Religion, Brain & Behavior

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/rrbb20

A cognitive account of manipulative sympathetic

magic

Ze Hong

To cite this article: Ze Hong (2022): A cognitive account of manipulative sympathetic magic,
Religion, Brain & Behavior, DOI: 10.1080/2153599X.2021.2006294

To link to this article: https://doi.org/10.1080/2153599X.2021.2006294

Published online: 24 Jan 2022.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=rrbb20
RELIGION, BRAIN & BEHAVIOR
https://doi.org/10.1080/2153599X.2021.2006294

A cognitive account of manipulative sympathetic magic

Ze Hong
Department of Human Evolutionary Biology, Harvard University, Cambridge, MA, USA

ABSTRACT ARTICLE HISTORY

Frazer’s theory of sympathetic magic has been extremely influential in Received 20 July 2021
both anthropology and comparative religion, yet the manipulative Accepted 2 September 2021
aspect has not been adequately theorized. In this paper, I formalize
KEYWORDS
sympathetic magical action and offer a naturalistic explanation of Sympathetic magic; causal
manipulative sympathetic magic by attributing it to a combination of cognition; spatial
environmental regularities (i.e., things that are similar and/or physically autocorrelation; evolutionary
proximate tend to co-vary) and human causal cognition (i.e., the psychology
tendency to mistake correlation as causation), and supply ample
ethnographic and historical evidence for my arguments. In doing so I
also specify the variables involved and re-classify sympathetic magic
into four distinct types for analytic convenience.

1. Introduction
More than a century ago, the pioneering anthropologist James Frazer1 published his seminal work
The Golden Bough which would shape the study of comparative religion and cultural anthropology
thereafter (Hanegraaff, 1998), and much work in anthropology has been devoted to the discussion
of his ideas (Horton, 1993; Mauss, 1902/2001; Tambiah, 1990). In The Golden Bough, Frazer (1890,
p. 19b) explicitly formulates two principles of magic:
First, that like produces like, or that an effect resembles its cause; and second, that things which have once been
in contact with each other continue to act on each other at a distance after the physical contact has been
severed.

The first principle is referred to as the Law of Similarity and the second is the Law of Contact or
Contagion, which are collectively termed “sympathetic magic.” Although Frazer does not formally
define the two laws (for example, he never fully specifies what “act[ing] on each other” means), The
Golden Bough provides ample ethnographic examples to illustrate these principles.
Frazer’s definition of “magic” has been extremely influential and has inspired countless scholarly
discussions (Hanegraaff, 1998). We should note, however, that Frazer’s theorizing of magic
occurred over a century ago and he apparently had a particular agenda in mind when writing
The Golden Bough (Strenski, 2006). Specifically, Frazer (along with Edward Tylor whom he greatly
admired) rejected Christianity and treats magic and religion as separate developmental stages of
human societal evolution (Bremmer, 1999). Magic, Frazer contends, is humans’ most primitive
attempt at manipulating the world, which is to be substituted by religion (belief in an omnipotent
deity) and eventually science. Such stagist views of human social and cultural evolution have been
widely criticized and are now largely rejected in anthropology (Kundt, 2017), but the resulting cat-
egorization of magic and religion into different types continue to exert significant influence and is
still the subject of much discussion among contemporary researchers (Hanegraaff, 1998). For many,

CONTACT Ze Hong ze_hong@g.harvard.edu

© 2021 Informa UK Limited, trading as Taylor & Francis Group
2 Z. HONG

his exclusion of supernatural agency from the category “magic” is problematic, as much of what we
normally consider “magic” practices today are directed towards supernatural agents. Here I do not
wish to engage in the debate of what the proper definition of “magic” should be, as it is a truly
thorny issue (Lindeman & Svedholm, 2012) and beyond the scope of this paper. Rather, I will
focus on the types of magical practices defined by Frazer, i.e., sympathetic magic based on the prin-
ciples of similarity and contagion.
Of course, one may question whether Frazerian sympathetic magic is a “natural kind” at the phi-
losophical level; I suggest, however, that Frazer at the very least identified two analytically useful
principles of human thought and action that deserve a serious (re)analysis. Note that the fact
that Frazer may have been wrong2 about the developmental stages of human social evolution
does not mean we should abandon the concept of sympathetic magic altogether, and I argue that
a fresh theoretical examination of sympathetic magic is important for three reasons. First, while
the name of James Frazer has been a “embarrassment” for many anthropologist today (Strenski,
2006, p. 65) and his work not talked about very much, his definition of magic as not involving
supernatural agents and thus belonging to a different category from religion has persisted well
into the present day, indicating that many researchers likely have found it a useful category to ana-
lyze and compare cultural practices. Figure 1 shows the frequency of the phrase “sympathetic
magic” in Google’s English text corpora (Ngram) from 1800 to 2019, and we can clearly observe
a temporal trend here: the frequency of the term “sympathetic magic” has increased dramatically
after the publication of The Golden Bough in 1890 and remained at rather high levels ever since.3
Second, cultural practices that derive from such principles have been repeatedly observed in tra-
ditional, small-scale societies that lead some early theorists to conclude that sympathetic magic is
a universal feature of human societies (Mauss, 1902/2001). Though systematic meta-analysis on
the prevalence of sympathetic magic has not been conducted yet, a simple keyword search in
Human Relations Area Files (a comprehensive database on human societies and cultures primarily
in the form of annotated ethnography) show that a substantial proportion of ethnographic records
contain the key phrase “sympathetic magic” (125 out of 359 total cultures in the database). Finally,
the psychological intuitions behind sympathetic magic have been shown in contemporary Ameri-
cans (Hood et al., 2010; Rozin et al., 1986), further suggesting that these principles may be quite
fundamental to human cognition.
Note that the majority of the magical practices described in The Golden Bough are manipulative;
that is, the magician believes that he can produce an effect by imitating it (Similarity) and that what-
ever he does to an object will affect equally the person with whom the object was once in contact
(Contagion). For example, people in Haiti would create voodoo dolls and symbolically insert pins
into it in order to inflict harm on whom the doll represents (Armitage, 2015), and the Malays would
use the body parts of the intended victim (nails, hair, etc.) to create a figure of the intended victim

Figure 1. Frequency of the key phrase “sympathetic magic” from 1800 to 2019 using Google’s Ngram Viewer.
 RELIGION, BRAIN & BEHAVIOR 3

with wax from a bee’s comb, and scorch the figure in order to cause harm to the actual victim (Fra-
zer, 1890). If we focus on the manipulative aspect of sympathetic magic, the two laws can be
specified as the following:
Law of similarity: If two objects A and B are similar in their properties, then a change in A will cause a cor-
responding change in B.

Law of contagion: If two objects A and B were once in contact, then a change in A will cause a corresponding
change in B, even if A and B are no longer in contact.

