be regarded synonymous. If properties of one ontological category
are borrowed to characterize an entity in another ontological category, the following
statement is not merely a falsehood but it is a category mistake (Carey, 1985; Chi, 1992;
Keil, 1979, 1994). Accordingly, we deWne superstitious, magical, and paranormal beliefs
as category mistakes where the core attributes of mental, physical, and biological entities
and processes are con- fused with each other.
However, all category mistakes are not superstitions. For example, many adults regard
physical force as a material substance (Reiner, Slotta, Chi, & Resnick, 2000), which it is
not. What diVerentiates other category mistakes from superstitions is that in superstitions
the category mistakes always include a confusion of core knowledge, and perhaps
therefore lay people seem to recognize some beliefs as false beliefs and others as
superstitions quite easily and implicitly.
Second, category mistakes can be perceived as superstitions only insofar as the state-
ments are believed to be literally true. Thus, metaphorical and allegorical expressions that
deliberately confuse the properties are not superstitions (e.g., “A well-functioning memory
is a goldmine”).
 While preschool children understand surprisingly much about physical, biological, and
psychological phenomena, they nevertheless make at Wrst the same category mistakes that
can be found in superstitions (Kelemen, 1999a, 1999b; Rosengren, Johnson, & Harris,
2000). However, equating adults’ superstitions with small children’s misconceptions does
not imply that superstitious individuals are cognitively at the level of small children. The
deWnition should be understood through the basic tenet of dual-process theories that peo-
ple have two modes of processing information, intuitive and analytical reasoning, which
rely on diVerent databases and have diVerent rules of operation (Evans, 2003; Pacini &
Epstein, 1999; Sloman, 1996; Stanovich & West, 2000; Sun, 2004). According to the dual-
process theories, analytical processes and rational knowledge do not replace intuitive pro-
cesses and contents as children mature. Rather, both types of processes and knowledge
exist and develop throughout one’s life, and therefore two conXicting beliefs can coexist
in an adult’s mind, one more rational and verbally justiWable (e.g. “Death is Wnal”), the
other operating more automatically and being more resistant to logical arguments (e.g.,
“The soul continues to exist though the body may die”).
In line with our argument that superstitions are based on early intuitions, previous stud-
ies have shown that superstitions are more strongly related to a tendency to rely on intui-
tive thinking than to analytical thinking (Aarnio & Lindeman, 2005; Epstein, Pacini,
Denes Raj, & Heier, 1996; Wolfradt, Oubaid, Straube, BischoV, & Mischo, 1999), and
that the same magical beliefs that are typical to preschool children are implicitly preserved
and activated among adults even though the representations have been devaluated in the
face of more rational knowledge (Subbotsky, 2000, 2001).
To investigate whether the new conceptualization is warranted, we hypothesized that in
comparison to skeptical individuals, superstitious individuals more often assign mental attri-
butes to physical and biological entities and material attributes to mental entities, and con-
fuse intentional processes with non-intentional processes. We also hypothesized that various
types of superstitious, magical, and paranormal beliefs are similarly related to ontological
confusions and that the ontological confusions are positively related to a tendency to rely on
intuitive thinking. Finally, we expected that ontological confusions and intuitive thinking
are more important correlates of superstitions, magical and paranormal beliefs than two
other correlates that have repeatedly been suggested to account for the beliefs, namely low
rational thinking and emotional instability (reviews: Vyse, 1997; Zusne & Jones, 1989).

2. Method

2.1. Participants

Two hundred and thirty-nine Finnish volunteers participated in the study: 96 female
skeptics, 27 male skeptics, 88 female, and 28 male superstitious individuals. Their age
range was 16–47 with a mean of 24.2 years. The great majority (94%) were full-time
students who represented a wide variety of disciplines, including the natural, behavioural,
medical, and social sciences, technique, business and trade, and services.

2.2. Procedure

The participants were recruited from a larger population (ND3261) of participants in a

study on superstition, which was conducted one year earlier. We sent an invitation to those
whose overall superstition scores (see Section 2.3) were among the upper or lower 10% for
their gender. The groups were balanced for gender because in the earlier study, women
(M D2.16) had higher superstition scores than men (M 1.94), D F (1, 3245) 106.44,
p < .001, 4D 2
D password and a hyper-
.032. The recruitment message contained a personal
link to the questionnaire on the Internet. The web survey was implemented as a Java Serv-
let, running on an Apache web server. Due to incomplete contact information, a total of
14% of the participants of the earlier study were eventually invited to take part. Fifty-three
percent of them returned a completed questionnaire within the time limit.

2.3. Measures

2.3.1. Ontological confusions

We constructed measures of ontological confusion by utilizing Chi and her associates’
(Chi, 1992; Chi, Slotta, & de Leeuw, 1994) conceptualizations and assessment methods of
ontological categories and their distinctions.
First, we examined the participants’ conceptions of properties of diVerent ontological
entities by 34 statements in which attributes of one ontological category were assigned to
another one. The participants were asked whether they understood the statements as meta-
phorically or literally true (1 Donly metaphorically; 5 only literally).
In 16 of the statements, material entities (artifacts, liquids, solids, and plants) held men-
tal attributes (e.g., beliefs, desires, and kindness). Example statements include: “Old furni-
ture know things about the past” and “When summer is warm, Xowers want to bloom.”
These items constituted the measure of mentalizing matter (aD.94). Ten of the statements
assessed physicalizing mental (aD.88) and they described mental phenomena (a thought, a
human mind) with attributes of physical matter (e.g., has volume, aVects a material entity
if it touches it). Example statements include “An unstable human mind is disintegrating.”
Six statements assessed biologizing mental D (a .84) and they described mental phenomena
(a thought, a human mind) with attributes of biological entities (e.g., is living, can be
contam- inated), for example, “An evil thought is contaminated.” For comparison, there
were also eight fully metaphorical statements (e.g., “A wailing wind isDa Xute”; a .88) and
four fully literal statements (e.g., “Running water is D Xuid”; a .86).
Confusion between intentional and non-intentional events was analyzed by giving the
participants descriptions of 18 non-intentional events that lead to personally relevant out-
comes, and by asking whether they saw purpose in that event (1 D the event had no purpose,
5D the event clearly had a purpose). Following the work of Chi and her associates (Chi,
1992; Chi et al., 1994), the 18 statements included three types of non-intentional events.
Six of the events were random (e.g., dealing cards in a card game), six were artiWcial
(e.g., a server failure), and six were natural events (e.g., fog). For each set of six events,
positive, negative, and neutral outcomes were described. The positive and negative
outcomes were derived from the Life Experience Survey (Sarason, Johnson, & Siegel,
1978). An example statement of a random event with a negative outcome is: “You play
cards but get only clubs and spades and therefore run into large debts. Was there a purpose
in getting those cards?” An example statement of an artiWcial event with a positive
outcome includes: “The brakes of your car fail and you smash up with a stranger whom
you will later marry. Did the brakes fail for a purpose?” An example statement of natural
events with a neutral out- come is: “A lightning cuts down a big tree in the garden of your
house but no harm is brought to you by this. Did the lightning have a purpose?” Sum
scales of the statements
dealing with random (aD.92), artiWcial (aD.93), and natural events (a .94) were con-
structed by averaging the items. D
For comparison, there were also four statements, which described true intentional
events (e.g., a kiss, running race, and bullying) with positive, neutral, and negative out-
comes (e.g., a beginning of a romantic relationship, reduced working ability). However,
the reliability of this scale was low D
(a .53), possibly because the participants consistently
assessed intentional events with positive outcomes to be more purposeful than the other
events, F (1, 238)D 343.25, p < .001.
Also, a total score for ontological confusions was calculated by averaging all the related
items. The reliability (a) was .92.