To a modern reader, these principles and their associated practices may appear striking, as the mod-
ern mechanistic/materialistic worldview does not allow mystical causations (Murdock, 1980), in
particular, action at a distance without a proper mechanistic explanation4 (Williamson, 2011). In
the classic anthropological literature, the psychological reason for such objectively faulty laws is
usually attributed to a mistake in the association of ideas. Mauss (1902/2001) puts it
straightforwardly:
E. B. Tylor and others after him have noticed that these laws are none other than the association of ideas, with
one difference, that here the subjective association of ideas leads to the conclusion that there is an objective
association of facts, or in other words that the fortuitous connexion between thoughts is equivalent to the cau-
sal connexion between things. (p. 79)

Here, Mauss, in summarizing the conclusions of previous thinkers, points out that the laws of simi-
larity and contagion exist because the thoughts of objects that are similar or in contact are naturally
associated, and the mistake of the magician (and those who believe in magic) is that they treat this
thought association as a real association with causal properties. The most extreme form of such a
mistake would of course be to treat the image as equal to the real object, and the part as equal to the
whole. As such, whatever one does to the image affects the real object, and the part affects the whole.
Elegant as it seems, this classical account suffers two difficulties. First, in the jargon of evolutionary
biology, this classic explanation is “proximate” (Tinbergen, 1963) in that it only tells us how people
mistake the image/part as the real/whole, but not why people would make such cognitive mistake
from adaptive/functional perspectives. Second, it is not particularly cognitively difficult to recognize
that an image of a person or his body parts and the person himself are different entities, and to my
knowledge, there is no developmental or ethnographic evidence showing that people confuse the
identity of objects at such a fundamental level.
More recently, psychologists and cognitive scientists, most notably Paul Rozin and colleagues
(Nemeroff & Rozin, 2000; Rozin et al., 1986; Rozin & Nemeroff, 1990, 2012), have sought to
show that the psychology that sustains such “magical thinking” also exists in contemporary societies
and have provided functional and adaptive rationale for Frazer’s sympathetic magic. For example,
they find that people in contemporary American societies show disgust to replicas of disgusting
objects; in a series of experiments, Rozin et al. (1986) showed that most participants exhibited a
preference for a normally shaped piece of fudge over fudge shaped like dog feces. Such effect has
been demonstrated at the physiological level: US adults exhibit more electrodermal activity when
a photograph of a sentimental object is destroyed despite maintaining that photographs have no
physical connection with the real objects (Hood et al., 2010). The explanation offered is that simi-
larity magic is related to the principle of generalization, i.e., treating objects that share some prop-
erties as potentially sharing more properties, and the tendency to generalize is usually very useful to
the survival of our species (Nemeroff & Rozin, 2000). In a similar vein of argument, Rozin and col-
leagues suggest that the law of contagion, when conceived as transfer of properties, makes rather
good adaptive sense: microbes, for example, do travel from one body to other through physical con-
tact and cause illness, and our disgust towards contaminated food for fear of getting ill either con-
sciously or unconsciously may extend to other domains, indeed anything that we perceive
negatively. College undergraduates, for example, would strongly prefer to not wear a sweater
worn by someone who experienced a misfortune (e.g., an amputated leg) (Rozin et al., 1994).
4 Z. HONG

The above accounts have contributed much to our understanding of the nature and functions of
sympathetic magic, but they largely ignore the manipulative aspect, i.e., why would change in one
object cause change in another object that is/was in physical contact with or similar to it? More
recently, Rozin et al. (2018) have attempted to address this question and termed the contagious
kind “backward magical contagion.” They showed that American Mturkers express discomfort
when some of their personal items (e.g., hair, signature photocopy, travel diary) is possessed by
a negative third party (rapist or enemy), yet reach no firm conclusions regarding why such phenom-
enon occurs.5
In this paper, I aim to show that the manipulative aspect (Paul Rozin’s “backward causation”)
may be explained by a combination of certain regularities of the world (i.e., things that are similar
or physically closer tend to co-vary in their properties) and the ordinary human cognitive error of
mistaking correlation as causation (Bleske-Rechek et al., 2015; Kida, 2009; Stanovich, 2009). In the
rest of the paper, I first formalize the manipulative aspect of sympathetic magic and then provide a
verbal argument of how the environment and human cognitive processes collectively produce
manipulative sympathetic magic. In doing so, I also propose two other types of magical principles
that have not been formally proposed and have received relatively little attention from researchers,
i.e., that physical proximity and similarity may induce each other as a result of the positive spatial
autocorrelations in nature, as well as providing relevant ethnographic and historical examples.
Finally, I discuss the plausibility and limitations of this account and offer some alternative
possibilities.

2. Mathematical formulation of Frazer’s sympathetic magic

Suppose there are a grid of objects in a two-dimensional space. Each object has two sets of attri-
butes, location (x, y) and properties (p1 , p2 , p3 , . . . , pn ). The distance between any two objects Oi
and Oj is thus the standard Euclidean distance:
























dij = (xi − xj )2 + (yi − yj )2

Define the property similarity between two objects Oi , Oj as

1 n
sij = | pik − p jk |
n k=1

Now suppose that the properties of objects change over time due to internal or external factors. For
example, if we take the objects under consideration to be plants and animals, then they may natu-
rally grow in size with seasons (internal factor) and could be affected by natural events such as fire
and disease at particular locations (external factor) as well. Denote object Oi ’s properties at time t1
as ( pt11 , pt21 , pt31 . . . ptn1 ), t2 as ( pt12 , pt22 , pt32 . . . ptn2 ) and so forth, and create a new variable to represent
the property change of Oi from time tq−1 to time tq :

t −tq−1 1 n
t t
DOiq = | p q − pikq−1 |
n k=1 ik

Now we create a vector to represent the entire sequence of temporal change from time t1 to time tq :
t −tq−1
= (DOti2 −t1 , DOti3 −t2 , . . . , DOiq
change
Oi )
We are interested in the covariance of the overall temporal change in properties of any two objects
change change
Oi and Oj , Cov(Oi , Oj ), as this variable plays a crucial role in representing the manipulative
aspect of sympathetic magic. With the above notations, we can now formally express the law of
similarity and the law of contagion as:
 RELIGION, BRAIN & BEHAVIOR 5

change change
Similarity: an increase in sij causes an increase in Cov(Oi , Oj )
change change
Contagion: a decrease in dij causes an increase in Cov(Oi , Oj )

The above formalizations state that making two objects more similar or physically closer will cause
them to co-vary more regarding their properties. If one intends to induce a certain kind of change in
Oj (usually a person), according to these magical principles they only need to obtain an object Oi
that is either in closer contact with Oj or similar to Oj and then manipulate Oi . In practice, of course,
the physical proximity of Oi and Oj usually occurs some time in the past, because if Oi and Oj are
physically in contact in present time, one may either directly manipulate Oj (e.g., performing heal-
ing rituals directly on the person instead of his bodily parts/clothes) or not wish to perform the
manipulation in front of Oj (e.g., induce harm in Oj when Oj is a person).

3. The natural environment and human cognition that give rise to manipulative
sympathetic magic
Why do people think a change in similarity or physical proximity would have a causal relationship
with their co-variation? As alluded to in the Introduction, I argue that (1) in the environment where
humans live, objects that are physically closer and/or similar do tend to co-vary in their properties,
and (2) humans have a strong tendency to detect patterns and establish causality and may mistake
change change
correlation as causation (Matute et al., 2015). Because Cov(Oi , Oj ) correlates with both dij
and sij , humans may mistakenly believe that change in dij and sij will cause changes in
change change
Cov(Oi , Oj ). For example, an increase in dij would lead to an increase in
change change change change change
Cov(Oi , Oj ). If Oi is externally manipulated, the only way for Cov(Oi , Oj ) to
change
increase is for Oj to change in the same direction to an extent such that its change matches
change
the change in Oi even more.
In plain language, what I am suggesting above is that objects that are physically close or similar in
their attributes are more likely to co-vary, and therefore people may believe that by making two
things closer or similar they can induce such co-variation, and when people have control over
one object, they may mistake correlation as causation again in thinking that manipulating this
object will cause similar changes in the other object such that the co-variation is increased or main-
tained. Take the voodoo doll as an illustrative example; we may characterize the procedure as hav-
ing two separate steps. To induce harm in an enemy, step one is to create a voodoo doll (a symbolic
object) that either resembles the enemy in some aspect or made of something that was once phys-
ically close to the enemy such as clothes, hair, or fingernails, with the rationale being that creating
similarity or physical proximity causes co-variation; step two is to harm the voodoo doll (induce
change in the symbolic object), with the rationale being that changing one object causes change
in the other object in the same way or direction.
There is a large empirical literature on humans’ tendency to mistake correlation for causation in
domains such as health (Oh, 2016), finance (Heyns & Vlok, 2014), environmental management
(Hilborn, 2016), and scientific research (Rohrer, 2018). Indeed, mistaking correlation for causation
is such a rampant phenomenon that researchers have lamented that this cognitive bias “leads us
astray practically every day” (Dobelli, 2013, p. 110). On the theory front, mechanisms that allows
for the evolution of hyper causation detection have been proposed, particularly in the context of
superstitious behaviors. Foster and Kokko (2009), for example, discover that natural selection
can favor strategies that lead to errors (assigning causality between two events when there is
none) as long as the occasional correct response carries a large fitness benefit. Abbott and Sherratt
(2011) model a situation where individuals need to decide whether to exploit (act to maximize
fitness given the available information regarding the causal relationship between action and out-
come) or to explore (act to generate more information about the true nature of causal relationship)
and find that superstitious behaviors (exploiting a non-existent causal relationship) may evolve
when the cost of superstition is low relative to the perceived benefits. Given the large literature
6 Z. HONG