2.3.2. Beliefs
Information on participants’ beliefs (and on intuitive and analytical thinking, and emo-
tional instability) was acquired in an earlier study. Most beliefs were measured with the
Revised Paranormal Belief Scale (Tobacyk, 2004), which is a slightly revised version of
the most commonly used measure of superstitious, magical, paranormal, and religious
beliefs (Tobacyk & Milford, 1983). Because the items on the RPBS cover only some
aspects of superstitious and magical beliefs, it was supplemented with a number of items
to cover a wider spectrum of beliefs. The 55 items were measured on a Wve-point
rating
D scale (1 strongly disagree, 5 strongly agree).
Belief in paranormal agents (aD.83) was measured with 14 items on belief in witches,
extraordinary life forms, and extraterrestrial life (e.g., “Ghosts exist”). Belief in the para-
normal abilities of human beings (aD.83) was assessed with 10 items focusing on beliefs in
telepathy, spiritualism, precognition, and psychokinesis (e.g., “A person’s thoughts can
inXuence the movement of a physical object”). Religious beliefsD(a .88) was gauged by
the 4 items in RPBS (e.g., “I believe in God”). Luck beliefs D (a .83) were measured with 9
state- ments about omens of luck, rituals, and amulets (e.g., “Amulets, for instance a
speciWc piece of jewelry, bring good luck”). Belief D in astrology (a .89) was
operationalized with Wve items (e.g., “The position of the stars at the time of birth
D in the claims of feng shui (a .89) was measured with Wve
inXuences personality”). Belief
items (e.g., “Furnishing accord- ing to the principles of feng shui balances your
environment and thus aVects your health and success in a positive way”). In addition, a
mean score of all the itemsDwas used to mea- sure overall superstition (a .96).
Analytical and intuitive thinking were assessed by the Rational-Experiential Inventory
(Pacini & Epstein, 1999). The inventory consists of two 20-item scales (1Dstrongly dis-
agree, 5Dstrongly agree). The Rationality subscale (a .87) assesses the extent to which an
individual engages in and enjoys rational, analytic, eVortful, aVect-free, and logical think-
ing. The Experientiality subscale (a D.88) assesses the extent to which an individual
engages in and enjoys automatic, preconscious, holistic, non-verbal, and associationistic
thinking. Example items include: “I usually have clear, explainable reasons for my deci-
sions” (analytical or rational thinking) and “I believe in trusting my hunches” (experiential
or intuitive thinking).
Emotional instability was measured by the Neuroticism subscale of the Finnish version
of the NEO Five-Factor Inventory (McCrae & Costa, 1987; Pulver, Allik, Pulkkinen, &
Hamalainen, 1995). The subscale consists of 48 Wve-point items (1 strongly D disagree,
5 Dstrongly agree), which measure anxiety, depression, self-consciousness, vulnerability,
impulsiveness, and hostility. The reliability (a) of the scale was .93.
3. Results

The results showed that, as hypothesized, the believers mentalized matter more (MD2.30)
than the skeptics (MD 1.88), F (1,237) D 16.66, p <.001, 42 .07, they physicalized mental
more D
(M D 2.84) than the skeptics (M D 2.34), F (1,D 237) 26.44, p <.001, 42 .10, and they
biologized D
mental more (M D 2.68) than the skeptics (M 2.01),
D F (1, 237)D34.07, p <.001, 42 .13. The
believers did not diVer from the skeptics whenDthey assessed the truth of purely literal state-
ments, F (1,237)D0.15, ns, or the purely metaphorical statements, F (1,237) 1.18, ns.
In addition, the believers also assigned more (MD2.82) purpose to natural events than
the skeptics (M D1.24), F (1, 235)D 197.49, p < .001, 4D 2
.46; more (M 2.75) purpose to
artiWcial events than the skeptics (MD D 1.28), F (1, 235)D 169.53, p < .001, 42 .42; and more
(M D 3.00) purpose to random events than the skeptics D (M 1.40),
D F (1, 235) 200.62,
p < .001, 4D 2
.46. The purpose of intentional events was assessed D equally by the believers
and the skeptics, F (1, 237)D1.57, ns.
The results also showed that the believers relied more on intuitive thinking (MD3.69) than
the skeptics (M D2.99), F (1, 239)D 97.51, p <.001, 42 .29; but less on analytical thinking
D
(M D3.78) than the skeptics (M 4.05),
D F (1, 239) D14.50, p <.001, 42 .06; and that they
D
were emotionally less stable (M D2.62) than the skeptics (M D 2.86), F (1, 239) 9.68, p < .01,
42 D
.04. No gender diVerences were found between the groups, D z2 (1) 0.16, ns. This is intelli-
gible, because the participants were selected from individuals whose overall superstition
scores in the earlier study were among the upper or lower 10% for their gender.
To analyze how the ontological confusions relate to various types of beliefs, intuitive and
analytical thinking, and emotional instability, correlations were obtained. Because only skep-
tics and believers were included in the sample, the originally continuous belief variables were
not normally but bimodally distributed. Therefore, Spearman’s rho, which transforms ratio
level variables into rank-orders and thus linearizes the relationship, was used (Table 1).