on this topic, I shall not further belabor this point, and will mainly discuss the plausibility and appli-
cability of the first point that objects in contact (more generally, objects in close physical proximity)
or objects that are similar tend to co-vary with regard to their attributes.
On the relationship between similarity and co-variation, human cognition systems tend to cat-
egorize objects into different kinds (Harnad, 2017) according to their background theories of the
world, and these categories often match pretty well with “natural” kinds in the world, with folk
biology being a prime example (Atran, 1999). In the environment where humans live, it is often
true that things that are of the same kind (1) change similarly in a temporal fashion (e.g., plant
organisms grow with seasons) and (2) respond to external factors in similar ways (e.g., certain infec-
tious disease would cause illness in similar types of animals, such as those of the same species,
genus, or family). More generally, because the change induced by external factors on objects
often heavily depends on the properties of these objects, those with similar properties would natu-
rally respond in similar ways. Importantly, things of the same kind tend to occur in close proximity
and resemble each other in properties. Our cognitive tendency to detect correlations and mistake
them as causations, therefore, may pick up such environmental regularities and form the manipu-
lative sympathetic magical intuitions over developmental time. Note that the formation of such a
heuristic bias need not require actual causation but only perceived causation: for example, observ-
ing that domesticated animals (which are physically proximate) falling ill one after another may
lead to the impression that the illness of one animal causes the illness of another animal,6 and
with one more inferential step we get manipulative sympathetic magic: inducing illness in one ani-
mal causes illness in a different animal that is physically proximate.
Before discussing the relationship between physical proximity and co-variation, let us first turn
to a related claim that physically closer objects tend to be similar in attributes, or more formally, dij
negatively correlates with sij . To argue for this claim, I draw on the concept of spatial autocorrela-
tion which essentially means that geographically nearby values of a variable tend to depend on one
another (Getis, 2010). Although in theory spatial autocorrelations can be both positive and negative,
in reality, positive spatial autocorrelations vastly outweigh negative ones (Chun & Griffith, 2018;
Griffith, 2019). For example, negative spatial autocorrelation is found in only 8 out of 361 agricul-
tural plant breeding trials (Wu et al., 1998) and 80 out of 2801 US intra-county population density
geographic distributions (Griffith et al., 2003). This means that we are more likely to observe nearby
objects to be similar than dissimilar; in other words, physical proximity of objects correlates with
their similarity. Ethnographically, people also notice the co-localization of things that are of similar
properties. The popular Chinese proverb “things find one another according to their kinds” (物以
类聚) first appeared in Zhan Guo Ce (5th–3rd century BCE) and in English there is the familiar
saying of “birds of a feather flock together.”
change change
Now, since we have already argued for the correlation between Cov(Oi , Oj ) and sij as
change change
well as sij and dij and if we take these arguments as valid, Cov(Oi , Oj ) and dij are likely
to be correlated as well (Lian et al., 2020), especially when the first two correlations are high (Castro
Sotos et al., 2009). Of course, we need to bear in mind that strictly speaking the transitivity of cor-
relation is not a mathematical guarantee (Langford et al., 2001). There are also more straightfor-
ward reasons why physically proximate objects tend to change together: usually, the extent to
which objects are affected by an external factor depends on their proximity to it. For example, in
a forest fire, the closer objects are to the center of the fire, the more damage will be incurred
(death of living organisms or burning of inanimate objects). The aforementioned example of infec-
tious disease applies here as well: it is a well-known fact in both epidemiology (Lawson et al., 2016)
and folk intuition (Brown et al., 2011) that physical proximity is one of the primary factors in pre-
dicting who will get infected.
Though people in traditional, small-scale societies rarely theorize their belief systems and every-
day practices in explicit terms, in societies with a literate tradition we do occasionally find explicit
theorizing of the relationship between these variables, especially similarity and covariation. Con-
sider, for example, ancient Chinese culture. The Confucian scholar Dong Zhongshu (179–104
 RELIGION, BRAIN & BEHAVIOR 7

BCE) formally proposed the principle that “things of the same kind activate one another” (Dong,
179–104 BCE/2015):
Now if you pour water on level ground, it will avoid the dry area and run to the wet area, but if you expose two
similar pieces of firewood to fire, the fire will avoid the wet piece and go to the dry one. All things avoid what is
different from them and follow what is similar to them. Therefore, if qi are the same in kind, they will come
together; if [musical] tones match, they will respond to each other … This has nothing to do with spirits. Their
regularities make them so. A beautiful thing calls forth things that are beautiful in kind; an ugly thing calls
forth things that are ugly in kind, for things of the same kind arise in response to each other. For example,
when a horse neighs, horses will respond; when an ox lows, oxen will respond.

Bear in mind that Dong’s theorization occurred two millennia before Frazer, and it importantly
differs from Frazer’s writing in that Dong genuinely believes that such a principle is factually cor-
rect. Dong’s main point here is that things of the same kind (with the same type of qi) will “activate”
one another, and he supplies his sweeping claim (“all things avoid what is different from them and
follow what is similar to them”) with some musical and biological examples. Interestingly, some of
Dong’s examples are factually true: when a horse neighs, other horses will indeed likely respond
with neighs either because they are communicating with one another or because they are all
responding to some common stimulus. The overall conclusion, however, has a distinctive causal
flavor. It is worth noting that Dong’s writing has not only since been very influential in the learned
circle and was extensively discussed by later scholars, but also was perceived as sensible and intui-
tive by the lay people and exerted a great influence on their everyday life. For example, the principle
of “things of the same kind activate each other” permeates traditional Chinese medicine (Jiang et al.,
2018), and his proposed method of rainmaking based on the same principle was practiced as late as
the Qing dynasty (1636–1911 CE) (Liu, 2013).

4. Two additional magical principles: similarity induced closeness and proximity

induced resemblance
Although in both Frazer’s original formulation and subsequent treatment of sympathetic magic
similarity and contagion/proximity are treated as distinctive types of magic, as the above analysis
have shown, there is a deeper connection between similarity and proximity in that they themselves
are correlated as well. This suggests that the human tendency of mistaking correlation as causation
should, in theory, produce two other types of magical principles:

Similarity induced closeness: Making two objects more similar will cause them to be physically
closer.
Proximity induced resemblance: Making two objects physically closer will cause them to be similar.

To be fair, in The Golden Bough, Frazer does provide some “similarity induced closeness” examples,
but he groups them under the same “like produced like” category and does not offer additional
explanations (Frazer, 1890):
The Indians of British Columbia live largely upon the fish which abound in the seas and rivers. If the fish do
not come in due season, and the Indians are hungry, a Nootka wizard will make an image of a swimming fish
and put it into the water in the direction from which the fish generally appear. This ceremony, accompanied
by a prayer to the fish to come, will cause them to arrive at once. (p. 18)

In this case, the indigenous people are not manipulating the image fish in the sense of causing it to
change in its properties; rather they simply place it at a certain location hoping that real fish will also
appear in that location. Later in the text Frazer uses the more explicit language of “attraction” (Fra-
zer, 1890):
The Toradjas of Central Celebes believe that things of the same sort attract each other by means of their
indwelling spirits or vital ether. Hence they hang up the jawbones of deer and wild pigs in their houses, in
8 Z. HONG

order that the spirits which animate these bones may draw the living creatures of the same kind into the path
of the hunter. (p. 18)

This “things of the same kind attract each other” idea deserves a special attention, because it was
explicitly theorized in other cultures. Wang Chong (27CE–97CE), a skeptic thinker of Eastern Han
dynasty of China, made a very similar point on fake fish attracting real fish:
The fisherman carved wood into a fish-shape and painted the fish with red lacquer. When the wooden fish
floats against the water and stirs it, fish [in the water] think it is real and swim towards it to meet it. (Lunheng,
chapter 47)

Here, Wang Chong’s comments on wooden fish attracting real fish occurs in a larger context where
he defends the thesis that one can “seek to obtain something using something that is similar” (以类
求之). In medieval Europe, thinkers such as Girolamo Fracastoro also believed that the attraction
and repulsion of two bodies takes place on the basis of their similarity and dissimilarity, respectively
(Nejeschleba, 2006).
The other type of magic, proximity induced resemblance, is simply the reversing of the causal
arrow of similarity induced closeness. To reiterate, it states that by making two things closer they
will become more similar. In a sense, this is same the type of contagious magic that Paul Rozin
and colleagues discussed where they focus on the psychological discomfort of the participants
when the negative properties of some objects “diffuse” into other objects that they have to inter-
act with. Ethnographically, this magic principle mostly manifests itself in the negative form, that
is, if one contacts or comes into close proximity with something perceived to be negative, they
will “acquire” the negativity which will lead to some kind of misfortune and therefore such con-
tacts should be avoided. I shall note a positive use of such a magical principle from my own field-
work experience with the Yi. The Yi in Liangshan area will sometimes diagnose the location and
nature of the illness of a patient by using a chicken to sweep over the patient’s body and then
sacrifice the chicken by drowning it in water. After the chicken is dead a dissection is performed
and it is believed that abnormalities found in the chicken body correspond to pathologies
(usually hidden) in the patient. For example, a broken bone in the chicken wing would signify
a broken bone in the patient’s arm. The local people would jokingly tell me that this method is
the “X-ray of the Yi”.7
The kinds of magical principles discussed so far can be shown in a simple diagram (Figure 2).
Unlike previous accounts of sympathetic magic, this graphical representation of the four principles
clearly illustrates the different variables involved and the direction of causality. Note that we don’t
see causal arrows going from covariation to similarity of physical proximity because when it is

Figure 2. Types of magical principles due to a combination of environmental regularities (correlation amongst similarity, proxi-
mity, and co-variation) and human cognitive tendency (mistaking correlation as causation).
 RELIGION, BRAIN & BEHAVIOR 9

possible to induce covariation in both objects we presumably have both of them at hand, and to
induce similarity it would be much easier to change one so that it looks more like the other; and
to induce proximity they could simply be placed by each other. Therefore, in practice we almost
always see the causal arrow going towards rather than from covariation.

5. Discussion
In this paper, I have argued that the intuitive plausibility of manipulative sympathetic magic is a
result of the regularities in the environment (the correlation between physical proximity, similarity,
and covariation) and a by-product of human causal cognition (mistaking correlation as causation),
Again, this paper does not attempt a general theory of magic (broadly defined); practices directed
towards supernatural entities, for example, clearly have rather different underlying psychological
mechanisms such as agency detection (Andersen, 2019), intuitive dualism (Bloom, 2007), and men-
talizing (Barrett, 2004; Guthrie, 1995). Like beliefs in god which has been suggested by some8 to be
evolutionary by-products (Boyer & Bergstrom, 2008), I argue that manipulative sympathetic magic
is also best viewed as a by-product of human cognitive evolution and does not serve adaptive func-
tions. Cultural evolution, however, may shape the performative aspects of magic in ways that fit our
genetically evolved (Singh, 2017) or culturally transmitted intuitions (Hong, forthcoming). In rea-
lity, the above four sympathetic magic principles are almost always used in combination with a
number of other features to increase the intuitive plausibility of magical practices. Among the Yi
in Southwest China, for example, when diagnosing causes of illness shamans (called suni) would
use both sympathetic magic (rubbing an egg against one’s body, dropping the egg yolk into a
bowl with water, and examining the patterns) and an animistic agent (a benevolent spirit which
is said to attach itself to the shaman) in order to identify the causative ancestral ghost (Hong, sub-
mitted). Therefore, it may be difficult to pinpoint these magical principles from actual ethnographic
observations.
Although the present account provides a novel way of understanding the nature and function of
manipulative sympathetic magic, it nonetheless presents a few challenges. Below I address these
difficulties and discuss some alternative explanations.

5.1. What counts as “being similar”?

In the context of sympathetic magic, similarity serves as an important explanans and as such an
independent account of how humans perceive similarity is crucial. Yet, the notion of similarity
is in fact a very difficult concept both philosophically and psychologically (Goodman, 1972/2012;
Tversky, 1977). Traditionally, similarity is often invoked in the discussion of categorization (Gold-
stone, 1994) and analogical reasoning (Gentner & Markman, 1997) where much effort has been
devoted to understanding the specific features and relational structures that make two entities simi-
lar. However, human similarity judgements are highly dependent on context, and require not only
selection of relevant features but some weighing of their relative importance. People with different
backgrounds will not only pay attention to different features but also weigh them differently when
deciding the extent to which two objects are similar (Goodman, 1972/2012). Experimental evidence
shows that similarity judgment of human subjects often exhibit substantial flexibility and is suscep-
tible to context manipulation: for example, raccoon and snake were judged to be less similar when
no explicit context was provided than when a context was created by placing the word pets above
the comparison (Barsalou, 1982). Additionally, human subjects often judge similarity between
objects in an asymmetrical manner: for example, North Korea is typically judged to be more similar
to China than China is to North Korea (Tversky & Gati, 1982). To address these theoretical issues
and empirical observations, many models have been proposed on how exactly humans process sen-
sory input information and categorize objects into different groups (Decock & Douven, 2011;
Medin, 1989; Verguts et al., 2004).
10 Z. HONG

More recently, advances in brain science and cognitive neuroscience have shed considerable
light on cognitive mechanisms of similarity perception, such as deep convolutional neural networks
where sophisticated computational models of brain representations are developed to mimic human
similarity judgment and categorization (He et al., 2015; Simonyan & Zisserman, 2015). Data driven
approaches have also helped to reveal the underlying dimensions of human similarity judgment:
Hebart et al. (2020), for example, identified 49 meaningful dimensions (e.g., animals, colorful, cir-
cular, fire) using a comprehensive dataset of real-world images and over one million triad responses
generated by online workers.
In the mathematical formulation in Section 2, I essentially used a simplified version of the geo-
metric model of similarity which is known to be subject to a number of problems (Decock & Dou-
ven, 2011). This is for illustrative purpose only, and I make no strong claims about the human
psychological reality of similarity computation and/or categorization. As far as my argument is con-
cerned, the ways humans judge similarity for the purpose of manipulative sympathetic magic need
only roughly match similarity of objects in nature regarding their tendency to change in a synchro-
nized manner. In practice, of course, we may observe objects used to activate or influence each other
that are “similar” only in a very minimal sense by most accounts (Mauss, 1902/2001). This is
because during cultural evolutionary time humans may culturally construct theories that group cer-
tain objects together for purposes (in cognitive science, this is sometimes referred to as the “theory
theory,” see (Gopnik, 2003; Murphy & Medin, 1985)) that has nothing to do with sympathetic
magic. Once these objects are recognized as belonging to the same kind, however, they may then
be (mis)used for manipulative sympathetic magic. For example, dragons and rain by most natur-
alistic accounts (and to most westerners) have very little in common, yet in traditional Chinese cul-
ture these two types of objects share some fundamental properties and therefore are believed to be
of the same kind. As such, traditional Chinese rainmaking activities frequently involve the use of
dragon images (Hong, forthcoming).