Table 1
Spearman’s rank-order correlations (rs) between beliefs, ontological confusions, intuitive and analytical thinking,
and emotional instability
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Beliefs
1. Paranormal agents
2. Paranormal abilities .85
3. Luck beliefs .72 .75
4. Astrology .80 .82 .81
5. Feng shui .70 .72 .70 .75
6. Religious beliefs .78 .78 .69 .75 .67
Confusions
7. Physicalizing mental .37 .40 .30 .31 .35 .33
8. Biologizing mental .41 .41 .37 .41 .42 .43 .75
9. Mentalizing matter .39 .41 .35 .37 .40 .39 .97 .88
10. Purpose in random events .62 .65 .61 .64 .64 .67 .44 .50 .49
11. Purpose in artiWcial events .62 .65 .59 .62 .63 .64 .43 .46 .47 .88
12. Purpose in natural events .66 .69 .61 .66 .66 .68 .42 .48 .48 .90 .93
13. Intuitive thinking .49 .51 .43 .54 .51 .43 .29 .34 .32 .39 .32 .37
14. Analytical thinking ¡.21 ¡.19 ¡.22 ¡.28 ¡.27 ¡.21 ¡.20 ¡.17 ¡.20 ¡.27 ¡.24 ¡.24 ¡.06
15. Emotional instability .22 .22 .23 .28 .22 .21 .10 .12 .11 .18 .16 .21 .18 ¡.33
rs 7 .13, p < .05. rs 7 .17, p < .01. rs 7 .20, p < .001.
 Table 2
Standardized discriminant function coeYcients and pooled within-groups correlations (structure coeYcients) for
prediction of membership in superstitious or skeptic individuals
Variable Discriminant function coeYcients Structure coeYcients
Ontological confusions .74 .82
Intuitive thinking .56 .63
Anaytical thinking ¡.16 ¡.24
Emotional instability .06 ¡.20

To Wnd the best function that separates believers and skeptics, and to compare the
importance of the predictors of group membership, a standard discriminant function anal-
ysis was performed using the total score of ontological confusions, intuitive and analytical
thinking, and emotional instability as predictors of group membership. One discriminant
function was calculated, with a z2 (4) D168.69, p < .001 (Table 2). The loading matrix of
correlations between predictors and the discriminant function suggests that the best pre-
dictor for distinguishing between superstitious and skeptics were ontological confusions,
followed by intuitive thinking. Other predictors were not important because, by conven-
tion, correlations lower than .33 are not considered eligible (Tabachnick & Fidell, 1996).
The loadings reveal only the full correlation between the discriminant function and the
predictors, and the variance shared with ontological confusions and group membership
may thus be predictable from other predictors. The structure coeYcients, in turn, show the
unique contribution the predictors make. As can be seen from Table 2, after adjustment for
other predictors, ontological confusions are still the most important predictors of being a
skeptic or a believer.

4. Discussion

Historically, researchers have considered beliefs in superstition, magic, and the paranor-
mal in some indeWnite way both as distinct and interrelated phenomena. Also, the deWni-
tions of the concepts have repeatedly been insuYcient. The present study oVered a uniWed
conceptual framework where the concepts were diVerentiated from other unfounded
beliefs and deWned identically as an ontological confusion between the core attributes of
mental, physical, and biological entities and processes. In addition, we examined whether
superstitious individuals confuse the attributes of ontological categories, as the deWnition
suggests.
The results gave justiWcation for the deWnition. Compared with the skeptics, the super-
stitious individuals assigned more physical and biological attributes to mental phenomena.
Thus, they understood such notions as a mind that can touch objects and an evil thought
that may be contaminated more literally than the skeptics. Superstitious individuals also
assigned more mental attributes to water, furniture, rocks, and other material things than
skeptics did and accepted that entities like these may—literally, not only metaphorically—
have psychological properties such as desires, knowledge, or a soul. In particular, supersti-
tious individuals saw natural, random and artiWcial (i.e., non-intentional) events like fog
or a server failure as having a purpose when the processes had led to episodes with a
person- ally relevant outcome such as a marriage. The believers did not, however, di Ver
from the skeptics when they assessed the literal truth of purely literal or purely
metaphorical state- ments, or the purposefulness of truly intentional acts like kissing.
74 M. Lindeman, K. Aarnio / Journal of Research in Personality 41 (2007) 731–744

The results also showed that various manifestations of the beliefs, for example beliefs in
astrology, feng shui and paranormal abilities of human beings, were associated with onto-
logical confusions and with higher intuitive thinking, and—albeit only slightly—to lower
analytical thinking and emotional instability. Because the believer and skeptic groups were
beforehand balanced for gender, the eVect of gender was not analyzed. However, it should
be noted that women’s lower analytical and higher intuitive thinking have been shown to
be the generative mechanisms for women’s higher endorsement of paranormal beliefs
com- pared to men (Aarnio & Lindeman, 2005).
The discriminant analysis indicated that the best measures to distinguish believers from
skeptics were ontological confusions, and secondarily intuitive thinking. Neither analytical
thinking nor emotional stability could discriminate the groups from each other. These
results support the argument that superstitions and other paranormal beliefs arise from the
intuitive system and not from a malfunctioning analytical system and are in line with the
earlier Wndings that people who rely more on intuitive thinking hold more superstitions
than others (Epstein et al., 1996; Wolfradt et al., 1999). When assessing the importance of
the measures’ unique contribution, after variance associated with other predictors had
been removed, susceptibility to ontological confusions remained the most important pre-
dictor of group membership. As a whole, then, the results are in line with the notion that
ontological confusions are deWning properties of superstitions, magical, and paranormal
beliefs.
The new deWnition of superstition has evident beneWts compared with the earlier deWni-
tions. For one thing, some of the earlier deWnitions have been too wide in that
superstitions have been regarded equal to any unsupported notion. By pointing out the
category mistake of core knowledge, the present deWnition enables us to identify how
superstitions diVer from other unfounded beliefs. Accordingly, many beliefs that have
previously been regarded as paranormal, magical, or superstitious are simply
unsubstantiated beliefs, not superstitions. These include, for example, belief in
graphology or biorhythms. Likewise, beliefs obeying the laws of contagion and
similarity are here regarded as superstitions only insofar as the idea of contagion is
stretched beyond the biological domain and similarity is used to draw inferences about
entities from diVerent ontological domains. Thus, disgust towards a piece of clothing
worn by a sick person is not a superstition whereas disgust towards clothes worn by
Hitler is. Similarly, reasoning that whales are Wsh because of simi- lar appearance and
habitats is simply a false belief whereas a belief that needles stuck in a doll cause pain
in the person the doll represents is a superstition.
On the other hand, so far no deWnition has been wide enough to cover all types of
super- stitions, magical and paranormal beliefs. The present deWnition, in contrast, makes
it possi- ble to understand that the beliefs can manifest itself in endless ways. They can
range from primeval animism to the trendy (albeit ancient) idea that furniture and
buildings have liv- ing energy, from children’s beliefs that the moon is an animate being to
educated adults’ beliefs in astrology, and from animate paintings in Harry Potter’s milieu
to the magical ideas about contamination and external forces found among patients
suVering from obses- sive-compulsive disorder and schizophrenia. We suggest that
confusion of core knowledge is the common denominator for all these, and that the
prevailing culture as well as believ- ers’ explicit knowledge and other mental resources
shape the more speciWc contents and manifestations of the beliefs.
In our view, the new conceptualization enables researchers to make more elaborated
theoretical statements regarding superstition. So far, theoretical arguments have lacked
statements speciWc to superstitions. For example, one of the main reasons for the
existence of superstitions has traditionally been taken to be people’s search for causal
explanations and organizing the world in a meaningful and consistent fashion to impose
order and pre- dictability on it (Malinowski, 1948/1992; Tambiah, 1990). However,
because the same rea- sons also apply to numerous other endeavors, such as scientiWc
work, the explanatory power of these statements for superstition has been weak.
The present study suggests that future theory formation and studies on superstition
might beneWt from addressing intuitive thinking and its knowledge base. Subsequent
eVorts may provide more powerful indications that superstitious individuals’ knowledge
about the world is inaccurate in that their early, as yet undeveloped intuitive conceptions
about psychological, biological, and physical phenomena have retained their autonomous
power and co-exist side by side with later acquired rational knowledge. While the role of
insuYcient or incorrect knowledge in superstitions has been acknowledged for long (Mali-
nowski, 1948/1992; Piaget, 1929/1951), it is surprising that so far no research attention has
been paid to superstitious individuals’ knowledge about entities and processes in the
world. In short, whereas developmental psychology research has focused on the strength
of core knowledge among children, research on superstition would beneWt from
studies that address the vulnerabilities of this knowledge among adults.
In addition, future research might attempt to analyze whether superstitions can be
understood in terms of a common essence. There is increasing evidence that children’s
con- ception about what unites members of the same category (e.g., dogs and cats) and
what diVerentiates members of one category from another (e.g., toys and dogs) is based on
the notion of an essence (Gelman, 2004; Johnson & Harris, 1994). Similarly, Meigs (1984)
has suggested that a central theme in the versatile magical beliefs found among the Hua
people in New Guinea is a common, vital essence, nu. This common essence parallels the
view that in superstitions, there is typically an interconnected cosmos, a fundamental
relation between a part and a whole, a human being and a universe, and a single event and
the future (Malinowski, 1948/1992; Piaget, 1929/1951; Werner, 1948). Hence, it might be
hypothesized that confusion between the core properties of ontological categories implies
a notion of a common essence between the categories and thus leads to thinking in terms
of connections and undivided totalities.
In many ways, the results are only preliminary. Most importantly, ontological confu-
sions should be examined much more thoroughly in future studies. In addition, only verb-
alizable conceptions about the ontological categories of physical, biological, and mental
phenomena were examined. Although the results showed that at least part of these concep-
tions are explicit and can be tapped with a questionnaire, our assumption is that majority
of this knowledge is outside of conscious awareness, and are thus held at diVerent levels
of explicitness, ranging from spontaneous and indistinct “as if” feelings to explicit beliefs
(NemeroV & Rozin, 2000; Subbotsky, 2001). Therefore, in future studies, ontological con-
ceptions should be examined with research methods focusing on implicit knowledge.
Moreover, only believers and skeptics were included in the study. We focused on
extreme groups because we were concerned about the possibility that the middle group
consists of habitual believers, that is, of individuals who actually do not believe in
supersti- tions but take them as entertainment, or as Vyse (1997) have suggested, as habits
that help pass the time. If this were the case, the middle group would be qualitatively
diVerent than the extreme groups. However, this is only speculation, and the results need
replication on samples more representative of the general population. Finally, while the
use of
correlational design is justiWed in research in its initial stages, future development of the
propositions will require experimental designs.
Despite these limitations, we have provided a conceptual framework that integrates the
existing heterogeneous deWnitions, links prior and present Wndings to theory, and opens
avenues for future studies on the enchantment of superstition, magic, and the paranormal.
What has been established is an initial understanding that it may be a question of vague
category boundaries, as Woody Allen (1980, p. 15) already has found out: “There is no
question that there is an unseen world. The problem is how far it is from midtown and how
late is it open?”