5.2. Is contact the same as physical proximity?

So far in the paper, I have been using “contact” and “physical proximity” more or less interchange-
ably. Both Frazer himself and subsequent scholars, however, emphasize “contact” as an essential
component of sympathetic magic (hence the name “law of contagion”). In an obvious sense, the
most physically proximate two objects can be is actual contact, yet for sympathetic magic to
work, the two objects need only be recognized as belonging to the same system by virtue of physical
proximity. Neuroscientific work in “systematizing mechanism” shows that the human brain may
focus on one detail (the input) and observes what happens to the input when it is manipulated
by just one factor (the operation), and logs the result of the transformation of the input by the oper-
ation (the output) (Baron-Cohen et al., 2003), and such a cognitive process likely leads humans to
see lawful, potentially causal patterns (Baron-Cohen & Lombardo, 2017). In proximity induced
resemblance, proximity itself is often sufficient to induce change; for example, in rural Ghana in
the 1940s, the potency of certain medicine was believed to be spoiled by the proximity of a men-
struous woman in the house (Field, 1970). Similarly, the war scouts of the Kpelle people in Western
Africa in the early twentieth century would carry “magic bananas” which would indicate the proxi-
mity of the enemy through sympathetic response (it would be too late to wait for contact!) (Wester-
man & Schütze, 1921). In South Asia, the Telugu believe that auspicious influences can be
transmitted by both touch and proximity (my emphasis) (Tapper, 1987). Regarding manipulative
magic depending on the principle of contagion, it is true that most practices recorded ethnographi-
cally and historically involve contact at least in the folk sense. I suggest that this is because amongst
the many objects that are physically proximate to the focal object, the ones that actually are in con-
tact with the focal object are much more cognitively salient (and more likely to be recognized as
belonging to the same system) and may enjoy an advantage during cultural transmission (Claidière
& Sperber, 2007; Sperber, 1996).
 RELIGION, BRAIN & BEHAVIOR 11

5.3. What about the transfer of essence?

In both classical anthropological accounts and later work on magic, one explanation of the law of
contagion that frequently emerges in the literature is the transfer of essence or vital properties
through contact. To be sure, my account does not involve any transfer of properties, and I offer
two arguments regarding the relevance of the “transfer of essence” idea. First, it may be a result
of an extremely common human thinking habit, post hoc rationalization (Cushman, 2019) of the
intuition that objects that are in contact/physically proximate influence each other. In other
words, the mechanism that people come up with to justify their intuitions may not explain why
they have such intuitions in the first place. Second, while transfer of essence does offer a plausible
mechanism for proximity induced resemblance, it does not really explain manipulative magic, at
least not in any straightforward way. Why would objects that share the same essence causally co-
vary? One possibility is that once the essence of an object is transferred to another object, the
two objects become similar and we are back at the puzzle of manipulative magic based on similarity.
Sørensen (2007), in his extensive treatment of magic from cognitive perspectives, proposes that the
transfer of essence establishes a kind of “essence link” such that the objects involved may sympath-
etically respond to each other. This account unfortunately still begs further questions: why would
contact establish an essence link, and why does the essence link make the two objects respond to
each other’s change in similar ways?9 As a result, it feels more of a re-description of the phenom-
enon rather than a genuine explanation.

5.4. The possibility of false generalization

The present account does not rule out an important alternative explanation: false generalization
based on a few examples. The idea is that in nature there are indeed cases where things activate
each other at a distance. Dong, for example, in illustrating of principle of “things of the same
kind activate one another,” uses the phenomenon of sympathetic resonance in musical instruments
with strings: “Pluck the note gong, and other gong notes will respond to it; pluck the note shang, and
other shang notes will respond to it.”10 Indeed, many early Chinese writers used the actual obser-
vation that if one string of an instrument is plucked, a similarly tuned string on a nearby instrument
will vibrate (Schliesser, 2015). Western scholars such as Plotinus (204–270 CE) and Francis Bacon
(1561–1626 CE) also noticed the sympathetic resonance of strings in musical instruments (Schlies-
ser, 2015). Plotinus, in particular, uses sympathetic vibration in music to argue for the efficacy of
prayers (Gurtler, 2015):
But the sun, or another heavenly body, does not hear his prayer. And that which he prays for comes about
because one part is in sympathetic connection with another, just as in one tense string … when one string
is plucked, another has a kind of sense of this by its concord and the fact that it is tune to the same scale.
But if the vibration can even pass from one lyre to another in so far as a sympathy exists, then there is also
one single harmony in the All … (IV.4.41.1-4)

Plotinus here jumps from the existence of sympathetic resonance between strings to “one single
harmony in All,” a quite hasty generalization. We now know through modern acoustics and physics
that such sympathetic resonance occurs when the natural oscillatory periods of sonorous vibrators
are harmonically related and the vibrations are isochronous (Schliesser, 2015), and is a rather
specific case and thus would be a poor foundation to construct grand generalizations.
The most typical natural example of sympathetic action, of course, is magnetism. This phenom-
enon has attracted much theorizing from scholars and thinkers across cultures and historical times.
In the same paragraph where Wang Chong makes the case that fake fish can attract real fish, he also
uses magnetism to justify “like activates like”: “sea turtle shells, after being rubbed, may attract small
and light items; magnet stones can attract needles made of iron. This is because they are all really of
the same kind, not different kinds” (Lunheng, chapter 47). In the western tradition, magnetic attrac-
tion was similarly used to illustrate the principles (or at least the plausibility) of sympathetic
12 Z. HONG

influence (Nejeschleba, 2006; Pollitt, 2019), including many notable enlightenment scientists such
as Margaret Cavendish, Anne Conway, and Gottfried Leibniz (Meyns, 2018).
Aside from music and magnetism, miscellaneous cases can also be grouped together under the
same category “sympathy.” Aristotle, for example, when discussing the idea of sympathy, starts with
yawn contagion11 (“Why do men generally themselves yawn when they see others yawn?”) yet he
also includes the urge to urinate when people are close to a river and the contagious spread of dis-
ease (Barnes, 2014).
All of this is to say that there are indeed cases where change in one object (spontaneous or
manipulated) induces change in other objects, and such cases were often used to infer some general
law of sympathetic action. Of course, pre-modern theorizing of sympathetic action in literate
societies is not exactly the same as Frazerian sympathetic magic, but they do share the underlying
logic that things (usually of the same kind) can influence each other in some way. The key difference
is perhaps that the enlightenment thinkers do not believe one can create deep similarity by means of
superficial resemblance. A voodoo doll, for example, would be an image and only an image in the
eyes of an enlightenment thinker, and as such would not be categorized as belonging to the same
kind as the real person.
I suggest that both mechanisms (mistaking correlation amongst similarity, proximity, and cov-
ariation as causation and false generalization) contribute to the phenomenon of manipulative sym-
pathetic magic, though their relative importance may depend on the specific historical and cultural
contexts. The limitation of the false generalization account may be that it does not address the
manipulative aspect of magical actions as much as the correlation-causation account, as there is
nothing intrinsic in this account that requires change in one object leading to change in another
object in the same direction. In medieval theorizing of sympathetic actions, for example, sympathy
is often discussed in conjunction with antipathy, and manipulation of objects mainly affects the
presence or absence of their sympathetic power (e.g., garlic deprives magnets of their attractive
power (Sander, 2019)) rather than the corresponding change in objects that are in sympathetic
relationships with them. In any case, it is perhaps not surprising that multiple cognitive and cultural
mechanisms sustain a phenomenon as rich and prevalent in human societies as sympathetic magic.
Nonetheless, we should keep in mind that the principles of manipulative sympathetic magic do not
account for all magic/superstitious practices in a broad sense: as already mentioned, practices such
as deity worship clearly requires quite different psychological bases from those required for sym-
pathetic magic. This paper also only addresses the evolved intuition aspect12 of magic practices
as it explains why cultural practices with certain forms are more successful than others (Miton
et al., 2015). Obviously, people must have encountered numerous empirical failures when attempt-
ing to use any magic (in fact, one of the proposed defining features of magic is that it doesn’t work,
see Lindeman and Svedholm (2012)), yet empirical failures can be easily explained away and rarely
fundamentally challenge the validity of culturally accepted practices13 (Hong, forthcoming; Hong
et al., forthcoming).

6. Conclusion
In this paper, I formalize Frazer’s classic account of magic and argue that the puzzling practice of
sympathetic magic can be explained by a combination of environmental regularities and the human
tendency to mistake correlative patterns as causality. I draw from a rich source of ethnographic and
historical record in support of the plausibility of the account and at the same time suggest that a full
understanding of sympathetic magic may require taking into consideration multiple mechanisms.

Notes
1. Although Tylor (1871) had already described magic practice in traditional, small scale societies and alluded to
some general principles, he never formally proposes sympathetic magic the way Frazer did.
 RELIGION, BRAIN & BEHAVIOR 13

2. Today, we know that he was almost certainly wrong about the magic-religion-science developmental stages of
human social and cultural evolution based on ethnographic and archeological knowledge accumulated over
the past century.
3. Though it is possible that some of the “sympathetic magic” usage during the 20th century occurred in the con-
text of criticizing it for not being a useful analytic category, it is highly unlikely that such uses of the term could
account for such a consistently high frequency of its occurrence.
4. It should be pointed out that the phenomenon of quantum entanglement in physics resembles manipulative
sympathetic magic principles, where change in one particle may induce change in a different particle at some
considerable distance (Popkin, 2018). However, quantum entanglement was discovered rather late in human
history, and it was unlikely that such phenomena (often unobservable to the naked eye) would have influenced
the cognitive evolution of our species.
5. Note that to my knowledge this is the only example of experimental evidence showing manipulative
sympathetic magic in contemporary western societies. However, the fact we usually do not observe
such biases in contemporary modern societies does not mean these biases do not exist; rather, they
may be triggered in particular social and cultural context and under the appropriate conditions lead
to the development of concrete beliefs and practices. As I alluded to in the main text, The prevalent
mechanistic worldview in modern societies may have profoundly influenced our intuitions regarding
what’s possible (manipulative sympathetic magic would be deemed largely impossible); additionally,
people’s everyday inferences are rarely affected by manipulative sympathetic magical intuitions (simi-
larity and contagion) alone, but rather almost always a result of multiple psychological, social, and
cultural factors.
6. In fact, of course, such covariation is caused by unobserved variables (pathogens).
7. There are very few informants who think that in performing this ritual the illness is transmitted to the
chicken as well (therefore the patient is healed after such transfer). However, most people seem to
think that it only diagnoses rather than cures, and additional healing procedures (usually in the form
of propitiating the aggressive ancestral spirits and sending them away) need to be performed to cure
the patient.
8. For a discussion on the adaptationist-byproduct debate on the evolution of religion, see Sosis (2009).
9. Presumably, “essence link” could be defined this way, but this would not add to our understanding of the
phenomenon.
10. Both gong and shang are notes of traditional Chinese pentatonic scales.
11. Note that unlike the previous two examples of cosmic harmony where different events/entities are linked via a
mechanistic pattern (under the implicit assumption of the fundamental patterns in the physical universe), the
yawning case is based on social perception of empathy that is also observed in other primates (Anderson et al.,
2004).
12. More precisely, the present paper only accounts for evolved intuition regarding sympathetic magical actions:
i.e., why we find similarity and contagion as attractive causal principles. For other work that looks at the
evolved aspect of magical thinking, see Legare and Souza (2012), Singh (2017), and Boyer (2020).
13. Keith Thomas, the celebrated historian of medieval magic comments that “ … once their initial premises are
accepted, no subsequent discovery will shake the believer’s faith, for he can explain it away in terms of the
existing system. Neither will his convictions be weakened by the failure of some accepted ritual to accomplish
its desired end, for this too can be accounted for … The reaction against magic could thus never come from
the cumulative resentment of disappointed clients. It had to arise from outside of the system altogether” (Tho-
mas, 1971/2003, pp. 767–768).

Acknowledgements
I thank Manvir Singh for his continued encouragement, Joseph Henrich for providing helpful comments, and Tiffany
Hwang for carefully proofreading the final draft of the manuscript.

Disclosure statement
No potential conflict of interest was reported by the author(s).

ORCID
Ze Hong http://orcid.org/0000-0002-5343-3008
14 Z. HONG

References
Abbott, K. R., & Sherratt, T. N. (2011). The evolution of superstition through optimal use of incomplete information.
Animal Behaviour, 82(1), 85–92. https://doi.org/10.1016/j.anbehav.2011.04.002
Andersen, M. (2019). Predictive coding in agency detection. Religion, Brain and Behavior, 9(1), 65–84. https://doi.
org/10.1080/2153599X.2017.1387170
Anderson, J. R., Myowa-Yamakoshi, M., & Matsuzawa, T. (2004). Contagious yawning in chimpanzees. Proceedings
of the Royal Society B: Biological Sciences, 271(suppl), S468–S470. https://doi.org/10.1098/rsbl.2004.0224
Armitage, N. (2015). European and African figural ritual magic: The beginnings of the voodoo doll myth. In
C. Houlbrook & N. Armitage (Eds.), The materiality of magic: An artifactual investigation into ritual practices
and popular beliefs (pp. 85–100). Oxbow Books.
Atran, S. (1999). The universal primacy of generic species in folkbiological taxonomy: Implications for human bio-
logical, cultural, and scientific evolution. In R. Wilson (Ed.), Species: New interdisciplinary essays (pp. 231–261).
MIT Press.
Barnes, J. (2014). The complete works of aristotle. In The Complete Works of Aristotle.
Baron-Cohen, S., & Lombardo, M. V. (2017). Autism and talent: The cognitive and neural basis of systemizing.
Dialogues in Clinical Neuroscience, 19(4), 344–352. https://doi.org/10.31887/dcns.2017.19.4/sbaroncohen
Baron-Cohen, S., Richler, J., Bisarya, D., Gurunathan, N., & Wheelwright, S. (2003). The systemizing quotient: An
investigation of adults with Asperger syndrome or high-functioning autism, and normal sex differences.
Philosophical Transactions of the Royal Society B: Biological Sciences, 358(1430), 361–374. https://doi.org/10.
1098/rstb.2002.1206
Barrett, J. (2004). Why would anyone believe in God. AltaMira Press.
Barsalou, L. W. (1982). Context-independent and context-dependent information in concepts. Memory & Cognition,
10(1), 82–93. https://doi.org/10.3758/BF03197629
Bleske-Rechek, A., Morrison, K. M., & Heidtke, L. D. (2015). Causal inference from descriptions of experimental and
non-experimental research: Public understanding of correlation-versus-causation. Journal of General Psychology,
142(1), 48–70. https://doi.org/10.1080/00221309.2014.977216
Bloom, P. (2007). Religion is natural. Developmental Science, 10(1), 147–151. https://doi.org/10.1111/j.1467-7687.
2007.00577.x
Boyer, P. (2020). Why divination? Evolved psychology and strategic interaction in the production of truth. Current
Anthropology, 61(1), 100–123. https://doi.org/10.1086/706879
Boyer, P., & Bergstrom, B. (2008). Evolutionary perspectives on religion. Annual Review of Anthropology, 37(1), 111–
130. https://doi.org/10.1146/annurev.anthro.37.081407.085201
Bremmer, J. (1999). The birth of the term “magic.” Zeitschrift Für Papyrologie Und Epigraphik, 126, 1–12.
Brown, P., Armelagos, G., & Maes, K. (2011). Humans in a world of microbes: The anthropology of infectious disease.
In M. Singer & P. Erickson (Eds.), A companion to medical anthropology (pp. 253–270). Blackwell Publishing.
https://doi.org/10.1002/97814443953
Castro Sotos, A. E., Vanhoof, S., van den Noortgate, W., & Onghena, P. (2009). The transitivity misconception of
Pearson’s correlation coefficient. Statistics Education Research Journal, 8(2).
Chun, Y., & Griffith, D. A. (2018). Impacts of negative spatial autocorrelation on frequency distributions. Chilean
Journal of Statistics, 9(1), 3–17.
Claidière, N., & Sperber, D. (2007). Commentary: The role of attraction in cultural evolution. Journal of Cognition
and Culture, 7(1–2), 89–111. https://doi.org/10.1163/156853707X171829
Cushman, F. (2019). Rationalization is rational. Behavioral and Brain Sciences, 43, e28–e28. https://doi.org/10.1017/
S0140525X19001730
Decock, L., & Douven, I. (2011). Similarity after Goodman. Review of Philosophy and Psychology, 2(1), 61–75. https://
doi.org/10.1007/s13164-010-0035-y
Dobelli, R. (2013). The art of thinking clearly: Better thinking, better decisions. HarperCollins Publishers.
Dong, Z. (2015). Luxuriant gems of the spring and autumn. Columbia University Press.
Field, M. J. (1970). Search for security: An ethno-psychiatric study of rural Ghana. In Norton library (p. 478).
W. W. Norton & Company. https://ehrafworldcultures.yale.edu/document?id=fe12-027.
Foster, K. R., & Kokko, H. (2009). The evolution of superstitious and superstition-like behaviour. Proceedings of the
Royal Society B: Biological Sciences, 276(1654), 31–37. https://doi.org/10.1098/rspb.2008.0981
Frazer, J. G. (1890). The Golden Bough: a study in comparative religion (Vol. 2). Macmillan.
Gentner, D., & Markman, A. B. (1997). Structure mapping in analogy and similarity. American Psychologist, 52(1),
45–56. https://doi.org/10.1037/0003-066X.52.1.45
Getis, A. (2010). Spatial autocorrelation. In M. Fischer & A. Getis (Eds.), Handbook of applied spatial analysis (pp.
255–278). Springer.
Goldstone, R. L. (1994). The role of similarity in categorization: Providing a groundwork. Cognition, 52(2), 125–157.
https://doi.org/10.1016/0010-0277(94)90065-5
 RELIGION, BRAIN & BEHAVIOR 15