Acknowledgments

We thank Tarmo Toikkanen and Marieke Saher for assistance. This study was sup-
ported by a grant from the Academy of Finland (200828).

References

Aarnio, K., & Lindeman, M. (2005). Superstition, education and thinking styles. Personality and Individual DiVer-
ences, 39, 1227–1236.
Allen, W. (1980). Complete prose. London: Picador.
Bering, J. M. (2003). Towards a cognitive theory of existential meaning. New Ideas in Psychology, 21, 101–120.
Broad, C. D. (1953). Religion, philosophy, and psychical research. New York: Harcourt & Brace.
Brugger, P., & Graves, R. E. (1997). Testing vs. believing hypotheses: magical ideation in the judgment of contin-
gencies. Cognitive Neuropsychiatry, 2, 251–272.
Campbell, C. (1996). Half-belief and the paradox of ritual instrumental activism: a theory of modern superstition.
British Journal of Sociology, 47, 151–166.
Carey, S. (1985). Conceptual change in childhood. London: MIT Press.
Carey, S. (1996). Cognitive domains as modes of thought. In D. Olson & N. Torrance (Eds.), Modes of thought
(pp. 187–215). Cambridge: Cambridge University Press.
Carey, S., & Spelke, E. (1994). Domain-speciWc knowledge and conceptual change. In L. A. Hirschfeld & S. A. Gel-
man (Eds.), Mapping the mind. Domain speciWcity in cognition and culture (pp. 169–200). Cambridge: Cam-
bridge University Press.
Chi, M. T. H. (1992). Conceptual change within and across ontological categories: examples from learning and
discovery in science. In R. N. Giere (Ed.), Cognitive models of science (pp. 129–186). Minneapolis, MN: Uni-
versity of Minnesota Press.
Chi, M. T. H., Slotta, J. D., & de Leeuw, N. (1994). From things to processes: a theory of conceptual change for
learning science concepts. Learning and Instruction, 4, 27–43.
DeLoache, J. S. (2002). Early development of the understanding and use of symbolic artifacts. In U. Goswami
(Ed.), Blackwell handbook of childhood cognitive development (pp. 207–226). Oxford: Blackwell.
Epstein, S., Pacini, R., Denes Raj, V., & Heier, H. (1996). Individual diVerences in intuitive-experiential and
ana- lytical–rational thinking styles. Journal of Personality and Social Psychology, 71, 390–405.
Evans, J. S. B. T. (2003). In two minds: dual-process accounts of reasoning. Trends in Cognitive Science, 7, 454–
459.
Fallon, A. E., Rozin, P., & Pliner, P. (1984). The child’s conception of food: the development of food rejections
with special reference to disgust and contamination sensitivity. Child Development, 55, 566–575.
Frazer, J. G. (1922/1963). The golden bough. A study in magic and religion. New York: Macmillan.
Gelman, S. A. (2004). Psychological essentialism in children. Trends in Cognitive Science, 8, 404–409.
Goswami, U. (Ed.). (2002). Blackwell handbook of childhood cognitive development. Oxford: Blackwell.
Harris, P. L., Brown, E., Marriot, C., Whithall, S., & Harmer, S. (1991). Monsters, ghosts and witches: testing the lim-
its of the fantasy-reality distinction in young children. British Journal of Developmental Psychology, 9, 105–123.
Johnson, C. N. (2000). Putting diVerent things together: the development of metaphysical thinking. In K. S.
Rosengren, C. N. Johnson, & P. L. Harris (Eds.), Imagining the impossible. Magical, scientiWc and religious
thinking in children (pp. 179–211). Cambridge: Cambridge University Press.
 M. Lindeman, K. Aarnio / Journal of Research in Personality 41 (2007) 731– 743
744
Johnson, C. N., & Harris, P. L. (1994). Magic: special but not excluded. British Journal of Developmental Psychol-
ogy, 12.
Kalish, C. (1999). What young children’s understanding of contamination and contagion tells us about their
concepts of illness. In M. Siegal & C. C. Peterson (Eds.), Children’s understanding of biology and health (pp. 99–
130). Cambridge: Cambridge University Press.
Keil, F. C. (1979). Semantic and conceptual development: An ontological perspective. Cambridge, MA: The MIT
Press.
Keil, F. C. (1994). The birth and nurturance of concepts by domains: the origins of concepts of living things. In
L. A. Hirschfeld & S. A. Gelman (Eds.), Mapping the mind. Domain speciWcity in cognition and culture (pp. 234–
254). Cambridge: Cambridge University Press.
Kelemen, D. (1999a). Function, goals and intention: children’s teleological reasoning about objects. Trends in
Cognitive Science, 3, 461–468.
Kelemen, D. (1999b). Why are rocks pointy? Children’s preference for teleological explanations of the natural
world. Developmental Psychology, 35, 1440–1452.
Leslie, A. M. (1994). ToMM, ToBY, and Agency: core architecture and domain speciWcity. In L. A. Hirschfeld &
S. A. Gelman (Eds.), Mapping the mind. Domain speciWcity in cognition and culture (pp. 119–148). Cambridge:
Cambridge University Press.
Leslie, A. M., Friedman, O., & German, T. P. (2004). Core mechanisms in ’theory of mind’. Trends in Cognitive
Science, 8, 528–533.
Malinowski, B. (1948/1992). Magic, science and religion. Prospect Heights, Ill.: Waveland Press.
Malle, B. F. (1997). The folk concept of intentionality. Journal of Experimental Social Psychology, 33, 101–121.
McCrae, R. R., & Costa, P. T. (1987). Validation of the Wve-factor model of personality across instruments and
observers. Journal of Personality and Social Psychology, 52, 81–90.
Meigs, A. S. (1984). Food, sex, and pollution. A New Guinea religion. New Jersey: Rutgers University Press.
National Science Foundation (2002). Science and engineering: Indicators 2002. Science and technology: public
attitudes and public understanding. Science Wction and pseudoscience. Retrieved March 17, 2005, from the
World Wide Web: http://www.nsf.gov/sbe/srs/seind02/c7/c7s5.htm.
NemeroV, C. J. (1995). Magical thinking about illness virulence: conception of germs from “safe” versus
“danger- ous” others. Health Psychology, 14, 147–151.
NemeroV, C., & Rozin, C. (2000). The makings of the magical mind. The nature and function of sympathetic mag-
ical thinking. In K. Rosengren, C. Johnson, & P. Harris (Eds.), Imagining the impossible: The development of
magical, scientiWc, and religious thinking in contemporary society (pp. 1–34). Cambridge: Cambridge University
press.
Pacini, R., & Epstein, S. (1999). The relation of rational and experiential information processing styles to person-
ality, basic beliefs, and the ratio-bias phenomenon. Journal of Personality and Social Psychology, 76, 972–987.
Padgett, V. R., & Jorgenson, D. O. (1982). Superstition and economic threat: Germany, 1918–1940. Personality
and Social Psychology Bulletin, 8, 736–741.
Piaget, J. (1929/1951). The child’s conception of the world. London: Routledge and Kegan.
Pulver, A., Allik, J., Pulkkinen, L., & Hamalainen, M. (1995). A big Wve personality inventory in two non-Indo-
European languages. European Journal of Personality, 9, 109–124.
Rakison, D., & Poulin-Dubois, D. (2001). Developmental origin of the animate–inanimate distinction. Psycholog-
ical Bulletin, 127, 209–228.
Reiner, M., Slotta, J. D., Chi, M. T. H., & Resnick, L. B. (2000). Naive physics reasoning: a commitment to sub-
stance-based conceptions. Cognition and Instruction, 18, 1–34.
Rice, T. (2003). Believe it or not: religious and other paranormal beliefs in the United States. Journal for the Scien-
tiWc Study of Religion, 42, 95–106.
Rosengren, K. S., Johnson, C. N., & Harris, P. L. (Eds.). (2000). Imagining the impossible. Magical, scientiWc and
religious thinking in children. Cambridge: Cambridge University Press.
Rozin, P. (1990). Development in the food domain. Developmental Psychology, 26, 555–562.
Rozin, P., & Fallon, A. E. (1987). A perspective on disgust. Psychological Review, 94, 23–41.