Goodman, N. (2012). The structure of appearance (Vol. 53). Springer Science & Business Media. (Original work pub-
lished 1972)
Gopnik, A. (2003). The theory theory as an alternative to the innateness hypothesis. In L. Antony & N. Hornstein
(Eds.), Chomsky and his critics (pp. 238–254). Blackwell Publishing. https://doi.org/10.1002/9780470690024.ch10
Griffith, D. A. (2019). Negative spatial autocorrelation: One of the most neglected concepts in spatial statistics. Stats, 2
(3), 388–415. https://doi.org/10.3390/stats2030027
Griffith, D. A., Wong, D. W. S., & Whitfield, T. (2003). Exploring relationships between the global and regional
measures of spatial autocorrelation. Journal of Regional Science, 43(4), 683–710. https://doi.org/10.1111/j.0022-
4146.2003.00316.x
Gurtler, G. M. S. J. (2015). Plotinus ennead IV. 4.30-45 & IV. 5: Problems concerning the soul. Parmenides Publishing.
Guthrie, S. E. (1995). Faces in the clouds: A new theory of religion. Oxford University Press on Demand.
Hanegraaff, W. J. (1998). The emergence of the academic science of magic: The occult philosophy in Tylor and
Frazer. In A. Molendijk & P. Pels (Eds.), Religion in the making (pp. 253–275). Brill.
Harnad, S. (2017). To cognize is to categorize: Cognition is categorization. In H. Cohen & C. Lefebvre (Eds.),
Handbook of categorization in cognitive science (pp. 21–54). Oxford: Elsevier Science & Technology. https://doi.
org/10.1016/B978-0-08-101107-2.00002-6
He, K., Zhang, X., Ren, S., & Sun, J. (2015). Delving deep into rectifiers: Surpassing human-level performance on
imagenet classification. Proceedings of the IEEE International Conference on Computer Vision, 1026–1034.
https://doi.org/10.1109/ICCV.2015.123
Hebart, M. N., Zheng, C. Y., Pereira, F., & Baker, C. I. (2020). Revealing the multidimensional mental representations
of natural objects underlying human similarity judgements. Nature Human Behaviour, 4(11), 1173–1185. https://
doi.org/10.1038/s41562-020-00951-3
Heyns, J. H., & Vlok, P. J. (2014). Correlation and causation: a potential pitfall for efficient asset management.
Hilborn, R. (2016). Correlation and causation in fisheries and watershed management. Fisheries, 41(1), 18–25. https://
doi.org/10.1080/03632415.2016.1119600
Hong, Z. (forthcoming). Dream interpretation from a cognitive and cultural evolutionary perspective: The case of
oneiromancy in traditional China. Cognitive Science.
Hong, Z., Slingerland, E., & Henrich, J. (forthcoming). Magic and empiricism in early Chinese rainmaking – A cul-
tural evolutionary analysis. Current Anthropology.
Hong, Z. Submitted. Ghost, divination, and magic among the Yi.
Hood, B. M., Donnelly, K., Leonards, U., & Bloom, P. (2010). Implicit voodoo: Electrodermal activity reveals a sus-
ceptibility to sympathetic magic. Journal of Cognition and Culture, 10(3–4), 391–399. https://doi.org/10.1163/
156853710X531258
Horton, R. (1993). Patterns of thought in Africa and the west : essays on magic, religion, and science. Cambridge
University Press.
Jiang, T., Bao, S., & Lv, Y. (2018). 董仲舒 “天人相应” 对《 伤寒论》 的影响. 中国中医基础医学杂志, 5.
Kida, T. E. (2009). Don’t believe everything you think: The 6 basic mistakes we make in thinking. Prometheus Books.
Kundt, R. (2017). Contemporary evolutionary theories of culture and the study of religion. Bloomsbury Publishing.
Langford, E., Schwertman, N., & Owens, M. (2001). Is the property of being positively correlated transitive? American
Statistician, 55(4), 322–325. https://doi.org/10.1198/000313001753272286
Lawson, A. B., Banerjee, S., Haining, R. P., & Ugarte, M. D. (2016). Handbook of spatial epidemiology. In M. D.
Ugarte, A. B. Lawson, R. P. Haining, & S. Banerjee (Eds.), Handbook of spatial epidemiology (pp. 3–39). Boca
Raton: Chapman and Hall/CRC. https://doi.org/10.1080/01621459.2021.1880230
Legare, C. H., & Souza, A. L. (2012). Evaluating ritual efficacy: Evidence from the supernatural. Cognition, 124(1), 1–
15. https://doi.org/10.1016/j.cognition.2012.03.004
Lian, R., Zhou, C., & Goertzel, B. (2020). The probabilistic support Kendall correlation and its transitivity properties.
Communications in Statistics – Theory and Methods, 49(2), 485–499. https://doi.org/10.1080/03610926.2018.
1543776
Lindeman, M., & Svedholm, A. M. (2012). What’s in a term? Paranormal, superstitious, magical and supernatural
beliefs by any other name would mean the same. Review of General Psychology, 16(3), 241–255. https://doi.org/
10.1037/a0027158
Liu, W. (2013). 清代求雨禳灾叙事的伦理意蕴与民俗信仰. 福建师范大学学报 (哲学社会科学版), 6(22).
Matute, H., Blanco, F., Yarritu, I., Díaz-Lago, M., Vadillo, M. A., & Barberia, I. (2015). Illusions of causality: How they
bias our everyday thinking and how they could be reduced. Frontiers in Psychology, 6, 888–888. https://doi.org/10.
3389/fpsyg.2015.00888
Mauss, M. (2001). A general theory of magic (Routledge classics). Routledge. http://www.amazon.co.uk/General-
Theory-Magic-Routledge-Classics/dp/0415253969. (Original work published 1902)
Medin, D. L. (1989). Concepts and conceptual structure. American Psychologist, 44(12), 1469–1481. https://doi.org/
10.1037/0003-066X.44.12.1469
Meyns, C. (2018). Sympathetic action in the seventeenth century: Human and natural. Philosophical Explorations, 21
(1), 60–75. https://doi.org/10.1080/13869795.2017.1421691
16 Z. HONG

Miton, H., Claidière, N., & Mercier, H. (2015). Universal cognitive mechanisms explain the cultural success of blood-
letting. Evolution and Human Behavior, 36(4), 303–312. https://doi.org/10.1016/j.evolhumbehav.2015.01.003
Murdock, G. P. (1980). Theories of illness a world survey. University of Pittsburgh Press. https://philpapers.org/rec/
MURTOI.
Murphy, G. L., & Medin, D. L. (1985). The role of theories in conceptual coherence. Psychological Review, 92(3), 289–
316. https://doi.org/10.1037/0033-295X.92.3.289
Nejeschleba, T. (2006). The theory of sympathy and antipathy in wittenberg in the 16th century. Acta Universitatis
Palackianae Olomoucensis, 7, 81–91.
Nemeroff, C., & Rozin, P. (2000). The makings of the magical mind: The nature and function of sympathetic magical
thinking. In K. S. Rosengren, C. N. Johnson, P. L. Harris, C. N. Johnson, & P. L. Harris (Eds.), Imagining the
Impossible (pp. 1–34). Cambridge University Press. https://doi.org/10.1017/cbo9780511571381.002
Oh, S. W. (2016). Vitamin D studies: Mistaking correlation for causation. Korean Journal of Family Medicine, https://
doi.org/10.4082/kjfm.2016.37.4.203
Pollitt, B. (2019). Sympathy, magnetism, and immoderate laughter: The feather in Cook’s last voyage. Art Bulletin,
101(4), 70–94. https://doi.org/10.1080/00043079.2019.1602454
Popkin, G. (2018). Einstein’s ‘spooky action at a distance’ spotted in objects almost big enough to see. AAAS—Science. https://
www.Sciencemag.Org/News/2018/04/Einstein-s-Spooky-Action-Distance-Spotted-Objects-Almost-Big-Enough-See.
Rohrer, J. M. (2018). Thinking clearly about correlations and causation: Graphical causal models for observational
data. Advances in Methods and Practices in Psychological Science, 1(1), 27–42. https://doi.org/10.1177/
2515245917745629
Rozin, P., Dunn, C., & Fedotova, N. (2018). Reversing the causal arrow: Incidence and properties of negative back-
ward magical contagion in Americans. Judgment and Decision Making, 13(5), 441–450.
Rozin, P., Markwith, M., & McCauley, C. (1994). Sensitivity to indirect contacts with other persons: AIDS aversion as
a composite of aversion to strangers, infection, moral taint, and misfortune. Journal of Abnormal Psychology, 103
(3), 495–504. https://doi.org/10.1037/0021-843X.103.3.495
Rozin, P., Millman, L., & Nemeroff, C. (1986). Operation of the laws of sympathetic magic in disgust and other
domains. Journal of Personality and Social Psychology, 50(4), 703–712. https://doi.org/10.1037/0022-3514.50.4.703
Rozin, P., & Nemeroff, C. (1990). The laws of sympathetic magic: A psychological analysis of similarity and conta-
gion. In J. W. Stigler, R. A. Shweder, & G. H. Herdt (Eds.), Cultural psychology: Essays on comparative human
development (pp. 205–232). Cambridge University Press.
Rozin, P., & Nemeroff, C. (2012). Sympathetic magical thinking: The contagion and similarity “heuristics.” In T.
Gilovich, D. Griffin, & D. Kahneman (Eds.), Heuristics and Biases (pp. 201–216). Cambridge University Press.
https://doi.org/10.1017/cbo9780511808098.013
Sander, C. (2019). Magnets and garlic: An enduring antipathy in early-modern science. Intellectual History Review, 30
(4), 523–560. https://doi.org/10.1080/17496977.2019.1648924
Schliesser, E. (2015). Sympathy: A History. Oxford University Press.
Simonyan, K., & Zisserman, A. (2015). Very deep convolutional networks for large-scale image recognition. 3rd
International Conference on Learning Representations, ICLR 2015 – Conference Track Proceedings.
Singh, M. (2017). The cultural evolution of shamanism. Behavioral and Brain Sciences, 41, 1–e66. https://doi.org/10.
1017/s0140525x17001893
Sosis, R. (2009). The adaptationist-byproduct debate on the evolution of religion: Five misunderstandings of the
adaptationist program. Journal of Cognition and Culture, 9(3–4), 315–332. https://doi.org/10.1163/
156770909X12518536414411
Sørensen, J. (2007). A cognitive theory of magic. Rowman Altamira.
Sperber, D. (1996). Explaining culture: A naturalistic approach. Blackwell. https://www.wiley.com/en-us/Explaining
+Culture%3A+A+Naturalistic+Approach-p-9780631200451.
Stanovich, K. E. (2009). What intelligence tests miss: The psychology of rational thought. Yale University Press.
Strenski, I. (2006). Thinking about religion: An historical introduction to theories of religion. Blackwell Publishing.
Tambiah, S. J. (1990). Magic, science, religion, and the scope of rationality. Cambridge University Press.
Tapper, B. E. (1987). Rivalry and tribute: Society and ritual in a Telegu village in South India. In Studies in sociology
and social anthropology (pp. xvi, 309). Hindustan Publishing Corporation. https://ehrafworldcultures.yale.edu/
document?id=aw17-002.
Thomas, K. (2003). Religion and the decline of magic: Studies in popular beliefs in sixteenth and seventeenth-century
England. Penguin UK. (Original work published 1971)
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift Für Tierpsychologie, 20(4), 410–433. https://doi.
org/10.1111/j.1439-0310.1963.tb01161.x
Tversky, A. (1977). Features of similarity. Psychological Review, 84(4), 327–352. https://doi.org/10.1037/0033-295X.
84.4.327
Tversky, A., & Gati, I. (1982). Similarity, separability, and the triangle inequality. Psychological Review, 89(2), 123–
154. https://doi.org/10.1037/0033-295X.89.2.123
 RELIGION, BRAIN & BEHAVIOR 17

Tylor, E. B. (1871). Primitive culture: Researches into the development of mythology, philosophy, religion, art, and cus-
tom (Vol. 2). J. Murray.
Verguts, T., Ameel, E., & Storms, G. (2004). Measures of similarity in models of categorization. Memory and
Cognition, 32(3), 379–389. https://doi.org/10.3758/BF03195832
Westerman, D. H., & Schütze, F. (1921). Kpelle: A negro tribe in Liberia. Vandenhoeck & Ruprecht/J. C. Hinrichs’sche
Buchhandlung. Https://ehrafworldcultures.yale.edu/document?id=fd06-007
Williamson, J. (2011). Mechanistic theories of causality part I. Philosophy Compass, 6(6), 421–432. https://doi.org/10.
1111/j.1747-9991.2011.00400.x
Wu, T., Mather, D. E., & Dutilleul, P. (1998). Application of geostatistical and neighbor analyses to data from plant
breeding trial. Crop Science, 38(6) 1545–1553. https://doi.org/10.2135/cropsci1998.0011183X003800060023x