Rozin, P., Millman, L., & NemeroV, C. (1986). Operation of the laws of sympathetic magic in disgust and other
domains. Journal of Personality and Social Psychology, 50, 703–712.
Sarason, J. G., Johnson, J. H., & Siegel, J. M. (1978). Assessing the impact of life changes: development of the Life
Experience Survey. Journal of Consulting and Clinical Psychology, 46, 932–946.
Scheibe, K. E., & Sarbin, T. R. (1965). Towards a theoretical conceptualisation of superstition. British Journal for
the Philosophy of Science, 16, 143–158.
Schult, C. A., & Wellman, H. M. (1997). Explaining human movements and actions: children’s understanding of
the limits of psychological explanation. Cognition, 62, 291–324.
Shweder, R. A. (1977). Likeness and likelihood in everyday thought: magical thinking in judgments about person-
ality. Current Anthropology, 18, 637–658.
Sloman, S. A. (1996). The empirical case for two systems of reasoning. Psychological Bulletin, 119, 3–22.
Spelke, E. (2000). Core knowledge. American Psychologist, 55, 1232–1233.
Springer, K., & Keil, F. C. (1991). Early diVerentiation of causal mechanisms appropriate to biological and nonbi-
ological kinds. Child Development, 60, 637–648.
Stanovich, K. E., & West, R. F. (2000). Individual diVerences in reasoning: implications for the rationality debate?
Behavioral and Brain Sciences, 23, 645–726.
Subbotsky, E. V. (2000). Phenomenalistic reality: the developmental perspective. Developmental Review, 20, 438–
474.
Subbotsky, E. (2001). Causal explanations of events by children and adults: can alternative causal models coexist
in one mind? British Journal of Developmental Psychology, 19, 23–46.
Sun, R. (2004). Desiderata for cognitive architectures. Philosophical Psychology, 17, 341–373.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics. New York:
HarperCollins.
Tambiah, S. J. (1990). Magic, science, religion, and the scope of rationality. Cambridge: Cambridge University
Press.
Tobacyk, J. J. (2004). A revised paranormal belief scale. The International Journal of Transpersonal Studies, 23, 94–
99.
Tobacyk, J., & Milford, G. (1983). Belief in paranormal phenomena: assessment instrument development and
implications for personality functioning. Journal of Personality and Social Psychology, 44, 1029–1037.
Tylor, E. B. (1871/1974). Primitive culture: Research into the development of mythology, philosophy, religion, art,
and custom. New York: Gordon Press.
Vyse, S. A. (1997). Believing in magic: The psychology of superstition. New York: Oxford University Press.
Wang, S., Kaufman, L., & Baillargeon, R. (2003). Should all stationary objects move when hit? Developments in
infants’ causal and statistical expectations about collision events. Infant Behavior and Development, 26, 529–
567.
Wellman, H. M. (2002). Understanding the psychological world: developing a theory of mind. In U. Goswami
(Ed.), Blackwell handbook of childhood cognitive development (pp. 167–187). Oxford: Blackwell.
Wellman, H. M., & Gelman, S. A. (1992). Cognitive development: foundational theories of core domains.
Annual Review of Psychology, 43, 337–375.
Wellman, H. M., & Gelman, S. A. (1998). Knowledge acquisition in foundational domains. In D. Kuhn & R. S.
Siegler (Eds.), Handbook of child psychology. Cognition, perception, and language (Vol. 2, pp. 523–573). New
York: Wiley.
Werner, H. (1948). Comparative psychology of mental development. New York: International Universities Press.
Wolfradt, U., Oubaid, V., Straube, E. R., BischoV, N., & Mischo, J. (1999). Thinking styles, schizotypal traits and
anomalous experience. Personality and Individual DiVerences, 27, 821–830.
Zusne, L., & Jones, W. H. (1989). Anomalistic psychology: A study of magical thinking (2nd ed.). Hillsdale, NJ: Erl-
baum.